Digital Pricing Strategy: Capturing Value from Digital Innovations 1032127724, 9781032127729

Citation preview

‘The “must-read” book for the C-level and executives, combining theories and best practices and providing practical considerations and guidance on how to drive digital transformation successfully.’ Jens Pfennig, Global Head of Pricing & Business Operations, Software AG ‘This book reinforces the message I have been spreading for many years—that digital innovations and data-driven opportunities need to be monetized and priced the right way! Data, analytics, and AI without a value mindset bring organizations no benefits.’ Bill Schmarzo, Dean of Big Data and author of The Economics of Data, Analytics, and Digital Transformation: The Theorems, Laws, and Empowerments to Guide Your Organization’s Digital Transformation ‘Pricing is a matter of life and death in the digital age. Digital transformation changes business models; it should also change pricing. This book is a trailblazer that has managed to bring together the knowledge of leading experts in highly relevant areas such as value-based pricing, the value of data, and artificial intelligence in pricing. A must-read for both service providers and service buyers to reach win-win.’ Ari Hirvonen, Chief Digital Officer, University of Jyväskylä, President of Finnish Universities CIO Network ‘This well-rounded book features a wealth of information and thought-leader expertise on all the digital pricing challenges your organization will face. By focusing on monetization, new business strategies, data projects, and real-world challenges, Digital Pricing Strategy: Capturing Value from Digital Innovations will give you and your team a blueprint for success for today and times ahead. Liozu and Hinterhuber have assembled a great roster of collaborators here to help guide companies forward.’ Kevin Mitchell, President of the Professional Pricing Society



Digital Pricing Strategy

Digital Pricing Strategy provides a best-practice overview of how companies design, analyze, and execute digital pricing strategies. Bringing together insights from academic and professional experts globally, the text covers essential areas of the value and pricing of data, platform pricing, pricing of subscriptions, and the monetization of the global environment. Case studies, examples, and interviews from leading organizations, including Zuora, Honeywell, Relayr, Alcatel Lucent, ABB, Thales, and General Electric, illustrate key concepts in practice. To aid student learning, chapter objectives, summaries, and key questions feature in every chapter, alongside PowerPoint slides and a test bank available online for lecturers. Comprehensive and applied in its approach, this text provides postgraduate, MBA, and Executive Education students with an understanding of the capabilities, processes, and tools that enable executives to effectively implement digital transformations and capture value from digital innovations. Stephan M. Liozu is the founder of Value Innoruption Advisors, a consulting boutique specializing in value-based pricing, monetization strategies, and digital pricing. Stephan sits on the Advisory Board of LeveragePoint Innovation and the Professional Pricing Society. He is the author of multiple books about pricing, including Pricing and Human Capital (2015), and co-editor of Innovation in Pricing (2012), The ROI of Pricing (2014), and Pricing and the Sales Force (2015). Andreas Hinterhuber is an associate professor at the Department of Management at Università Ca' Foscari Venezia, Italy, and is an equity partner at Hinterhuber & Partners, a consulting company specializing in pricing based in Innsbruck, Austria. He has published articles in leading journals including the Journal of Business Research and MIT Sloan Management Review, and has edited many books on pricing, including Innovation in Pricing (2012), The ROI of Pricing (2014), Pricing and the Sales Force (2016), and Value First then Price (2017).

Digital Pricing Strategy Capturing Value from Digital Innovations

Edited by Stephan M. Liozu and Andreas Hinterhuber

Cover image: © sutlafk First published 2023 by Routledge 4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2023 selection and editorial matter, Stephan M. Liozu and Andreas Hinterhuber; individual chapters, the contributors The right of Stephan M. Liozu and Andreas Hinterhuber to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-032-12771-2 (hbk) ISBN: 978-1-032-12772-9 (pbk) ISBN: 978-1-003-22619-2 (ebk) DOI: 10.4324/9781003226192 Typeset in Sabon LT Std by Deanta Global Publishing Services, Chennai, India Access the Support Material: www.routledge.com /9781032127712

Dedication We dedicate this book to all the digital innovators in the world. This book was prepared and packaged to help them become more successful with their digital ventures. We hope they price well and based on customer value. We invite them to join the pricing revolution.

Contents

List of Figures

xvii

List of Tables

xxi

List of Contributors

xxiii

Acknowledgments

xxvii

Introduction

1

Motivation for the book

2

Gathering of the best experts in the world

3

Structure of the book

4

Key questions and future areas of research

7

References

7

SECTION 1 DIGITAL PRICING

9

1



The Essential Ingredient for More Effective Digital Pricing: Value

11

Changing the script in digital pricing strategy

11

How to measure value for use in digital pricing

12

Leveraging technology to measure value

12

Evaluating value-based digital pricing over time

13

The dangers of a good/better/best model

14

The powerful pairing of value pricing and value selling

14

CVM and the direct impact on SaaS pricing

15

Building value synergies in digital pricing and selling

15

Bio

16

ix

Co n te n ts 2

3

4

5

x

Publish Your Prices

19

Setting the stage

19

Spirit Airlines versus Southwest Airlines

20

The SaaS imperative

20

Practicing value-based pricing

21

Lowering barriers to buy

21

Enabling customer self-selection

22

Demonstrating customer-centricity

23

Conclusion

24

Bio

25

References

26

Dynamic Pricing Process: How to Transition from Fixed to Dynamic Pricing?

27

overcoming the paradigm of fixed prices

27

Forms of dynamic pricing

28

Dynamic pricing: Price differences over time

28

Personalized pricing: Price differences among consumers

28

Dynamic pricing process

29

Dynamic pricing strategy

30

Dynamic price setting

30

Context factors for price setting

32

Dynamic pricing implementation

34

Dynamic pricing auditing

35

Conclusion

35

Acknowledgments

36

Bio

36

References

37

Realizing Your Monetization Potential Needs Customer Value Management

39

Introducing customer value management

40

The changing customer

41

All subscription businesses will need to behave like software businesses

41

So, what does a successful customer value management strategy look like?

42

Value management starts with discovery

42

The next evolution of CRM

43

Value of CVM to a B2B software company

44

Conclusion

45

Bios

45

Measure and Quantify the Value of Your Digital Solution

49

Competition redefined!

50

What is your basis for true differentiation?

52

Choosing and using dollarization techniques

54

Difficulties in dollarizing data-driven and digital offers

56

Value modeling: Expressing relative value in money terms

57



 Co n te n ts The must-do tasks for dollarization

59

note

60

SECTION 2 SOFTWARE AND SUBSCRIPTION-BASED PRICING 6

7

8



9

58

Bio

61

Price Increase for Discounted Customers in SaaS: Pricing Research Description and Success Story

63

Understanding value creation versus price realization

63

Case study introduction

65

Revenue engine diagnostics

66

Impact modeling and risk reduction

66

Marking leakage implementation and project RoI

67

Conclusions

68

Bio

68

SaaS Pricing: From Subscriptions to Usage-Based Pricing Models

71

Pricing: An important, yet neglected activity

71

Price-setting approaches in the software industry

72

Price getting in the software industry: Extreme levels of discounting

74

Software pricing models

74

Best practices of usage-based pricing models

75

Capabilities critical for implementing usage-based pricing models in SaaS

77

Conclusion

79

References

80

The Digital Pricing Framework: Best Practices in B2B Pricing and Offer Design

83

Introduction

83

The digital pricing framework

83

The road to value-based digital commercial offers

85

Strategy (the 10,000-foot view)

87

Price-value analysis (the 2,000-foot view)

87

Commercial structure (the 500-foot view)

90

Financial analysis (ground-level view)

94

Conclusion

95

Bio

95

References

96

Tapping into the Subscriber Psychology with Good/Better/Best: Is There an Optimal Ratio between Tiers?

97

Executive summary

97

The shift to usership is accelerating

97

The shift from product-centric to subscriber-centric

98

Linear, product-centric model

98

xi

Co n te n ts Dynamic, subscriber-centric model

10

101

The beer trial

103

Finding the right balance: How to build a GBB package

103

Design tiers to address the needs of your customer segments

111

Deploy pricing strategy based on your organizational maturity and market positioning

114

Bios

115

References

118

Value-Based Pricing of Smart-Product-Service Offerings in the Manufacturing Industry

11

98

Why good/better/best?

119

Introduction

119

Framework for pricing smart-product-service offerings

120

Smart-product-service-system design

121

Value determination

123

Price model design

125

Price metric design

127

Four pricing patterns for smart-product-service offerings

129

Conclusions

131

Bio

131

References

132

Price Sensitivity Meter and Conjoint Analysis as Tools for Setting Your Industrial Subscription Pricing

135

Part 1: How to run a price sensitivity meter exercise

135

Bio

139

Part 2: A practical guide to conjoint analysis: Injecting more confidence into your strategic pricing decisions Bio

xii

139 147

SECTION 3 THE VALUE AND PRICING OF DATA

149

12

151

Overcoming Real-World Challenges in B2B Digital Pricing Transformation Introduction

151

The team

152

Project or product?

152

Data

153

Technology and technical skills

153

Algorithms and models

154

Testing and auditability

155

Integrations

155

Continuous experimentation (A/B testing)

155

Conclusions

155

Bio

156



 Co n te n ts 13

14

15

16

159

Introduction

159

Industrial market segmentation

161

Alternative framework: Defining the key triggers from a buyer’s perspective

162

The process: Results of pilot project

167

Conclusions

169

Bio

170

References

171

Three Considerations for Data Monetization and Value Creation in the Digital Age

173

Introduction

173

First consideration: The four stages of data monetization

173

Second consideration: Creating a data strategy that delivers value

176

Third consideration: Value engineering—The secret sauce for data science success

180

Bio

184

note

185

References

185

The Economics of AI: How to Shift Data Projects from Cost to Revenue Center

187

Introduction

187

The economics of AI

187

The cost of enterprise AI

188

Capitalization and reuse

191

Capitalization and reuse with an AI platform

192

Conclusion

193

Bios

193

References

194

The Pricing of Data: An Interview with Jian Pei, Simon Fraser University

195

Bio

201

SECTION 4 THE PRICING OF PLATFORMS AND MARKETPLACES

203

17

Marketplace Monetization Methods

205

18



Holistic Approach to Market Segmentation of Industrial Smart Services: What Is the True Value of Data?

Monetization methods

205

Current trends in marketplace monetization

210

Bio

212

note

213

The Monetization of Marketplaces and Platforms in the Context of Web 3.0

215

Introduction

215

Web 1.0: The read-only web

215

Web 2.0: The interactive and social web

216

In the latter part of Web 2.0, the sharing economy took hold

216

xiii

Co n te n ts

19

Web 3.0 is about to take hold. here is where things get really interesting

218

The future benefits

220

Conclusion

221

Bio

221

Reference

222

Pricing in Platforms and Marketplaces: A Primer in Understanding All the Dimensions of the Pricing Problem and Opportunity in Marketplace Platforms

223

Platforms offer complex pricing opportunities

223

Pricing in the marketplace

223

Pricing the ‘product side’ of your platform strategy

226

Another pricing context: Your extensions platform

228

Pricing and unit economics

229

How to use pricing strategically in platforms and marketplaces

230

Conclusions

233

Bio

233

References

234

20 Online Pricing Experimentation Introduction

235

Why experiment

236

Steps in a pricing experiment

238

How should pricing experimentation be organized?

245

Summary

246

Acknowledgments

246

Bio

246

References

247

SECTION 5 PRICING AND ARTIFICIAL INTELLIGENCE

249

21

Artificial Intelligence and Its Impact on Pricing Technology

251

Introduction

251

Fourteen ways AI impacts and improves pricing

251

Bio

254

References

255

22 Why AI Transformations Should Start with Pricing

xiv

235

259

Why pricing is an excellent target

259

A CPG company’s AI pricing transformation

260

Three key steps to success

261

Bios

262

Acknowledgments

262

References

263



 Co n te n ts 23 Digitization of B2B Pricing: A Fundamental Shift Required

265

Introduction

265

Changing B2B buyer needs

267

How does B2B pricing need to change to meet the needs of the new B2B buyer?

270

Bio

272

References

273

24 Value-Based Offers Assisted by Artificial Intelligence

275

AI methodology

277

Making the value-based offers relevant and personalized

280

Making the platform agile, aligned, and actionable

282

Bios

283

References

285

25 Digital Transformation: How to Convert a Buzzword into Real Bottom-Line Value

287

Demystifying digital transformation

287

Empowering B2B buyers and sellers with digital commerce

288

Improve the omnichannel experience with deal negotiation

288

Where human intelligence and artificial intelligence converge

290

Achieving and measuring commercial excellence

293

A practical example: World-leading water technology company Xylem enables valuebased pricing for 1 million product configurations

295

not a project, a transformation

298

A practical example: Molex’s commitment to commercial excellence

298

So, who owns pricing?

298

A practical example: How a $150 billion+ integrated health care services company reimagined its strategic outlook toward pricing Conclusion

300

Bios

302

References

303

Index



300

305

xv

Figures





xvii

F igures

xviii



 F igures





xix

Tables





xxi

List of Contributors

Camille Brégé is a managing director and partner in BCG’s Paris office. Arnaud Bassoulet is a principal in BCG’s Paris office. Simone Cicero is Founder of Platform Design Toolkit. He is an entrepreneur, facilitator, thinker, writer, and host, with a focus on open business models, ecosystemic organizations, and platforms. He was featured in the Thinkers50 Radar class of 2020. John V. Colias, PhD, is a leader with both university teaching and business consulting experience, focuses on predictive modeling and prescriptive analytics. He is Senior Vice President, Research and Development, at Decision Analyst. Louis Columbus is a software product marketing and product management leader. He holds an MBA from Pepperdine University and completed the Strategic Marketing Management and Digital Marketing Programs at the Stanford University Graduate School of Business. Darius Fekete is Managing Consultant for Vendavo’s Value Consulting team, with more than a decade of professional experience in pricing and digital transformations. Alexis Fournier is Regional VP, EMEA, AI Strategy at Dataiku. He began his career as a

data scientist in the telecommunications industry before joining an international organization. Alexis supports organizations in the understanding of the value of AI in the enterprise and its processes, as well as on the different paths to everyday AI.



xxiii

L ist o f Co n tribut o rs Claire Gubian leads the Business Transformation practice at Dataiku, which helps customers accelerate their transformation thanks to AI. Claire is a seasoned leader who spent most of her career in management consulting, Mrinal (MG) Gurbaxani is the CEO and Co-Founder of Cuvama. He holds a degree in mathematics from the College of Wooster (USA) and an MBA from INSEAD (France/ Singapore). Joël Hazan is a managing director and partner in the Paris office of Boston Consulting

Group (BCG).

Andreas Hinterhuber is Associate Professor at the Department of Management at

Università Ca’ Foscari Venezia, Italy. Matt Johnston is the Founder and CEO of EPIC Conjoint. Mitchell D. Lee is VP Product Marketing and Profit Evangelist at Vendavo with more than three decades of experience in improving technical, operational, marketing, and commercial processes. Tobias Leiting, MSc, Calvin Rix, MSc, Regina Schrank, MLitt, and Dr.-Ing. Lennard Holst

are project managers in the service management department at the Institute for Industrial Management (FIR) at RWTH Aachen University. The research and consultancy activities focus on the transformation of manufacturing companies from producers to suppliers of smart-product-service-systems by enabling companies and business units to design, market, and efficiently deliver offerings for their external and internal customers. Stephan M. Liozu is the Founder of Value Innoruption Advisors, a consulting boutique

specialized in value-based pricing, industrial pricing, and digital and subscription-based pricing. Stephan holds a PhD in management from Case Western Reserve University (2013). Jacek Łubiński is a software engineer turned finance professional turned VC. He has nine years of experience in VC. He holds three master’s degrees from Austrian and Polish universities (two in computer science and one in finance) and enjoys reading books and drinking yerba mate. Scott Miller, CPA, CMA, is Founder and President of Miller Advisors Inc (www.miller

-advisors.com). ​ Michael Mansard is a seasoned subscription economy business strategist. He currently serves

as Principal Director, Business Transformation and Subscription Strategy within Zuora’s Chief Revenue Officer’s group. Michael is also the Subscribed Institute’s EMEA Chair. Over 27 years at AT&T, Stella S. Park has held a variety of leadership roles in sales, customer research and analytics, global marketing, strategy and execution, and currently in global business channel marketing.

x xiv



L ist o f Co n tribut o rs Jian Pei is a professor in the School of Computing Science at Simon Fraser University

and an associate member of the Department of Statistics and Actuarial Science. He is a well-known leading researcher in the general areas of data science, big data, data mining, and database systems. Robert Phillips is former Director of Pricing Research at Amazon and Director of Marketplace Data Science at Uber. John Porter is CTO and co-founder of DecisionLink, John and his team have engineered more value solutions than any other company while helping to solve some of the most complex business modeling challenges for some of the largest companies in the world. Luis Prato has more than 15 years of experience in turning technology into commer-

cially viable services for critical infrastructure. He is an affiliated PhD candidate at the Rotterdam School of Management. Murali Saravu is the Founder and CTO of Monetize360, a technology company that

delivers dynamic monetization solutions to sales and finance departments. Bill Schmarzo is Chief Data Monetization Officer | Recognized innovator, educator, prac-

titioner in Data Science, Design Thinking | Creator Big Data MBA | Author of four books including ‘Economics of Data, Analytics and Digital Transformation.’ Alex Smith is CCO and Co-Founder of Cuvama. Alex has worked for 17 years in presales and value-selling across a wide range of B2B and B2C industries, including traditional wholesale distribution, financial services, and industrial manufacturing. Gaurav Sonpar is a Business Strategist at Zuora, a leading subscription economy evangelist. Gaurav holds a Bachelor of Engineering degree and an MBA from Lancaster University. Jean-Sébastien Verwaerde is a managing director and partner in BCG’s Paris office. Arnd Vomberg is Professor of Digital Marketing and Marketing Transformation in the Marketing and Sales Department of the University of Mannheim. Before joining the University of Mannheim, he was an Associate Professor with Tenure at the University of Groningen (The Netherlands). Lalit Wadhwa is Executive Vice President and Chief Technology Officer at Encora’s client services. He leads Encora’s Technology Practices Areas, in support of the company’s sustained growth. Kyle T. Westra is a pricing strategist who focuses on elevating technology companies.



He holds an MBA with distinction in business strategy and marketing analysis from the Kellstadt Graduate School of Business at DePaul University and a BA in political science and economics from Tufts University.

xxv

L ist o f Co n tribut o rs Maciej Wilczyński is a pricing expert and partner at Valueships, a consultancy boutique specializing in software, cloud, subscription, and digital businesses. He holds a PhD in strategic management. Craig Zawada is the Chief Visionary Officer at PROS. A widely published author,

Zawada is perhaps best known for co-authoring The Price Advantage, which has been recognized as one of the most pragmatic books available on pricing strategy.

xxvi



Acknowledgments

We want to thank the entire Routledge publishing team for helping make this project a reality. We also thank all authors for sharing their knowledge and expertise graciously and publicly. We want to recognize their efforts and patience with the writing and publishing process. We encourage all readers to connect with them and engage in positive discussions about how value and pricing can revolutionize digital transformations.



xxvii

Introduction Stephan Liozu and Andreas Hinterhuber

The world of technology is evolving at the speed of light. It is amazing to read about the progress made in just the last 10 years. One can only imagine what the next decade has in store. From quantum technology to advancement in smart and rational artificial intelligence (AI), to digital platforms, and connected ecosystems, the sky is the limit. This is overly exciting for the tech world of science and technology. We wish we could say the same about the worlds of value and pricing management. Over the past 20 years, we have witnessed the emergence of interesting and improved pricing technologies. More pricing software companies have popped up and have invested in AI pricing technology. There are more discussions in consulting circles about the power of data-driven and AI-infused pricing. Let us consider the following facts. On the one hand, massive investments are made in digital transformations. ■ ■

■

More than 80% of companies are conducting some form of digital transformation project (Zuora, 2021). Every year vast investments are allocated and spent to support digital transformations. In 2019 alone, according to IDC (BusinessWire, 2019), $1.18 trillion in investments were made in technologies and services to enable digital transformations. IDC also predicted that $6 trillion would be invested between 2018 and 2022. That is a massive amount of cash. IDC reports that in 2023, the number of investments in digital and technology will be over $2 trillion. They estimate that of these, around $754 billion will be allocated to monetization technologies. A larger share of these investments will be done in payments, payment processing, and subscription management software (IDC, 2019).

DOI: 10.4324/9781003226192-1

1

S tepha n L i o z u a n d A n dreas H i n terhuber

On the other hand, the impact is not fully realized, and investments in pricing are small. ■ ■

■ ■

■ ■

The pricing software space is small and growing at a modest pace (Liozu, 2019b). Investments in pricing science and human capital are not at the level we expect given the rate of penetration of digital transformation projects and data analytics initiatives. Only 22% of global Fortune 500 companies have dedicated pricing teams with 20 staff or more (Liozu, 2019a). It is reported that most startups spend less than 6 hours discussing and setting prices. It is a running joke that startup leaders spend more time discussing janitorial supplies than the pricing of their offers (ProfitWell, 2019, SaaS Pricing). Only 6% of companies have managed to create monetary impact from their digital investments (Accenture, 2019). Only 30% of transformations met or exceeded their target value and resulted in sustainable change; companies in this group are in the win zone. Some 44% created some value but did not meet their targets which resulted in only limited long-term change; these companies are in the worry zone. And in the woe zone, 26% created limited value (less than 50% of the target), producing no sustainable change (Forth et al., 2020).

We have been working for two decades in the field of value and pricing management and have seen firsthand the lack of sophisticated capabilities in critical areas of digital transformations and digital programs. While we can demonstrate a causal effect, we can certainly refer to all the consulting reports discussing the failure rate in IoT projects, the lack of impact of digital transformations, and the rate of startup failures due to a lack of product–market fit or lack of customer segmentation. There are countless examples of companies managing digital strategies and value/pricing strategies very well. There are many more showing the opposite. The current level of disruption and volatility since 2019 generated hope that we could expect a dramatic boost in pricing software with the emergence of direct-to-consumer models, the boom in e-commerce, or the move to recurring business models. The hard reality is that value and pricing management remain second thoughts. They are merely included in firms’ marketing strategies or buried in the product and packaging discussion. Digital factories might have a few value or pricing coaches, but these people are no pricing specialists. Motivation for the book

The current level of penetration of value and pricing science in the digital transformation process is still limited. It is a strong motivation for us to publish this book in partnership with Routledge. For the past 10 years, we have spent countless hours and resources filling the gaps in the pricing profession and bringing greater attention to the fields of value and pricing management. Today, we feel compelled to do this again to connect the world of digital strategies and pricing strategies. Frankly, we feel it is still a white space and new concepts for many digital professionals. We want to make them aware of critical value and pricing concepts that can increase their knowledge base, their capabilities, and their chances of success in future digital programs. This might sound like an ambitious agenda, but we feel strongly about improving the visibility of pricing science and strategies in the world of digital. 2



I n tr o ducti o n

At the same time, we wanted to challenge the pricing community to be as creative as technology and digital visionaries. We wanted to focus on inventing the future of value and pricing management and use this book as a vehicle to communicate new concepts, new methods, and new theories. They might not be new to the best pricing experts in the world, but they are extremely ‘out there’ when you connect them to digital strategies.

Gathering of the best experts in the world

With great ambitions come the need to put together an amazing roster of experts. It was priority one for us. We contacted the best experts in the world and asked them to participate in this challenging project. As you can imagine, finding experts with visionary and unique digital and pricing expertise was not an easy task. So, we made some bold moves and decided to cold contact many experts. Today, we are pleased to say that we have been able to bring many amazing thinkers from around the world to make this project a reality. We are grateful for their participation and their sharing of important content:



Arnaud Bassoulet Camille Brégé Simone Cicero John Colias Louis Columbus Darius Fekete Alex Fournier Claire Gubian MG Gurbaxani Joël Hazan Lennard Holst Matt Johnston Mitch Lee Tobias Leiting Jacek Łubiński Michael Mansard Scott Miller Stella Parks Jian Pei Robert Phillips John Porter Luis Prato Calvin Rix Murali Saravu Bill Schmarzo Regina Schrank Alex Smith Gaurav Sonpar 3

S tepha n L i o z u a n d A n dreas H i n terhuber

Jean-Sébastien Verwaerde Arnd Vomberg Lalit Wadhwa Kyle Westra Maciej Wilczyński Craig Zawada

Structure of the book

This book includes 25 chapters organized into five sections. We organized each section thematically and sought to have the right balance between sections. Our goal was to focus on the critical areas of digital transformations: pricing of data, subscriptionbased pricing, pricing in platforms and marketplaces, AI pricing, and digital pricing.

Section 1: Digital Pricing

Section 1 focuses on digital pricing and managing value in the context of digital transformation. Chapter 1, written by John Porter, CTO at DecisionLink, describes the need to embed value pricing in digital strategies in an automated way. The premise of this chapter is that technology is available to automate the management of pricing and value. Relying on manual processes does not work when digital strategies are designed to scale fast. The second chapter, written by Kyle Westra, discusses the need for pricing transparency in digital pricing and more specifically in SaaS. Current trends have shown that many companies are gating their pricing information and not publishing pricing for prospects and customers. Is that a good thing? Should customer pricing be self-service? These are questions that are treated in Chapter 2. In Chapter 3, we move to the retail world and discuss the need for value-based dynamic and personalized pricing. Arnd Vomberg introduces and discusses a systematic dynamic pricing process that provides practical guidance for decision-makers. This discussion enables managers to decide whether and how to implement a dynamic pricing approach in retail firms. Continuing with the theme of value-based pricing, Chapter 4 is written by the co-founders of Cuvama, a dedicated customer value management (CVM) platform enabling value strategies in companies. The two authors compare the CRM technology with the CVM technology to demonstrate that we are entering a new age where value management needs to have dedicated technology to boost value and pricing impact. Finally, Chapter 5 is dedicated to the topic of extracting true differentiation from digital innovations to manage customer value and pricing digital solutions based on value and not on cost or competition. This first section has a strong theme of both value and pricing and already makes a convincing case that one cannot be managed with the other. Technology also plays a critical role to enable the relationship.

Section 2: Software and Subscription-Based Pricing

Section 2 is dedicated to the pricing of software and using subscription-based pricing models to reach new markets and accelerate growth rates. The topic is not new. For the past 10 years, we have all experienced the subscription tsunami, especially in the

4



I n tr o ducti o n

business-to-consumer world. We invited book contributors to develop updated content for this section. Maciej Wilczyński starts us off in Chapter 6, with a case study on how pricing can be a powerful lever for software companies to boost profit. By analyzing data and focusing on the right metrics, the case study demonstrates the power of pricing research and analytics. In Chapter 7, Dr. Hinterhuber proposes a critical discussion of pricing approaches in the SaaS industry. This chapter reviews pertinent empirical studies, analyzes price-setting and price-getting practices in the industry, and outlines the capabilities required to master one important challenge: the move from subscription-based pricing to consumption-based pricing models. Several case studies provide best-practice examples. In Chapter 8, Scott Miller plunges us into his SaaS pricing framework. It is a unique roadmap for software product and pricing developers to conduct a successful transition from perpetual licenses to subscription-based pricing. In Chapter 9, Gaurav Sonpar and Michael Mansard provide innovative insights on best practices to set good/better/best packages when designing subscriptions. They claim that it is both an art and a science to do so successfully. Leveraging numerous client engagements and industry-leading practices, they have developed a 3D pricing and packaging (P&P) framework to help companies launch and refine a GBB-based packaging model. Along the way, they have embedded real-life examples, specific guidance, and benchmarks. Chapter 10 brings us to the world of Industry 4.0. Four scholars from the University of Aachen in Germany present their latest research on the pricing of smart products. They also propose a framework to extract customer value and price these products. Finally, the last chapter of section 2 highlights the need to conduct deep customer research to inform digital pricing decisions. For this chapter, Matt Johnston and Maciej Wilczyński have joined forces to present two pricing research techniques that can yield impressive insights to get digital pricing right. This second section is rich in methods and frameworks all focused on value-based pricing in the context of SaaS and subscription.

Section 3: The Value and Pricing of Data



This is the true novelty of this book. We wanted to offer a dedicated section on the value and pricing of data, and we achieved that. In Chapter 12, Lalit Wadhwa sets the stage. He proposes the challenges that leaders and technology teams must overcome to conduct successful digital transformation. Many of these challenges are data-related, as you can imagine. The value of data is the heart of Chapter 13, by Luis Prato. Luis discusses the results of his latest research, which consists of a holistic market-segmentation approach of industrial markets to derive the true value of data for end customers. His research is based on 100 interviews with business leaders and concludes that industrial natives need alternative market-segmentation frameworks that consider the value provided by their digital capabilities and the performance orientation of such services to better position these service offerings. Chapter 14, written by the Dean of Big Data, Bill Schmarzo, highlights three considerations for successfully monetizing data and extracting value in the digital age. Bill dives into the stages of data monetization strategies and data strategies that deliver value. In Chapter 15, Claire Gubian and Alex Fournier connect the topic of data and the value of artificial intelligence. They claim that without proper data management strategies, the economics of AI cannot be enabled and delivered to users. They offer interesting insights on the economics of AI

5

S tepha n L i o z u a n d A n dreas H i n terhuber

and the need to capitalize and reuse through data management. We conclude this section with an interview with Professor Jian Pei, an expert in the pricing of data sets. We discuss what makes data valuable and future trends in the data monetization sector.

Section 4: The Pricing of Platforms and Marketplaces

This is also a very innovative section. Searching for insights into the topic yielded very fragmented knowledge. We contacted the best people in the business, four of whom agreed to participate in the project. Chapter 17, by Jacek Łubiński, summarizes the techniques used to monetize marketplaces from a venture capitalist perspective. These techniques are proposed in a simple monetization matrix crossing popularity and attractiveness. Chapter 18, by Murali Saravu, CTO and Founder of Monetize360, discusses the monetization of platforms and marketplaces in the context of Web 3.0. Murali explains the differences in monetization between Web 1.0, Web 2.0, and Web 3.0. In Chapter 19, Simone Cicero proposes another view on the monetization of platforms and marketplaces from a business model design perspective. This chapter helps with understanding all the dimensions of the pricing problem and opportunity in digital marketplace-platforms. The last chapter of this section, by Robert Phillips of Nomis Solutions, discusses the benefits of running online experiments in digital marketplaces and how these can improve pricing power. Robert proposes a framework for online pricing experimentations.

Section 5: Pricing and Artificial Intelligence

Finally, we offer a full section on the emergence of artificial intelligence in pricing. This topic has increased in popularity, and we are witnessing the emergence of new pricing startups promoting new AI-powered pricing capabilities. In Chapter 21, Louis Columbus provides a review of all trends related to AI and pricing. This is an extremely useful list of statistics, quick facts, and nuggets of practical impact. Chapter 22 is a short essay by a group of pricing experts from the BCG group. They posit that any digital transformation with a strong AI component should begin with pricing. They argue that pricing science is already very much linked to AI and can quickly achieve impact and credibility in companies starting their AI journeys. Chapter 23, by Craig Zawada, Chief Visionary at Pros, highlights that AI by itself does not work. For AI in pricing to be successful, a company and its leaders need to think about a fundamental shift in how pricing and commercial activities are managed. AI can only work when process and culture are transformed simultaneously. In Chapter 24, Stella Parks and John Colias write on the topic of value-based offers assisted by AI. This chapter focuses on how AI can assist B2B sellers in identifying and interpreting unmet customer needs and creating value-based offers that will delight customers and exceed their expectations. Finally, Chapter 25 also deals with the all-important topic of making digital transformation real. Mitch Lee, Vendavo’s Profit Prophet, and Darius Fekete review critical aspects of digital transformation and how AI helps in achieving superior results. To unlock the true potential of digital transformation, organizations must take a comprehensive approach to commercial business processes, aligning experts and best practices with purpose-built, enterprise-ready technology capabilities to achieve commercial excellence.

6



I n tr o ducti o n Key questions and future areas of research

As you can imagine, we have only scratched the surface of the value and pricing topics in digital strategies. But there is just so much you can include in a single book. We decided to be very selective about the chapters we included, focusing on the most innovative topics to bring the discussion forward. Other topics deserve further attention and future research. First, consultants are beginning to write about pricing in the metaverse. This is the next big wave of investments. The metaverse connects several of the topics we touch on here: Web 3.0, marketplaces, and AI-based pricing. Second, another topic of future interest is the impact of quantum computing on pricing functions. How do supercomputers influence pricing decisions in the field, especially during times of volatility? Can you leverage the power of quantum computing with the current sophistication of commercial processes? Third, crossing pricing and digital transformation immediately raises legal and ethical red flags. Today, we already see the potential disruption of dynamic pricing for customer perceptions of fairness. Other authors also claim that AI in pricing can lead to pricing fixing and the emergence of dominant positions. We could write a whole book on this topic. Finally, we want to reinforce the connection between the scientific and the human side of pricing and technology. Can supercomputer and quantum computing be managed without quantum leadership in pricing and value management? The world needs further development in neuroscience and the development of superior levels of intuition. Will AI and computers make all the pricing decisions? Or will humans be able to decipher insights and use superior intuition to interpret the outcome of supercomputers? Philosophical discussions need to happen on all these topics. For now, we hope you enjoy reading this book and all the expert contributions we’ve assembled on this fascinating topic. Again, we thank Routledge for giving us the opportunity to work on this project, and we thank all contributors once more. Happy reading!



References Accenture. (2019). Your business as-a-service: Putting the right pieces in place. https://www​ .accenture​.com/​_ acnmedia ​/ PDF​-119​/Accenture​-As​-a​- Service​-Business​-Acceleration​.pdf. BusinessWire. (2019, April 24). Businesses will spend nearly $1.2 trillion on digital transformation this year as they seek an edge in the digital economy, according to a new IDC spending guide. https://www​.businesswire​.com​/news​/ home​/20190424005113​/en​/ Businesses​ -Spend​-1​. 2​-Trillion. Forth, P., Reichert, T., de Laubier, R., & Chakraborty, S. (2020). Flipping the odds of digital transformation success. Boston Consulting Group, October 29. https://www​.bcg​.com ​/ publications​/2020​/increasing​-odds​-of​-success​-in​-digital​-transformation. IDC. (2019). Market analysis perspective: Worldwide digital business models and monetization. Liozu, S. M. (2019a). Penetration of the pricing function among global Fortune 500 firms. Journal of Revenue & Pricing Management, 18(6), 421–428. Liozu, S. M. (2019b). State of the pricing management software industry. The Journal of Professional Pricing. https://www​.stephanliozu​.com​/wp​- content​/uploads​/2021​/02​/2019​ -Liozu​- State​-of​-the​-Pricing​-Management​- Software​-Industry​-JPP​- Q2​.pdf. Zuora. (2021, March). Subscription economy index report. https://www​.zuora​.com​/resource​/ subscription​- economy​-index/.

7

SECTION 1

Digital Pricing



C hapt e r 1

The Essential Ingredient for More Effective Digital Pricing Value John Porter

Setting the ‘right price’ has long been a conundrum for marketers. Many give lip service to value pricing, but most people building digital and SaaS solutions don’t fully understand the qualitative and quantitative value proposition for their customer. As a result, the specificity of cost-based pricing makes it easier and more appealing for many companies. It’s a formula: cost of goods + margin with a market/value gut check added for good measure. The ongoing challenge of this approach is that there’s always someone willing to do something for less. Ultimately, companies will undercut their price, and benchmarking against tangential solutions in the market can lead to a race to the bottom. Cost-based pricing is almost never informed by your product’s true market potential. Without a robust market value assessment, most companies are leaving money on the table when it comes to pricing their SaaS and other digital products.

Changing the script in digital pricing strategy

Value pricing offers a more market-driven approach to pricing for companies that can effectively quantify the value of a new product or service. By implementing value-based digital pricing, companies can effectively maximize the price that a product or service can receive. The good news is that technology and the growing degree of digitization in business today make quantification possible in a way that was previously unheard of. At the earliest stages of product ideation and development, companies should be laser-focused on identifying projected market impact and potential. This is essential from a digital pricing standpoint, but it also can have a long-term impact on sales success. In today’s tech-focused world, it’s extremely easy to fall into the feature and function battle. In addition to being an unsophisticated sales approach, this can also have a negative impact on SaaS pricing, as downward pressure to one-up the competition becomes the norm. A value-based sales strategy is key to sustaining a stronger digital price point and ensuring that discounting doesn’t take over during the renewal process. DOI: 10.4324/9781003226192-3

11

J o h n P o rter

While we may be getting ahead of ourselves, given that sales strategy comes many steps after pricing decisions, it’s important to understand the long-tail impacts of not focusing on value from the outset.

How to measure value for use in digital pricing

When it comes to value pricing, the challenge most companies face is that they simply don’t know where to begin. The key to understanding value is to focus on what’s most important to your target customer’s business. What are their priorities? What goals are they trying to achieve? What challenges frustrate them or hold them back? Evaluating how your product or service delivers on those needs makes SaaS pricing less theoretical and easier to justify. Your customers’ priorities, problems, and areas of focus are value drivers. You have to consider soft value (why they want your product) and hard value (how they can justify it to other internal stakeholders). These value drivers likely look different for different customers, but some key considerations as a starting point for value measurement are these: ■ ■ ■ ■

How much cost savings does a product or service enable? How does a product or service result in better use of resources? How does it impact productivity? What’s the bottom-line impact of reducing incidents?

Typical products or solutions may have anywhere from three to 12 value drivers. By first establishing the value drivers for your target market, you can create a business hypothesis that can then be tested. This is actually common practice in industrial automation companies. In manufacturing-driven organizations, value conversations are typically owned by product teams, which means that value is a core consideration during the product development process. Interestingly, this is not typically the way tech companies approach it, which makes digital pricing challenging for many. By merging industrial and technology development approaches, companies can get closer to a value-driven process: identify the problem. Enumerate the value drivers of solving that problem. Test across a variety of situations to quantify the return and impact based on those value drivers. And then assess the market potential for the value delivered. With this kind of well-thought-out approach, organizations have a clear way to prioritize development projects. Instead of working based on a gut feeling, the product development organization has tangible drivers to gather feedback around. At the same time, this becomes the foundation for establishing a value-based SaaS pricing strategy and a more powerful position in the market.

Leveraging technology to measure value

Before digital transformation, it was almost impossible or, at the very least, extraordinarily time-consuming to get to true measures of value. Today’s connected enterprise changes that. Technology solutions now give companies the ability to quantify the value their products and services provide. This is called customer value management

12

C hapter

1

T he E sse n tial I n gredie n t: C ust o mer Value

(CVM), and it is powered by a technology platform that connects a company’s existing systems to provide value insights across all customer touchpoints from marketing and sales to customer success, and in the case of value-based pricing, product development. With CVM, companies have the ability to centralize the collection of value data and apply core analytics to measure exactly what a product or service is doing for customers in dollars saved, productivity increased, top line enabled, and so forth. Whatever that customer’s value drivers are should be what’s measured. By establishing benchmark measures and quantifying value more regularly, a collaborative, more trusted advisory relationship becomes readily achievable between sellers and buyers. This deeper level of insight also facilitates the more productive value-based approach to digital pricing. A CVM platform gives companies an on-demand source for powerful pricing intelligence. This is bigger than point-in-time ROI measurements. CVM is an always-on tool for staying on top of value with access to real-time reporting and the ability to zoom in and out on value drivers across industries and segments and by using other variables. These tools make it possible to measure value from acquisition and retention to relationship expansion, providing product managers and marketers with a reliable, real-world source for benchmark data and intelligence to power more data-driven digital pricing strategies. To test value-pricing permutations, companies can use a CVM system to collect and store data in a benchmark value repository. Then, using the application engine, they can explore these permutations. Where data doesn’t exist, hypothetical values can be used and then further tested if evaluations look promising. At the same time, companies can leverage reference data from existing products to inform value pricing explorations and ensure statistically valid assessments. Being able to objectively assess your organization’s ability to articulate your own value is key to effectively leveraging value-based pricing in the digital arena. By determining how readily your team can measure, articulate, and defend the value of your product or service, you can determine how prepared you are to implement value-based pricing. Without the right foundation in place, sustainability of this strategy is likely unclear.

Evaluating value-based digital pricing over time

The importance of a strong value management foundation is key when you consider one of the major complications of a value-based pricing strategy: value itself changes over time. In fact, there is always a natural tension around what companies and individuals think the value of a product or service is. This is why continuous measurement and communication of value is essential. As value drivers evolve over time, companies with CVM tools can adapt their business value assessments for clients to reflect shifts and additions. With regular assessments, SaaS pricing can remain value-based, but with the recognition that those drivers may change over time. For many technology companies, one of the challenges of this can be that a valuebased pricing approach often results in products being categorized as a variable price by their customer (per user, interaction, incident, etc.). While the initial value may win the customer over, in time, they may struggle with not having a fixed line item that aligns with their desire for budget predictability. Over time, an effective value-realization

C hapter

1 

13

J o h n P o rter

practice (e.g., customer success teams using CVM to communicate value delivered to clients regularly) becomes essential for sustaining a value-based digital pricing strategy. To avoid such issues, in the past, sellers preferred to lock in three-year deals to provide predictability on both sides of the table. This has become a challenging proposition for a number of reasons. First, continuous innovation makes the time frame uncomfortably long for some and can lead to frustration on the part of the customer. Likewise, for companies that are good at communicating value delivered to their customers, it limits cross- and upsell potential. When things are going well, long-term contracts just tend to leave money on the table because they make discussions about add-ons and expansions more difficult to initiate.

The dangers of a good/better/best model

Organizations that have limited ability to articulate and measure value often fall into a good/better/best digital pricing model by happenstance. Many software companies use a freemium design to lure customers into a trial. The challenge is that in most cases, there must be some kind of compelling event for companies to move up to the next tier. At the same time, ongoing innovations begin to move functionality down the product path. While the goal is that innovations will drive customers to higher levels, the reality is that feature shifts can keep them at lower levels because they are getting the functionality they need without understanding the full value they could be enjoying.

The powerful pairing of value pricing and value selling

Companies that implement and leverage value-based digital pricing and value selling have a powerful market advantage. At the front end (sales cycle), focusing on value with prospects at every interaction directly impacts sales results. From shorter sales cycles and higher average sale prices (value pricing comes into play here, as well) to improved close rates and optimized positioning, value has the power to transform customer interactions. Value conversations are higher-level, which also means that pricing is not the salesperson’s only leverage. With value at the center of your sales and marketing strategies, you can price SaaS products and solutions based on their real-world impact in the market. There is no more confident a positioning. Value pricing and selling help you avoid pricing wars from the outset and minimize the risk of adversarial customer relationships. By using value as the foundation, there is natural governance in your interactions with customers, and an advisory relationship is more likely. A clear value-based pricing strategy governed by rules, guidelines, and transparency into the value drivers makes it significantly easier to get your sales team fully onboard and aligned in their efforts. This is key because the greatest goal of value-based selling is to justify and extract a price premium. This cannot be done without having put together customer value models for your products and the segment that quantifies how much value can be realized to justify the price. Likewise, if you implement value-based pricing without executing on it in the sales process, there’s no connection for the customer and the sales team is likely to revert to discounting and making concessions to the customer in order to win the sale.

14 

C hapter

1

T he E sse n tial I n gredie n t: C ust o mer Value

When it comes time for renewals, as discussed earlier, value is the key to expanded cross-sell and upsell conversations. When pricing and value are revisited as a part of this process, clear articulation of value helps defend your position, ensure customer retention, and avoid conflict. One of the challenges of renewal conversations is the transition of the conversation away from the actual users of your products to the procurement team. Procurement’s job is to negotiate the lowest possible cost for products and services rendered. Period. This makes them hungry for rationales for their negotiating stance. Sometimes this results in the use of overly simplistic metrics like raw user count. Even though the true economic value realized from a solution may be high, adoption may be gradual. With tangible data on the real value realized readily available, it becomes difficult to lean into less relevant metrics. At the same time, when customers have a clear picture of the value they’ve realized from a product or service, the demand for discounting is reduced. This directly lifts average net recurring revenue, a key performance measure. Digital transformation has made value data more accessible than ever, which makes all the difference in customer relationships and long-term profitability. With the ability to actively measure value, companies can avoid the debate over qualitative value assessments. With value pricing and value realization in place, organizations no longer have to look at renewals as an annual headache. Instead, these conversations become opportunities to deepen relationships, prove yourself an advisor, and build customers for life.

CVM and the direct impact on SaaS pricing

Verint, a customer engagement automation platform, began using a CVM platform in 2017. Since 2020 the company has been using CVM to quantify and communicate economic value throughout the sales cycle. This has driven powerful results for the company, including a 17% reduction in discounting and doubling the close rate of opportunities where business-value cases are used. Verint’s experience points to key opportunities for leveraging value throughout enterprise organizations: ■ ■ ■ ■ ■ ■

Supporting marketing in attracting and engaging customers quickly. Giving sales the ability to position high and differentiate with compelling business cases. Enabling the team to use business value cases with more prospects. Accelerating growth by closing bigger deals faster. Empowering customer success teams to track and prove value achieved for use in renewal discussions and to maintain price integrity. Improving renewal rates and growing share of wallet.

Building value synergies in digital pricing and selling

Value-based pricing and value selling build on the same inputs—customer value models that quantify value realization. In fact, value-based selling is the most effective way to operationalize value-based pricing and actually bring it to life. When pricing and sales teams partner in these efforts, both teams become stronger in their quantification and

C hapter

1 

15

J o h n P o rter

presentation of value. From measuring value through TCO, ROI, business value assessments, or economic value estimation modeling to tracking value delivered throughout the customer relationships, companies committed to value have the insights and inputs they need to inform renewals, future SaaS pricing strategies, and new sales efforts.

Bio

John Porter is a seasoned expert in the fields of value pricing and selling. As an inventive technologist, he has developed incredible knowledge and enthusiasm about customer value management. He is the first, and only, person to build an integrated platform for value management that aligns expectations, agreement, and realization of customer value across the entire customer journey. As CTO and co-founder of DecisionLink, John and his team have engineered more value solutions than any other company while helping to solve some of the most complex business modeling challenges for some of the largest companies in the world, including ServiceNow, DocuSign, Caterpillar, Adobe, and CrowdStrike. He has served in numerous management and leadership roles across sales, product management, and customer success at edocs, Siebel, Oracle, and SAP.

Key objectives 1. Understand the importance of value-based pricing for digital products, services, and solutions and the benefits it offers SaaS companies. 2. Outline the key steps for developing a value-based price for digital products and solutions. 3. Explore the technology resources currently available to support today’s digital pricing strategies. 4. Establish the importance of an integrated value-based pricing and value-based selling strategy for digital solutions.

Key summary points 1. Value-based pricing and selling maximize the price a company can extract for a digital product or service. 2. How companies can leverage a customer value management platform during the product development process to facilitate value pricing for SaaS solutions. 3. The importance of a combined value pricing and value selling effort in optimizing returns for digital and SaaS products. 4. The role of value throughout the customer life cycle to minimize discounting and build stronger customer relationships.

16 

C hapter

1

T he E sse n tial I n gredie n t: C ust o mer Value

Key questions 1. 2. 3. 4.

What is the key difference between cost-based and value-based digital pricing strategies? What are some typical value drivers for digital products? What are the pitfalls of a good/better/best product and pricing model? What role should value play in pricing considerations and conversations throughout the customer journey?

C hapter

1 

17

C hapt e r 2

Publish Your Prices Kyle T. Westra

The issue of price transparency is a perennial concern for business executives. And the conventional wisdom seems clear: there is a fundamental tradeoff between business success and transparency. Although greater transparency is disadvantageous for companies, the thinking goes, they have an ethical imperative to provide it to customers. I disagree with both aspects of this conventional wisdom. Price and pricing transparency are largely issues of competitive differentiation, not ethics, and businesses that promote transparency can in fact thrive. Perhaps nowhere is this truer than in the world of B2B SaaS. I argue that the dynamics behind successful SaaS companies are best supported by price and pricing transparency. While there is no simple right or wrong in matters of transparency, SaaS offerings provide the unparalleled ability to use transparency to differentiate from the competition, connect better with customers, and improve monetization.

Setting the stage

First, we should clarify what exactly is meant by price and pricing transparency: ■ ■

Price transparency: when a customer knows what they will pay at the onset of the transaction. Pricing transparency: when a customer knows how a company will arrive at the final price.

Companies can excel in one or the other, or both, or neither, a dynamic that I explore thoroughly in my book The New Invisible Hand: Five Revolutions in the Digital Economy (Westra, 2019). For our purposes here, let us see how these concepts apply to two nondigital companies.

DOI: 10.4324/9781003226192-4

19

K yle T. W estra Spirit Airlines versus Southwest Airlines

One of my favorite illustrations of price and pricing transparency comes not from the digital business world but from the low-cost airline industry. Two American carriers, Spirit and Southwest, demonstrate two radically different approaches to transparency. Spirit Airlines is emblematic of the common low-cost airline approach to price and price transparency. It advertises low upfront fares while charging substantial additional fees for what it calls ‘Optional Services’—items such as checked baggage, seat assignment, flight changes, and so forth. Seat assignment may cost anything from $1 to $250, depending on whether the customer purchases in advance, requests a larger seat, or has paid $69.95 for a year of ‘Spirit Saver$ Club Membership,’ which gives access to special discounts. Something as simple as a printed boarding pass may cost either $2 or $10, depending on whether the printing is done at a kiosk or by a gate agent (Spirit Airlines, n.d.). This is complicated pricing. The information is discoverable, but it takes time and effort. Additionally, it is unclear where one might fall in the fee ranges presented (e.g., $1 to $250 for a seat assignment) until making the purchase decision. Spirit Airlines, therefore, has relatively low price and pricing transparency. The price presented upfront is unlikely to be the final price one pays, and the constituent pieces of pricing that lead to the final price are not easy to follow. For Southwest, transparency is key to the story it tells about itself. It has coined the portmanteau ‘transfarency’ to emphasize the transparency with which it displays its fares. According to Southwest, transfarency is the ‘philosophy in which Customers are treated honestly and fairly, and low fares actually stay low—no unexpected bag fees, change fees, or hidden fees’ (Southwest Airlines, n.d.). Southwest highlights its price and pricing transparency as a key part of its branding. It wants to be perceived as the no-nonsense, slightly irreverent, low-cost option. Communicating the simplicity in its pricing serves to support Southwest’s branding and attracts the type of customers who value what the brand stands for. It may not always be the absolute least expensive option, but it is the straightforward one. In this manner, Southwest is using price and pricing transparency to differentiate itself in an industry often infamous for surcharges and additional fees. Its price transparency is high since the fee advertised (exclusive of taxes and other government and facility fees) will be what most customers actually pay. Pricing transparency is also high because of the lack of opaque fees that a customer may only uncover later. Spirit and Southwest take different approaches to transparency. Note, however, that I am not arguing that one method is ‘better’ than the other, whether in terms of business strategy or ethics. Each company is aligning its transparency with its contrasting mode of branding and customer segmentation. As Debbie Millman, an award-winning designer and educator, says, ‘branding is deliberate differentiation’ (as quoted in Stein, 2017). Customers can choose whether they want the simplicity of Southwest or the rockbottom fares of Spirit. It is not a matter of ethics whether the price of a paper boarding pass is explicit, as in Spirit’s case, or implicitly baked into the overall package, as it must be for Southwest. The SaaS imperative

While there isn’t an ethical imperative to adopt high levels of price and pricing transparency, the strategic imperative for SaaS companies is high. The digital economy gives 20

C hapter

2

P ublish Yo ur P rices

unparalleled ability to differentiate on transparency, making it an easier tool to wield for such companies. Additionally, there are several imperatives of a successful SaaS company that benefit from increased transparency: ■ ■ ■ ■

practicing value-based pricing lowering barriers to buy enabling customer self-selection demonstrating customer-centricity

Although these imperatives overlap and support each other, we explore each in turn. Practicing value-based pricing

Value-based pricing refers to pricing done from the point of view of value created by the offering for the customer. It contrasts with cost-plus pricing, in which pricing is largely determined by the cost of the offering for the producer, plus some desired profit margin. In all industries, cost-plus pricing is problematic for one simple reason: customers are not responsible for a company’s cost structure. Customers buy for value (benefits − price). If a company justifies having high prices because of its high costs and not because of its high value, customers will simply walk away. Cost-plus is an even worse guide for SaaS pricing. While marginal costs to produce are an important hurdle that price should pass over (sooner or later), they are relatively low for many SaaS companies. At the same time, the value that B2B software delivers in terms of customers’ increased revenue or decreased cost can be enormous. With such a large differential between cost to serve and value created, it would be absurd to constrain price by thinking in terms of cost-plus. The fact that SaaS companies typically have gross margins ([revenue − cost of goods sold] / revenue) upwards of 70%, much higher than many other industries, demonstrates the potential of pricing on value, not cost (’Impact of Gross Margin,’ 2021). SaaS companies should practice value-based pricing to maximize their profitability, and value-based pricing requires a certain level of price and pricing transparency to function well. Why must value and transparency go hand in hand? Value-based pricing requires that price and pricing be communicated well to customers. A company must demonstrate its value creation, showing how customers are better off buying its offerings than the competitive alternatives. To understand this story, customers must know what the prices are in order to make their own determination of value. Therefore, companies must have some measure of transparency for customers to understand the value claim inherent in value-based pricing. SaaS companies can create value that is orders of magnitude higher than their costs, making value-based pricing an especially strong imperative for such companies. For such strategies to work well, price and pricing must be communicated well to customers. In other words, value-based pricing requires transparency. Lowering barriers to buy

One key metric for SaaS businesses is the cost to acquire a customer, or CAC, defined as the sum of marketing and sales costs divided by the number of new customers. C hapter

2 

21

K yle T. W estra

CAC is important to calculate and track over time because it represents the amount of investment that must be overcome for a customer to become profitable. The more expensive it is to acquire a customer, the longer and harder it is to turn that customer into a profitable one. All else equal, the more obscure a company’s pricing, the higher its customer acquisition costs. If a company makes it so that a prospect must reach out and be handled by a sales representative instead of prequalifying itself or even completing a purchase online, the costs of that sales representative must be part of its CAC. Less transparent companies therefore will have higher CAC. The most straightforward way to reduce CAC is to decrease the numerator—that is, the sales and marketing costs associated with earning a new customer. And one of the best ways to do that is by increasing price and pricing transparency. As David Ulevitch, a general partner at the venture capital firm Andreessen Horowitz who focuses on SaaS investments, puts it: ‘You want to make it easy for your customers to give you money. And if your customers don’t understand your pricing, that’s a huge red flag’ (quoted in ’SaaS Go-To-Upmarket,’ 2020). This is especially true for serving small businesses where the customer lifetime value (CLV) typically cannot justify high CAC. In these situations, transparency can help bring such customers on board faster and more efficiently. Brad Coffey, then VP of Strategy at HubSpot, found that to be the case for their smaller customers, for whom they ‘focused on lowering CAC by removing friction from our sales process and moving more of our sales to the channel’ (quoted in Skok, n.d.). Doing so makes smaller customers more profitable to serve while freeing salespeople’s time for the larger, more complex sales in which their abilities can shine. SaaS companies can make it easier for their customers to purchase by increasing transparency around price and pricing. Doing so keeps acquisition costs low while increasing the speed at which customers start paying their subscriptions. As there is a trend toward greater transparency in SaaS generally, the competitive pressure to make it easy for prospects to turn themselves into customers will only increase. If a company is not easy to buy from, customers will go to one that is. Enabling customer self-selection

A successful sale requires not only that a prospect becomes a customer but that the customer purchases the right products and services for its goals. ‘Winning’ a sale that is against a customer’s interests is really sales creating a problem down the road for account management. It is important, therefore, that companies make it easier for customers to understand what exactly to buy. Price and pricing transparency help customers understand their options and selfselect the more appropriate one before getting far along a costly sales process and using up valuable sales time. This is especially true when a company has different versions or tiers of a product, such as good/better/best (Figure 2.1). In the above example, it would be impossible for a customer to properly select the right level of service in the absence of price transparency. Remember: value is simply benefits received minus price paid, so it is critical to delineate both benefits and price for a customer to make the right tradeoff for their unique situation. 22

C hapter

2

P ublish Yo ur P rices

ESSENTIALS You get: ● A

STANDARD You get: ● A ● B ● C

$19/mo.

PROFESSIONAL You get: ● A ● B ● C ● D ● E ● F ● G

$39/mo.

$99/mo.

Figure 2.1 Good/better/best product versioning.

The fit between a customer and the products and services purchased is an important indicator of a long-term positive commercial relationship. Equipping customers with the right information to make good selections themselves is likely to improve fit. This will positively impact customer churn (i.e., subscription cancelation), as the company avoids situations where, intentionally or not, a customer finds themselves paying too much for too much product and decides to walk away. Churn, along with CAC, is one of the most critical metrics for a SaaS company to track, as it dictates whether customers are staying with the offering long enough to turn profitable. Typically, SaaS companies depend not on a large upfront payment but on a series of subscription revenue to make a customer relationship profitable. If a customer churns before it becomes profitable, then the entire sale has actually made the company worse off. Customer self-selection is important not only at the moment of sale but as part of ongoing account management. If a customer’s goals change, they may need a different set of solutions. Ideally, the customer grows and requires a higher level of service. Transparency enables the customer to see that progression path and understand the relationship between price and benefits (that is, value) at different levels of service. However, transparency and a clear product hierarchy help even if the customer ends up needing a lower level of service in the future. Rather than churning, the customer may see how a smaller offering with a lower price still provides good value. Maintaining such a customer relationship enables the company to meet the customer’s new needs while keeping the door open to increasing service once again in the future. If higher transparency can reduce the risk of churn by increasing fit both at the first sale and throughout the account relationship, this is yet another way in which it can be the right strategy for a SaaS company.

Demonstrating customer-centricity

Many of the issues discussed so far wrap up into the concept of customer-centricity, something easy to understand in concept while hard to achieve in execution. What exactly is customer-centricity? Gartner, a leading technology research company, defines it as ‘the ability of people in an organization to understand customers’ C hapter

2 

23

K yle T. W estra

situations, perceptions, and expectations.’ But simple understanding is not enough: ‘Customer centricity demands that the customer is the focal point of all decisions related to delivering products, services, and experiences to create customer satisfaction, loyalty and advocacy’ (Gartner, n.d.). Making the customer the focal point of decision-making is easier said than done. However, there are decisions a company can make that clearly place it closer or farther away from customer-centricity. Choosing to hide prices and pricing can be a legitimate choice, but it is not customercentric. Lack of transparency is competitor-centric because it exposes that a fear of competitive reaction is what drives decision-making. ‘The most common reason not to display pricing is the idea that competitors could use price intelligence against them,’ argues business writer Kelly Main (2021) for Inc. Magazine. ‘But that line of thought places the focus on the competitors and not giving them what they want, rather than on consumers and giving them what they want.’ ‘If you are concerned about competition, the answer isn’t to hide pricing but to improve your offering,’ she concludes. Cost-plus pricing is also not customer-centric because it focuses a price justification on the producer’s cost, not the customer’s value. Remember: the customer is not responsible for covering a company’s cost structure. It is instead the company’s job to find a way to solve customers’ needs profitably. If a company hopes to serve customers profitably for the long term and better than the competition, it must practice customer-centricity.

Conclusion

The question of price and pricing transparency is less about right and wrong and more about what kind of relationship a company wants to have with its customers. Transparency can be a competitive advantage and a way to attract a different type of customer. Southwest is just one example of a company succeeding in part because of its high transparency, not despite it. Historically, most B2B companies have jealously protected their pricing. This can no longer be the default option, especially for SaaS companies. They must recognize that such a position comes with real costs. A SaaS company with low transparency must ask itself: what does that complexity and obfuscation get us? If the answer can be demonstrated to be stickier customers, higher pricing power, and/or better margins, those are real benefits. But even if such benefits exist, they must be weighed against the substantial drag to commercialization and monetization discussed earlier. There is no single right answer when it comes to price and pricing transparency. While there are strong forces in favor of increased transparency for SaaS companies, each organization must make an intentional, strategic decision that represents its particular circumstances. Just as there is no one-size-fits-all for companies, each customer segment should be evaluated separately. It is, of course, possible to have different levels of price and pricing transparency for different types of customers. Many B2B SaaS companies find it useful

24

C hapter

2

P ublish Yo ur P rices

to publish prices for small and medium-sized business customers (or lower levels of service) while protecting prices for enterprise customers (or the most premium level of support) behind a notice to the effect of ‘call us for more information,’ ‘price available upon request,’ and so on. This approach can be useful for multiple reasons. First, it acknowledges that any contracting at the specified product level is likely to include bespoke conditions and require the expertise of a salesperson to put together. Second, it maintains for the supplier a higher level of pricing power in a negotiation with a customer that may have a sophisticated procurement department. Third, as Prof. Utpal Dholakia (2021) of Rice University points out, ‘making the customer call for price raises their commitment to complete the buying journey,’ thus helping to make each interaction more likely to result in a sale. This all supports the premise that price and pricing transparency is neither an ethical imperative nor a straightforward driver of business success but a strategic tool to better interact with and sell to a company’s customers. There is no simple answer. Different companies can succeed by using transparency in different ways with different customer segments. With that said, the weight of technological forces, competitive dynamics, customer expectations, and particular commercialization goals of SaaS companies strongly suggest moving in a certain direction. If executives want better value-based pricing, faster sales cycles, more accurate customer segmentation, and increased customer-centricity, there is a simple solution: publish your prices.

Bio

Kyle T. Westra is a pricing strategist who focuses on elevating technology companies. He works with startups through multibillion-dollar global companies to design and execute strategies for product packaging, monetization, and commercialization. His Amazon-bestselling book, The New Invisible Hand: Five Revolutions in the Digital Economy, explores the promise and peril that new technologies are bringing to business strategy. He has also written for publications such as the Washington Post and the Journal of Revenue and Pricing Management. Kyle holds an MBA with distinction in Business Strategy and Marketing Analysis from the Kellstadt Graduate School of Business at DePaul University and a BA in Political Science and Economics from Tufts University.

Key objectives 1. Identify the differences between price transparency and pricing transparency. 2. Recognize the key ways in which higher transparency can commercially benefit a company. 3. Understand how these dynamics are especially strong in B2B SaaS companies.

C hapter

2 

25

K yle T. W estra

Key summary points 1. Price and pricing transparency are important strategic tools that companies can use to connect and transact with customers better than the competition can. 2. Especially strong SaaS dynamics that encourage higher transparency are these: ■ practicing value-based pricing ■ lowering barriers to buy ■ enabling customer self-selection ■ demonstrating customer-centricity 3. While low transparency may have some benefits, these must be weighed against the real disadvantages it creates for the customer buying experience.

Key questions 1. Do customers find it easy to transact with your company? ■ If not, what initiatives are underway to make this easier? 2. How can your company better differentiate itself from the competition using price and pricing transparency? 3. Has your company made strategic decisions about its level of transparency or accepted a status quo from industry competition?

References Dholakia, U. (2021, August 25). Availability-based price transparency. The Pricing Conundrum. https://thepricingconundrum​.substack​.com ​/p​/decoding​-three​-types​-of​-price​-related. Gartner. (n.d.). Customer centricity. https://www​.gartner​.com ​/en ​/marketing ​/glossary​/customer​ -centricity. Main, K. (2021, August 5). Salesforce’s $17B pricing lesson every startup needs to know. Inc. https://www​. inc​. com ​/ kelly​- main ​/salesforces​-17b ​- pricing​- lesson​- every​- startup ​- needs​- to​ -know​.html. SaaS go-to-upmarket. (2020, May 29). Future. https://future​.a16z​.com ​/podcasts​/saas​-go​-to​ -upmarket/. Skok, D. (n.d.). SaaS metrics 2.0—A guide to measuring and improving what matters. For Entrepreneurs. https://www​.forentrepreneurs​.com ​/saas​-metrics​-2. Southwest Airlines. (n.d.). Transfarency. Retrieved December 30, 2021, from https://www​ .southwest​.com​/ html​/air​/transfarency/. Spirit Airlines. (n.d.). Optional services. Retrieved December 30, 2021, from https://www​.spirit​ .com ​/optional​-services. Stein, L. (2017, December 18). The key to brand design is ‘deliberate differentiation.’ BrandingMag. https://www​.brandingmag​.com ​/2017​/12 ​/18​/the​-key​-to​-brand​-design​-is​ -deliberate​-differentiation. The impact of gross margin on SaaS valuations. (2021, March 12). Retrieved December 30, 2021, from https://softwareequity​.com​/the​-impact​-of​-gross​-margin​-on​-saas​-valuations/. Westra, K. T. (2019). The new invisible hand: Five revolutions in the digital economy. New Degree Press.

26

C hapter

2

C hapt e r 3

Dynamic Pricing Process How to Transition from Fixed to Dynamic Pricing? Arnd Vomberg

Overcoming the paradigm of fixed prices

Changes in the way companies set their prices have been at the center of interest for the last decade. Readers of the Wall Street Journal most likely have witnessed headlines such as these: ■ ■ ■

‘Coming Soon Toilet Paper Priced Like Airline Tickets’ (Angwin & Mattioli, 2012), ‘The High-Speed Trading Behind Your Amazon Purchase’ (Mims, 2017), and ‘Now Prices Can Change from Minute to Minute’ (Nicas, 2015).

Those headlines address an essential manner in which companies nowadays set their prices: companies increasingly break with last century’s norm of fixed prices and move to dynamic prices. Increased data availability and new technologies caused the rise of dynamic pricing. New technologies help companies to identify consumer behavior patterns more quickly and efficiently and align prices accordingly. Simultaneously, increased online price transparency pressures companies to monitor and respond in real time to competitive prices (Fisher et al., 2018). Depending on the industry, studies indicate a revenue-increasing potential of dynamic pricing between 2% and 8% and potential profit increases between 3% and 25% (BenMark et al., 2017; Kimes & Wirtz, 2003). However, such profit enhancements do not set in automatically but depend on retailers’ focal decisions (Vomberg, 2021). As pricing strategies are hard to reverse and costly to adjust, retailers need to carefully plan and execute a transition from fixed to dynamic pricing. This chapter develops a systematic dynamic pricing process that should help retailers unlock hidden profits with their pricing approach. I based this chapter on the state of the art of academic knowledge and industry reports. I will refer to in-depth interviews from Vomberg et al. (2020), which will provide important practical insights. We interviewed 11 online and multichannel retailers (referred to as R1–R11) and nine

DOI: 10.4324/9781003226192-5

27

A r n d V o mberg

solution providers (i.e., companies that develop commercial dynamic pricing software, referred to as S1–S9).

Forms of dynamic pricing

I will distinguish between two forms of dynamic pricing: time-based dynamic pricing and personalized pricing. Retailers who practice time-based dynamic pricing change their prices over time based on various factors (e.g., competitor prices, demand). Thereby, they focus on the temporal component of price differentiation. Time-based dynamic prices are the same for all consumers. By contrast, personalized pricing leads to different prices that consumers pay at the exact moment in time. Based on their individual willingness to pay—which retailers may approximate via consumer behavior (e.g., browsing history), personal characteristics (e.g., age), or their location information—consumers either pay individual prices or see product rankings in personalized order (i.e., price steering). It is worth noting that combinations between time-based dynamic pricing and personalized pricing are possible. Dynamic pricing: Price differences over time

Two key dimensions can describe dynamic pricing: frequency and range of price changes. Frequency refers to the number of price changes over a specific period (e.g., one month). For instance, media reports demonstrate that the price of a digital camera on Amazon​.c​om changed 275 times within three days. Another study counted three million price changes on Amazon in Germany for one day. The range can refer to the range of individual price changes or the range of prices within a specific time window (e.g., highest and lowest price in one week). A systematic review of online prices revealed that individual price changes could be substantial. The median absolute size of a price change is 11% in the US (Gorodnichenko et al., 2018). In addition, prices can differ broadly within a specific period. For instance, industry observers noted that a digital camera’s price changed within hours by up to 240% on Amazon​.co​m.

Personalized pricing: Price differences among consumers

Personalized pricing requires some form of individualization. Companies use the information that consumers leave behind as their ‘digital traces’ to individualize prices. Companies can offer personalized baseline prices, personalized coupons, a personalized ranking of products (‘price steering’), or base prices on consumers’ locations. Personalized baseline prices. Retailers can show different prices to different consumers when they enter the website. Retailers base those prices on some knowledge of the focal consumers’ interests and behaviors. Retailers may rely on information that consumers share deliberately with them. For instance, they could use personal details that consumers provide when creating an account on their website or information that consumers provide on ‘price wishlists’ (Vogelsang, 2020). Furthermore, retailers can (more) ‘secretly’ collect consumer information. Retailers can rely on cookie data to acquire demographic information, track a consumer’s browsing history, leverage customer journey information, consumer’s operating system (e.g., Windows vs. iOS), or offer different prices based on whether consumers are using a smartphone, tablet, or PC. 28

C hapter

3

D y n amic P rici n g P r o cess

Personalized coupons. Instead of displaying different baseline prices, retailers could individualize prices with personalized coupons. For instance, the Dutch supermarket Albert Heijn regularly offers special discounts based on consumer loyalty card information. Offering personalized coupons has several advantages over offering personalized prices. First, research indicates that consumers consider personalized coupons fairer than personalized prices (Weisstein et al., 2013). Second and relatedly, personalized coupons (in contrast to personalized base prices) create the impression among consumers to realize a bargain. A drawback, however, is that there is initial evidence that consumers can become used to receiving personalized discounts. Consumers then perceive the absence of such coupons as an adverse event, comparable to a price increase. Price steering. Retailers can also employ price steering: two consumers see different product results or the same products in a different order for the same search term. Search results presented earlier (e.g., on the first page) typically have a higher chance of being selected than products shown later. Thus, retailers can steer consumers toward buying higher-priced items. For example, the online travel agency Orbitz Worldwide Inc. inferred that Apple Inc.’s Mac computer users would be willing to spend around 30% more a night on hotels. So Orbitz steered those consumers to costlier travel options: Mac users have seen more expensive options first, whereas non-Mac users have seen less costly alternatives (Mattioli, 2012). Retailers also combine price steering with personalized coupons. For instance, a retailer elaborated:

R5a: Cookies automatically detect it: you belong to the sports category, you use a Mac, which means you probably accept higher prices; therefore, I rank the products with a higher price first, and I give you 10% on the sports segment.

Location-based pricing. Location-based pricing represents a basic form of personalized pricing. It refers to any price differences of products sold by the same retailer simultaneously between geographical locations. Expert interviews indicated that retailers value this option (Vomberg et al., 2020). Online retailers may rely on location information (e.g., IP addresses) to differ prices between countries. Studies observe online price differences between countries ranging from 21% (Mikians et al., 2012) to 700% (Iordanou et al., 2017). Initial anecdotal evidence also suggests location-based pricing within a country (e.g., based on GPS data). Reports revealed that the office-retail supplier Staples​.c​om displayed different prices to different people after estimating their locations. Staples​.c​ om might have also used information on the person’s distance to the next competing offline store for setting the price (Valentino-DeVries et al., 2012). Dynamic pricing process

Dynamic pricing represents an ongoing process (Figure 3.1). The dynamic pricing process covers the phases of dynamic pricing strategy, decisions about dynamic price setting, dynamic pricing implementation, and dynamic price auditing. In addition, the process contains feedback loops. C hapter

3 

29

A r n d V o mberg Phase

Exemplary questions

Dynamic Pricing Strategy

What are the retailer’s dynamic pricing goals? What is the retailer’s price positioning?

Dynamic Price Setting a) General Decisions b) Context Decisions

Which dynamic pricing approach to select? How to implement the dynamic pricing approach? What is the base for dynamic pricing? Which product portfolio factors are important? Which legal boundaries need to be considered?

Dynamic Pricing Implementation

How to implement dynamic pricing towards the customer? How to implement dynamic pricing internally? How to implement omnichannel dynamic pricing?

Dynamic Pricing Auditing

Have established goals been met? In which areas are adjustments necessary?

Figure 3.1 The dynamic pricing process with exemplary questions.

Dynamic pricing strategy

Before developing the dynamic pricing approach, retailers need to consider their general pricing strategy. Managers need to establish which goals they want to achieve with a transition to dynamic pricing. For instance, they could establish profit-, salesvolume-, or growth-oriented goals. Retailers need to evaluate how they want to position themselves relative to competitors in terms of price and the customer value they create (e.g., delivery times). For instance, they may establish a discounter position (low relative price and customer value), a middle-class position (medium relative price and customer value), or a premium position (high relative price and customer value). These strategic decisions set the framework within which retailers can and should develop their dynamic pricing approach.

Dynamic price setting Which dynamic pricing approach to select?

Retailers should implement the dynamic pricing approach that best aligns with their pricing strategy. Studies suggest that time-based dynamic pricing likely becomes a competitive standard (at least in online markets). For instance, a study documents that consumer electronics companies dynamically price between 36% (Conrad Electronics) and 65% (Media Markt) of their products. The same study finds comparable proportions for automotive accessory and mail-order companies (Verbraucherzentrale Brandenburg, 2020). Similarly, a US study shows that online retailers change the prices of around 10% of their products on a daily level (with values ranging from 1% to 37%) (Brown & MacKay, 2021). By contrast, studies hardly find evidence of personalized (baseline) prices (e.g., Iordanou et al., 2017). Studies provide evidence on location-based pricing; however, it still qualifies as a niche phenomenon (Hupperich et al., 2018; Iordanou et al., 2017). Industry experts note that location-based pricing is not frequently used within a country in online markets for two reasons (Vomberg, 2021). First, regulations (such as from 30

C hapter

3

D y n amic P rici n g P r o cess

the German Federal Ministry of Justice and Consumer Protection) stand in the way of this approach. Second, experts also mentioned technological problems: in their opinion, GPS data would currently not be sufficiently accurate for location-based pricing. Most dominantly, retailers rely on personalized coupons and price steering (Hannak et al., 2014; Vomberg, 2021). An expert argues:

S8: So, for Europe, individual pricing is the absolute exception. That has to be said quite clearly. The reason for this is a combination of technical limitations and a deliberate shying away from possible adverse effects.

How to implement the dynamic pricing approach?

Companies can either rely on self-developed algorithms or use external IT solutions. Expert interviews revealed that companies most dominantly rely on external IT solutions since their internal development would require extensive efforts (e.g., programming and maintenance) (Vomberg et al., 2020). I classify externally developed solutions into repricing software and more advanced pricing tools. Repricing software is one way to implement time-based dynamic pricing. It automatically lowers the retailer’s price until its price is below competitors’ prices or until it reaches a predetermined lower bound for the price—for instance, the software scrapes competitor prices from price comparison tools such as Idealo or Google Shopping. A crucial decision that retailers need to make is to specify the relevant competitors and which products or product categories prices should be adjusted. In addition, retailers need to specify the interval of price adjustments (e.g., real-time, daily, or weekly). Advanced pricing tools consider further data beyond competitor prices. For instance, these tools rely on internal inventory levels or derive optimal prices via machine learning algorithms. Such tools lower prices to beat competitors and conduct market experiments; they track and learn competitor and customer responses to price changes (Mims, 2017).

What is the base for dynamic pricing?

Overall, the effectiveness of the dynamic pricing approach depends on which determinants companies include in their algorithm. I classify these determinants into marketrelated, customer-related, and company-related factors. Figure 3.2 summarizes the discussion. Market-related factors. Most companies rely on competitor prices to inform their dynamic pricing strategy (Brown & MacKay, 2021). For instance, one expert notes:

S2: Experience shows that online retailers are exposed to such high competitive pressure that they have no choice but to respond to their competitors’ prices.

C hapter

3 

31

A r n d V o mberg

Dynamic Pricing

Time-based Dynamic Pricing

Personalized Pricing

• Same price for all consumers • Frequent price changes

• Different prices for different consumers • Different implementation forms: 1. Personalized baseline prices 2. Personalized coupons 3. Price steering 4. Location-based pricing

Potential information sources: • Competitor prices • Aggregated consumer price elasticity • Inventory level • Strategic considerations

Dynamic pricing information sources in combination with personal information such as: • Loyalty cards • Purchase histories • Browsing behavior • Operating systems • Device information • GPS-data • IP-addresses

Figure 3.2 Summarizing overview of dynamic pricing (Vomberg, 2021, reproduced with the kind permission of the publisher).

Media reports also discuss the time of the day or weather as potential determinants for dynamic pricing. This idea already dates back to the 1990s. The Coca-Cola Company experimented with raising prices in vending machines on hot days (Nicas, 2015). Expert interviews suggest that these factors are only adequate determinants of consumers’ willingness to pay in specific settings, rendering these factors less important dynamic pricing determinants (Vomberg et al., 2020). Customer-related factors. Companies could conduct randomized market experiments to learn how much consumers will buy at different price points (Fisher et al., 2018). Alternatively, solution providers indicated that they infer customers’ price sensitivities from real-time click and real-time transaction data (Vomberg et al., 2020). For instance, if more customers click on the same article, the focal article’s price might increase. This approach does not need to represent personalized pricing because it can rely on aggregated customer data and increase prices for all customers. Company-related factors. Retailers naturally also need to align their dynamic pricing approach with their internal goals. Companies can consider purchase prices and their inventory levels. For instance, fashion retailers may dynamically vary the prices of clothes during the season to avoid heavy discounts during end-of-season sales.

Context factors for price setting Which product portfolio factors are important?

Company-related factors. Companies need to decide which products they price dynamically and, relatedly, if they rely on different price-setting determinants for 32

C hapter

3

D y n amic P rici n g P r o cess

different products. This decision depends on the comparability of the products, the brand, and whether the product constitutes a key-value item. First, the ease with which consumers can obtain prices for different products determines how competitor prices should inform the dynamic pricing approach. Among the products with a high degree of comparability are consumer electronics or national brands. Product categories with a low degree of comparability are, for example, unique fashion items, furniture, or private brand products. Competitor prices are important for highly comparable products. Second, dynamic pricing likely differs between private-label brands, national brands, and luxury brands. Private-label brands are developed and produced directly by a retailer or manufacturer and generally sold exclusively by that retailer. Companies can focus on consumer price sensitivities and internal considerations since there are no competitor prices for private-label brands. By contrast, national brands often have easy-to-search names (e.g., Nike Airmax). Thus, they are highly comparable and therefore subject to intensive price pressure. Additionally, national brands are essential from a retailer’s perspective, as they pull customers into the shop. For those national brands, typically, competitor prices should be considered. Finally, experts claimed that luxury brands are less susceptible to price, therefore, competitor prices are relevant to a lesser extent (Vomberg et al., 2020). Frequent price changes may even negatively impact the brand’s high-quality image. Third, a final important consideration is the role of key-value items (sometimes referred to as product heroes). Key-value items pull customers into the store or initiate a purchase. In addition, key-value items drive customers’ perceptions of the overall price image of the assortment. For key-value items, retailers need to rely on competitive prices. Interviews indicated that even retailers that only cautiously implement dynamic pricing, systematically rely on competitive prices for key-value items. Likewise, most solution providers offer a particular module for key-value items, which calculates their optimal prices more frequently than other products (e.g., BenMark et al., 2017). Which legal boundaries need to be considered?

Recent regulations such as the General Data Protection Regulation (GDPR, enacted in 2018) or the California Consumer Privacy Act (CCPA, enacted in 2020) restrict consumer/user data usage. They require companies to disclose all data-processing activities. Furthermore, in many cases (e.g., GDPR), companies are only allowed to process consumer data after consumers give their consent. Any information that directly or indirectly relates to a person qualifies as personal data, or stated differently, any identifier connected to an individual qualifies as personal data. Instead of only focusing on the consumers’ names and addresses according to privacy regulations, IP addresses and cookie identifiers also qualify as personal data (Bleier et al., 2020). Such regulations affect time-based dynamic pricing to a lesser extent, as it typically relies on competitor-related and company-internal information and only analyzes aggregate consumer data. However, personalized pricing requires that companies transparently communicate to consumers that they use their data for personalized pricing and request consumer consent (Spann & Skiera, 2020). Experts raise severe doubts that consumers will consent to personalized pricing (Borgesius & Poort, 2017). C hapter

3 

33

A r n d V o mberg

Importantly, GDPR also rules out that retailers process certain kinds of data (also referred to as ‘sensitive data’) in general. Thus, companies are not allowed to perform personalized pricing on such sensitive data. Sensitive data are personal data revealing racial or ethnic origin, political opinions, religious beliefs, or trade union membership, and the processing of genetic data, biometric data for the purpose of uniquely identifying a natural person, data concerning health or data concerning a natural person’s sex life or sexual orientation. (Article 9(1) GDPR)

Dynamic pricing implementation How to implement dynamic pricing toward the customer?

A focal consideration is how consumers react to dynamic pricing. Research shows that consumers perceive both time-based dynamic pricing and personalized pricing as unfair. However, consumers react more negatively to personalized pricing (Haws & Bearden, 2006). Personalized pricing may lower a consumer’s trust in the retailer (e.g., Gabarino & Lee, 2003) and repurchase intentions (Gabarino & Maxwell, 2010). To overcome potential consumer resistance, retailers can lower the frequency of price changes. In addition, retailers can establish a complaint management approach and offer refunds. However, consumers likely become used to dynamic pricing. As a consequence, they likely react less negatively to dynamic pricing with time (Vomberg et al., 2021). At the same time, however, once consumers internalize frequent price changes, they may become more price-sensitive. Consequently, they may compare prices more frequently and switch to competitors (Vomberg et al., 2021).

How to implement dynamic pricing internally?

The implementation of dynamic pricing requires that companies delegate pricing decisions from humans to algorithms. While pricing managers or the procurement department traditionally set prices, using an automated dynamic pricing algorithm implies a loss of control of former pricing managers. Naturally, this requires that many companies go through significant cultural changes. One retailer notes the following:

R3: It’s not just a question of technology …, but also has to do with change and cultural change … So it’s a question of mentality … it’s part of the big picture, the extent to which a company is willing to adapt and radically introduce processes and changes in the company. So dynamic pricing is not just software … it’s usually a fundamental change in the way you’ve been doing things.

Implementing cultural changes requires that companies align formal and informal elements. Regarding formal elements, managers, for example, need to ensure that dynamic

34

C hapter

3

D y n amic P rici n g P r o cess

pricing algorithms align with incentive schemes. For instance, pricing algorithms that maximize profits conflict with incentive schemes based on volume targets. Regarding informal elements, managers need to create a cultural mindset that is open to pricing automation, which likely is a challenging task since people can have an aversion to algorithms—they refrain from using algorithms and do not trust them (Dietvorst et al., 2018). Managers should involve pricing managers in the configuration of pricing algorithms to lower such aversion. Furthermore, managers could roll out dynamic pricing for a subset of products and jointly evaluate the effectiveness with pricing managers (Bondi et al., 2021).

How to implement omnichannel dynamic pricing?

Dynamic pricing can constitute a crucial challenge for omnichannel retailers (i.e., simultaneous usage of online and offline channels). Omnichannel companies need to make two decisions: they need to decide whether they aim for price consistency or differentiation between online and offline channels. In addition, they need to determine how they implement dynamic pricing. First, empirical studies show that companies typically set equal prices between channels. One systematic empirical investigation shows that companies can hardly realize higher offline prices. Only for high-priced and take-away items do such higher offline prices seem possible (Homburg et al., 2019). Second, although companies typically strive for a price-consistency strategy, competitive online pressures can force retailers to employ dynamic pricing and accept temporary price differences between online and offline channels. Companies can resolve these tensions in different ways. Companies that do not sell all products in both their online and offline channels can focus on dynamic pricing only for the online channel and/or adopt dynamic pricing only for key-value items. Also, since the frequency of price changes is typically lower in offline channels, companies may reduce the frequency of price changes in online channels (Bondi & Sen, 2021).

Dynamic pricing auditing

Successful dynamic pricing needs ongoing auditing. Retailers need to determine which insights they want to gain from auditing (i.e., the goals of the analysis) and what should be studied (i.e., the unit of analysis). Retailers should evaluate whether the dynamic pricing approach meets specified goals. Moreover, retailers should identify which adjustments are necessary. Such adjustments can include changes in the dynamic pricing algorithm, which might be necessary due to changes in the competitive landscape or in consumer behavior.

Conclusion

Dynamic pricing can contribute to company success. However, retailers need to carefully design their dynamic pricing approach to unfold their full profit-creating potential. Furthermore, retailers need to overcome challenges when implementing a dynamic pricing approach. Based on the current state of academic research and interviews

C hapter

3 

35

A r n d V o mberg

with managers, I discussed focal decisions managers need to make in designing their dynamic pricing approach. In addition, I discussed how retailers could address focal implementation challenges. Overall, this chapter guides managers to unlock hidden profits with their dynamic pricing approach.

Acknowledgments

This chapter is an updated and revised version (with the kind permission of the initial publisher); Vomberg (2021) is an earlier version of this book chapter, published in The Digital Transformation Handbook with support from the Groningen Digital Business Center.

Bio

Arnd Vomberg is Professor of Digital Marketing and Marketing Transformation in the Marketing and Sales Department of the University of Mannheim. Before joining the University of Mannheim, he was an Associate Professor with Tenure at the University of Groningen (The Netherlands). Arnd Vomberg’s research focuses on digital marketing and marketing transformation. He studies omnichannel strategies, online pricing, the interplay between artificial intelligence and human beings, marketing automation, agile transformation, marketing technology, and marketing’s impact on employees. His research has been published in several leading journals of the field, including Journal of Marketing, Journal of Marketing Research, Strategic Management Journal, Journal of the Academy of Marketing Science, and International Journal of Research in Marketing. Arnd Vomberg has received several awards for his research, including the Ralph Alexander Best Dissertation Award from the Academy of Management.

Key objectives 1. Readers will be able to develop a dynamic pricing approach. 2. Readers will be able to evaluate different dynamic price-setting options. 3. Readers will learn how to overcome focal implementation challenges of dynamic pricing.

Key summary points 1. Retailers can unlock hidden profits with an adequate dynamic pricing approach. 2. Dynamic pricing represents an ongoing process including the phases of dynamic pricing strategy, dynamic price setting, dynamic pricing implementation, and dynamic pricing auditing. 3. Retailers need to overcome external and internal resistance to dynamic pricing.

36

C hapter

3

D y n amic P rici n g P r o cess

Key questions 1. Which forms of dynamic pricing can retailers select? 2. Which influencing factors should shape a company’s dynamic pricing approach? 3. How can retailers implement dynamic pricing internally and externally?

References Angwin, J., & Mattioli, D. (2012). Coming soon: Toilet paper priced like airline tickets. Wall Street Journal, September 5, 2012. https://www​.wsj​.com ​/articles​/SB1​0 000​8723​9639​0 444​ 9149​0 457​7617​3331​30724846. BenMark, G., Klapdor, S., Kullmann, M., & Sundararajan, R. (2017). How retailers can drive profitable growth through dynamic pricing. McKinsey & Company. https://www​.mckinsey​ .com/~ ​/media ​/ McKinsey​/ Industries ​/ Retail ​/ Our​%20Insights ​/ How​%20retailers​%20can​ %20drive​%20profitable​%20growth​%20through​%20dynamic​%20pricing​/ How​-retailers​ -can​-drive​-profitable​-growth​-through​-dynamic​-pricing​.pdf​?shouldIndex​=false. Bleier, A., Goldfarb, A., & Tucker, C. (2020). Consumer privacy and the future of data-based innovation and marketing. International Journal of Research in Marketing, 37(3), 466–480. Bondi, S., Goldrick, M., Reasor, E., Sen, D., & Wilkie, J. (2021). The dos and don’ts of dynamic pricing in retail. McKinsey & Company. https://www​.mckinsey​.com​/ business​-functions​/ marketing​-and​-sales​/our​-insights​/the​-dos​-and​-donts​-of​-dynamic​-pricing​-in​-retail. Bondi, S., & Sen, B. (2021). The power—and pitfalls—of dynamic pricing for omnichannel retailers. McKinsey & Company. https://www​.mckinsey​.com/~​/media ​/mckinsey​/ business​ %20functions​ / marketing​ % 20and​ % 20sales​ / our​ % 20insights​ / the​ % 20power​ % 20and​ %20pitfalls​%20of​%20dynamic​%20pricing​%20for​%20omnichannel​%20retailers​/ the​ -power​-and​-pitfalls​-of​-dynamic​-pricing​-for​-retailers​.pdf​?shouldIndex​=false. Borgesius, F. Z., & Poort, J. (2017). Online price discrimination and EU data privacy law. Journal of Consumer Policy, 40(3), 347–366. Brown, Z. Y., & MacKay, A. (2021). Competition in pricing algorithms. American Economic Journal: Microeconomics. Dietvorst, B. J., Simmons, J. P., & Massey, C. (2018). Overcoming algorithm aversion: People will use imperfect algorithms if they can (even slightly) modify them. Management Science, 64(3), 1155–1170. Fisher, M., Gallino, S., & Li, J. (2018). Competition-based dynamic pricing in online retailing: A methodology validated with field experiments. Management Science, 64(6), 2496–2514. Garbarino, E., & Lee, O. F. (2003). Dynamic pricing in internet retail: Effects on consumer trust. Psychology & Marketing, 20(6), 495–513. Garbarino, E., & Maxwell, S. (2010). Consumer response to norm-breaking pricing events in e-commerce. Journal of Business Research, 63(9–10), 1066–1072. Gorodnichenko, Y., Sheremirov, V., & Talavera, O. (2018). Price setting in online markets: Does IT click? Journal of the European Economic Association, 16(6), 1764–1811. Hannak, A., Soeller, G., Lazer, D., Mislove, A., & Wilson, C. (2014). Measuring price discrimination and steering on e-commerce web sites. In Proceedings of the 2014 internet measurement conference (pp. 305–318). Association for Computing Machinery. Haws, K. L., & Bearden, W. O. (2006). Dynamic pricing and consumer fairness perceptions. Journal of Consumer Research, 33(3), 304–311. Homburg, C., Lauer, K., & Vomberg, A. (2019). The multichannel pricing dilemma: Do consumers accept higher offline than online prices? International Journal of Research in Marketing, 36(4), 597–612.

C hapter

3 

37

A r n d V o mberg Hupperich, T., Tatang, D., Wilkop, N., & Holz, T. (2018). An empirical study on online price differentiation. In Proceedings of the eighth ACM conference on data and application security and privacy (pp. 76–83). Association for Computing Machinery. Iordanou, C., Soriente, C., Sirivianos, M., & Laoutaris, N. (2017). Who is fiddling with prices? Building and deploying a watchdog service for e-commerce. In Proceedings of the conference of the ACM special interest group on data communication (pp. 376–389). Association for Computing Machinery. Kimes, S. E., & Wirtz, J. (2003). Has revenue management become acceptable? Findings from an international study on the perceived fairness of rate fences. Journal of Service Research, 6(2), 125–135. Mattioli, D. (2012). On Orbitz, Mac users steered to pricier hotels. Wall Street Journal, August 23. https://www​.wsj​.com​/articles​/SB1​0 001​4240​5270​2304​4586​0 457​7488​8226​67325882. Mikians, J., Gyarmati, L., Erramilli, V., & Laoutaris, N. (2012). Detecting price and search discrimination on the internet. In Proceedings of the 11th ACM workshop on hot topics in networks (pp. 79–84). Association for Computing Machinery. Mims, C. (2017). The high-speed trading behind your Amazon purchase. Wall Street Journal, March 29. https://www​.wsj​.com​/articles​/the​-high​-speed​-trading​-behind​-your​-amazon​ -purchase​-1490532110. Nicas, J. (2015). Now prices can change from minute to minute. Wall Street Journal, December 14. https://www​.wsj​.com ​/articles​/now​-prices​- can​- change​-from​-minute​-to​-minute​-1450057990. Spann, M., & Skiera, B. (2020). Dynamische Preisgestaltung in der digitalisierten Welt. Schmalenbachs Zeitschrift für betriebswirtschaftliche Forschung, 72(3), 321–342. Valentino-DeVries, J., Singer-Vine, J., & Soltani, A. (2012). Websites vary prices, deals based on users’ information. Wall Street Journal, December 24. https://www​.wsj​.com ​/articles​/SB1​ 0001​4241​2788​7323​7772​0 457​8189​3918​13881534. Verbraucherzentrale Brandenburg. (2020). Unterschiedliche Preise im Netz. https://www​ .verbraucherzentrale ​-brandenburg​.de ​/ wissen ​/digitale ​-welt ​/onlinehandel ​/unterschiedliche​ -preise​-im​-netz​-28618. Vogelsang, M. (2020). Designing smart prices. Springer Gabler. Vomberg, A. (2021). Pricing in the digital age: A roadmap to becoming a dynamic pricing retailer. In Tammo Bijmolt, Thijs Broekhuizen, Bas Baalmans, & Nicolai Fabian (Eds.), The digital transformation handbook—From academic research to practical insights. https:// www​.rug​.nl​/gdbc​/ blog ​/pricing​-in​-the​-digital​-age. Vomberg, A., Homburg, C., & Sarantopoulos, P. (2021). Dynamic pricing: How do consumers react? And how should retailers respond? [Working Paper]. Vomberg, A., Lauer, K., & Weitkämper, K. (2020). Dynamic pricing: Preisfindung auf elektronischen Marktplätzen. In Handbuch Digitale Wirtschaft (pp. 653–677). Springer Gabler. Weisstein, F., Monroe, K., & Kukar-Kinney, M. (2013). Effects of price framing on consumers’ perceptions of online dynamic pricing practices. Journal of the Academy of Marketing Science, 41(5), 501–514.

38

C hapter

3

C hapt e r 4

Realizing Your Monetization Potential Needs Customer Value Management Mrinal (MG) Gurbaxani and Alex Smith

Most B2B industries are transforming digitally and leveraging their data to develop innovative business models, and with that, delivering flexible subscription-based offers to these new services and solutions. OEMs that produce trucks, ship engines, and computer servers are all moving toward delivering total solutions ‘as-a-service.’ Realizing the monetization potential of these flexible subscriptions requires industrial companies to develop a key new capability: customer value management. This trend began in the software industry, with Salesforce being the pioneer toward SaaS and, 20 years later, it’s gaining traction in SKU-oriented companies as well. In industrial companies, to steer clear of the commodity trap, savvy commercial excellence organizations are looking to transform from an SKU orientation to a solution orientation by packaging their product and wrapping in services, maintenance, support, upgrades, financing, monitoring, replenishment, and other value-add services. This enables businesses to create differentiation, increase margins, improve customer experiences, and create long-term stickiness (if the promised customer success is realized). When done right, it is a win-win. The SaaS industry quickly learned that business success, and true profitable growth, depends much more on the adoption of their solutions and renewals across their customer base than on acquiring a new customer. When selling products, businesses cared if you were successful, but once a product was purchased, businesses shifted to focus on the next sale in the pipeline. Each quarter, sales would reset to zero and there was a new quota number to achieve. But in this new evolution of selling SaaS and the establishment of annual recurring revenue (ARR) models, monetization models depend heavily on whether customers expand and renew, or not, based on the success and business value they see from their purchases. The subscription business model is therefore all about customer retention. Leading SaaS businesses invest heavily in customer success in the form of onboarding, implementation, and advisory services to ensure ARR growth. As a result, the focus on new

DOI: 10.4324/9781003226192-6

39

M ri n al ( M G ) G urbaxa n i a n d A lex S mith

KPIs like customer lifetime value or specifically net dollar retention rate in the boardroom has forced a culture of focusing on where customer value lies. While selling on value is an age-old concept, the need for consistent and scalable value management is amplified in the B2B subscription world. Customer-centricity can no longer be aspirational—a deep understanding of how your customers get value from your solution, and how much, is now critical to a subscription growth model. It is this structured understanding of customer benefits that allows for a measured landand-expand approach in driving ARR growth. Over the past decade, SaaS solutions have taken over the enterprise software market. As of 2018, SaaS solutions made up over 75% of the market, up from 6% in 2010, and the revolution shows no sign of slowing. In fact, customers and enterprise IT providers have adopted the idea with such enthusiasm that flexible-consumption models have been extended to just about everything in the corporate IT space, from hardware to cybersecurity services. But those economic benefits have come with a trade-off. Where previously customers might have paid five plus years’ worth of license fees upfront, effectively locking them in, the modern SaaS and subscription customer enjoys a very different power dynamic. The same will be true of all XaaS businesses, from equipment-as-a-service to infrastructure-as-a-service. This shift in power to the customer means that businesses are scrutinizing their purchases as never before. The subscription customer now has less invested, more demands, more options, more flexibility to buy, and more flexibility to churn. Their focus becomes ‘Am I realizing my business outcomes and seeing measurable value for money?’ As a direct consequence, B2B subscription companies are having to adapt, to ensure that after a customer buys, they continue to buy again and again. That means how companies manage customer relationships must also change. To stay competitive in the 2020s, software vendors now need to collaboratively discover the value, communicate and sell the value, and then continuously demonstrate and deliver the value to customers. Introducing customer value management

With this new power dynamic favoring the customer, subscription providers must rethink how they manage the customer relationship. With the increased customer focus on perceived value comes the mandate that providers must communicate their value at every interaction with the customer. Unfortunately, CRM solutions fall short of this task because they are built to manage the customer’s value to the vendor—not the other way around! New customer success platforms go a step toward solving this challenge with the concept of success plans—but based on how they are being deployed, they are fundamentally just traditional CRM extended to pass the buck to customer success managers (CSMs), leaving them on an island instead. To fill this gap, B2B subscription companies must evolve their disjointed processes and ad-hoc tools to encompass customer value management: a company-wide, connected strategy focused on the continuous understanding and optimization of the value your customers get from your solutions and services. 40

C hapter

4

R eali z ati o n M o n eti z ati o n The changing customer

The ease of launching new cloud-based applications and the attractiveness of the subscription model to investors have led to an overwhelming launch of new subscription solutions. Customers now have more choice than ever before and more options for how they engage and buy. In the software industry, implementation timelines and costs are constantly being driven down, interfaces and data protocols are being standardized, and userfriendly, intuitive applications that do not require week-long training courses are the norm. These innovations make it easier for businesses to adopt new technologies to gain a competitive advantage, and businesses are in constant search of better solutions: 71% of companies say they are willing to explore new providers for service-based IT solutions already in use in their organizations. However, that doesn’t mean that customer churn is a foregone conclusion. Although switching vendors is cheaper than it used to be, there are costs to the customer to switch. There are also opportunities for vendors to build lasting, mutually profitable relationships with their customers. As a result, customers are shining a spotlight on measurable sustainable value, and enterprise companies are more focused now than ever on enabling customer retention through customer value management. All subscription businesses will need to behave like software businesses

If a B2B software company wants the customer to buy, then they need to discover, communicate, and sell the value of the solution. If they want the customer to keep subscribing, then they need to consistently deliver and communicate value realized. Value selling and value delivery are not new problems, but they have yet to be solved consistently at scale. This problem is driving a change in focus for the industry. Previously, most software companies were ‘hunter–seller’ focused. This focus drove market share and new logos but did nothing to manage and deliver an ongoing customer experience. In an XaaS world with fewer practical barriers to customer churn, companies must proactively cultivate customer contract renewals. All this combines to drive the trend where the CSO hunter–seller sales-led model is being replaced with a CRO/CCO customer-success-led model. As a result, customer retention has become a key spending point for B2B XaaS companies, with 52% of providers increasing customer spending, and a further 54% treating add-on sales and upselling as priorities. The issue is not that customers aren’t getting any value from their XaaS subscriptions. The issue is that without a CVM strategy, service providers are blind to the real health of the account. Problematically, the customer is often blind to the value their subscriptions bring to the table as well. Fortunately, this can be addressed by developing and executing a robust CVM strategy. With a good CVM strategy, service providers gain visibility into which customers are getting value from their solutions, and which customers are not. For this reason, a CVM strategy is not just critical for ARR growth but also critical for ARR forecast accuracy. C hapter

4 

41

M ri n al ( M G ) G urbaxa n i a n d A lex S mith

Additionally, tracking the customer’s priorities in terms of what they see as valuable becomes the foundation for a smart retention and expansion strategy based on what matters most to the customer.

So, what does a successful customer value management strategy look like?

Today’s B2B subscription customer does not tolerate being forced to rearticulate the value they seek to realize from your solution each time they encounter someone from a new team within your company. In order to successfully drive customer satisfaction and renewals, a CVM strategy must be integrated across all customer-facing functions in the company. In particular, it requires the internal teams of sales, delivery and customer success, product, and marketing all to be connected through a common understanding and shared language based on customer value. Each of those teams can then better communicate that value to every customer.

Sales teams need to discover value and sell on outcomes: ■ ■ ■

A personalized value proposition is co-developed connecting the needs of the customer to the differentiated solution being sold. Use the value proposition to build a quantified business impact (KPIs + £$€). Needs, business objectives, and KPIs create version 1 of the ‘living’ customer success plan to kick-start the delivery and realization phase.

Delivery and customer success teams need to ensure that they deliver on value for shared success: ■ ■ ■

A digital record of pains selected, KPIs targeted, and relative importance is delivered in its entirety to customer success, giving them a roadmap from day one. The customer success plan is then jointly managed, becoming a ‘living’ document that drives the continuous delivery and communication of customer value. The solution delivery focuses on business outcomes, not technical requirements.

Product and marketing teams need to continually optimize the proposition based on customer insight: ■ ■ ■

Aggregated digital records of value sold, value delivered, and value realized become a nonstop market insight engine. Product and marketing teams continually improve the proposition (solution features, packaging and pricing, messaging) to align with and optimize customer value. CVM technology enables the proposition to be easily and consistently deployed through all channels to market.

Value management starts with discovery

The customer success team might appear to have the lead on ensuring that the customer adopts, achieves their goals, renews, and expands. However, the reality is that the most 42

C hapter

4

R eali z ati o n M o n eti z ati o n

important link in the customer value chain is the initial interaction to uncover the customer’s pains, needs, and goals: the discovery. Without discovery the deal is less likely to close, and, in addition, the relationship will not have been established on driving realized customer business value. With a relationship built on product features and price, your customer success manager never receives a hand-off and has to create their own roadmap, sometimes without access to the original buyer. However, while discovery is part of every sales methodology, few sales teams do it correctly. Different organizations may call it by different names, but the task of uncovering the customer’s business pains, collaborating on the value of solving those pains, and then connecting that to the solution is essential. Unfortunately, sales teams have been working on a script of pricing and features for so long that pivoting to discovering customer value is an unfamiliar and foreign landscape to many. Many sales teams feel out of their depth, untrained in the value drivers of a specific product for a specific industry, and unsure how to discover the unique combination of pain points for a customer. Not only is it a new area for them but it is also time intensive. However, value selling is critical for B2B success. The newly empowered B2B customer expects a partner who can address how a XaaS product will bring them value. Ninety-two percent of buyers want to hear a value proposition early on in the sales cycle, and 76% of customers state that they expect companies to understand their needs and expectations. To help address this, many larger B2B XaaS companies have invested in the ‘value’ professional: a strategic resource to help the sales team deliver a personalized business case. These business-value engineers/advisors/consultants often parachute into strategic sales cycles with Excel-based, or sometimes online, ROI calculators. The growth of this niche specialist team cannot be overlooked. In the last 6 years, the number of value professionals on LinkedIn has increased 10 times, from 5,000 in 2016 to over 50,000 today. Forward-thinking organizations have SVPs of business value engineering, and a chief value officer has been a rare new addition to the C-suite. However, value professionals are expensive, and people-based solutions scale poorly. To ensure that the value management strategy is executed consistently across the salesforce, you need a technology solution. The next evolution of CRM

Traditional CRM is not well positioned to help B2B XaaS companies shift to a customer-centric, value-driven focus. The primary issue is that CRM platforms were built around tracking the value of the customer to the software company by recording the customer’s purchases and interactions with the vendor. Recently, CRM has expanded its focus to include the lifetime value of the customer and the CSM’s activities. But the modern B2B customer is now demanding a 180-degree revolution: they want a relationship built on the value of the software company to the customer, not the reverse. CRM began from the desire to build a ‘single view of the customer,’ but that view was from the outside-in, from the perspective of the XaaS company. The system collects C hapter

4 

43

M ri n al ( M G ) G urbaxa n i a n d A lex S mith

data on each customer into one location, enabling the internal teams to collaborate through one single source of truth. This is useful data, to be sure, but to sell successfully in today’s enterprise XaaS market, a salesperson needs to know why a customer will buy or why they bought and will continue to buy. Traditional CRM just doesn’t connect the dots. Additionally, traditional CRM platforms are not built for direct collaboration with the customer. Value is in the ‘eye of the beholder.’ It is the customer who decides what is valuable to them, so an accurate representation of the customer’s priorities must be co-authored with them. Although CRM platforms can be deployed to enable collaboration across internal teams and with partners, these platforms were not built to drive interactive realtime collaboration with the customer, and that is a must in today’s customer-centric environment. CRM must evolve to become a platform focused on the value the customer is getting and on digital and virtual collaboration with the customer. Some new customer success platforms take a step toward solving this challenge with the introduction of, for example, success plans. However, current customer success platforms are really just an extension of the CRM model and have the same core limitations. Their focus is typically on extending the CRM’s sales automation, forecasting, and management to the CSM, effectively transferring the problem somewhere new. And, even in situations where there is a strong focus on the customer’s success, as we discussed in the prior section, discovery upfront by sales is the critical first link in the customer value chain. This is not addressed by a customer success platform that is focused on the CSM stage of the journey instead of a company-wide, end-to-end approach. Within the B2B XaaS industry, very few companies are using CRM technology to support the management of what the customer really wants, which is business value and outcomes. Value is no longer something that is an internal metric within a XaaS organization. Customer value must be discovered through the eyes of the customer. The XaaS customer needs to be centered in the conversation, and the value that they perceive will be entirely unique to their needs. To that end, CRM must evolve (see Table 4.1). Value of CVM to a B2B software company

The impact of evolving from CRM to a CVM strategy has massive economic potential for a B2B XaaS company. The benefits to sales teams adopting a consistent value-selling approach have been well documented in the past. These include higher win rates, reduced discounting, larger deal sizes, shorter sales cycles, and the faster onboarding of new salespeople. When sales teams do value discovery consistently as part of the integrated CVM approach, the benefits to the customer success team and the wider organization are, however, even more significant given the very high percentage of annual revenue that comes from existing customers for all modern software companies: Net revenue retention (NRR). Companies with a CVM strategy can achieve 130% NRR, an increase of 26 percentage points relative to the industry median NRR of 104%. 44

C hapter

4

R eali z ati o n M o n eti z ati o n Table 4.1  Traditional

CRM versus CVM comparison

Traditional CRM

CVM

Focuses on the lifetime value of the customers to the company Records details of the customer’s purchases, renewals, and interactions with the vendor Automates sales tasks and improves internal collaboration across teams within the vendor organization Tracks information external to the customer: email and social media communication, support tickets, advocacy events

Focuses on the lifetime value of the software company to the customer Sheds light on why the customer bought, why they use the solution, and why they should renew Automates collaboration with the customer

Tracks information internal to the customer: business pains and goals, the usage and adoption of the solution, their business KPIs and results

Company valuation. Research from SaaS Capital then shows that, due to the compounding benefit, each 1 percentage point increase in NRR correlates to an increase to the B2B SaaS company valuation of +12% over 5 years.

Conclusion

With growing competition within the B2B XaaS market, finding ways to differentiate your organization should be a priority as you focus on attracting a customer base that finds value in your products. Placing customer value at the core of your go-to-market model makes it clear to every potential customer what your solution will bring them and that you are committed to ensuring that they realize that value. In today’s business environment, B2B XaaS companies need to be partners in their customer’s success journey, and that relationship begins with the sales team connecting with your customers on value in the upfront discovery. For most organizations, this will be a sea change in how they manage the customer relationship, and traditional CRM technology is unable to support that change: modern value-based relationships require CVM. The growth of CVM will mimic the growth of CRM. Over the past 30 years, salespeople have evolved their relationship-tracking tools from pen and paper to spreadsheets and then to the CRM. Today, most businesses with value management strategies rely on spreadsheets and basic online ROI calculators to execute them. However, a consistent value management rollout and whole-organization follow-through require a CVM solution. And soon, every serious B2B XaaS company will have one.

Bios

Mrinal (MG) Gurbaxani MG is the CEO and Co-Founder of Cuvama. Over the last 17 years, MG has helped over 80 global B2B customers across manufacturing, distribution, high-tech, and C hapter

4 

45

M ri n al ( M G ) G urbaxa n i a n d A lex S mith

software realize their monetization potential. As the software industry moved to SaaS, MG recognized that the shift of power to the customer was inevitable, and built a boutique consultancy, mgpricing, dedicated to this new and growing market need. After many years of solving the same problem set, the consulting know-how was codified into a tech platform, and Cuvama was born. MG is passionate about game theory and is an active club chess and poker player. You’re likely to find him on a tennis court or golf course on sunny afternoons. He holds a degree in Mathematics from the College of Wooster (USA) and an MBA from INSEAD (France/Singapore). He lives with his wife and two young daughters in London. Alex Smith Alex is CCO and Co-Founder of Cuvama. Alex has worked for 17 years in pre-sales and value selling across a wide range of B2B and B2C industries, including traditional wholesale distribution, financial services, and industrial manufacturing. In addition to pricing, he has been part of and led SaaS sales teams selling multi-million-dollar contracts. Alex’s passion project now is helping fast-growth SaaS companies effectively communicate their value proposition and ‘get’ their value-pricing potential across the land, renew, and expand phases of a SaaS subscription relationship. Many years ago, Alex began life as a physicist getting his master’s at Oxford University—but he doesn’t remember much of that these days. Alex is an avid cyclist and lives in Manchester with his wife and two sons.

Key objectives 1. To understand how the evolution of B2B industries toward a subscription-based model requires companies to learn a brand-new discipline: customer value management. 2. The growth of the subscription model and increased choice is leading to changes in the way customers engage with vendors and make buying decisions. We explain what subscription businesses can learn from successful software companies and how customer value management is at the heart of a successful commercial relationship between vendor and customer. 3. Finally, we cover what a successful customer value management strategy looks like. Outlining some key considerations and how different teams can integrate CVM into their operations.

Key summary points 1. Customer value management (CVM) is an essential capability for B2B companies to master in the 2020s. As buying habits evolve, companies that can deliver and measure long-term customer value will have a substantial competitive advantage. CVM helps vendors rethink customer relationships and ensure that they build a company-wide, connected strategy that drives continuous understanding of how your customers get

46

C hapter

4

R eali z ati o n M o n eti z ati o n

value from your product or service. Once this understanding is in place, the value that customers receive can be optimized. 2. CVM’s importance is becoming more prevalent as reduced switching costs mean that buyers are increasingly willing to try new solutions. Vendors must focus on demonstrating to buyers that their solution helps them achieve key business goals. CVM helps vendors understand what the customer values and builds lasting relationships based on aligning their product with key business goals. 3. A successful CVM strategy should be integrated across all customer-facing functions of the company. Sales, delivery, customer success, product, and marketing should all be connected through common understanding and shared value based on customer value—enabling them to better communicate that value to customers.

Key questions 1. What are the objectives of customer value management and a relevant CVM strategy? 2. Can you achieve excellence in CVM without cross-functional engagement? 3. How is the future of CRM impacted by the emergence of the CVM platform?

C hapter

4 

47

C hapt e r 5

Measure and Quantify the Value of Your Digital Solution Stephan M. Liozu

The net effect of smart, connected products will be different in every industry, but across the board the nature of competition will change. Michael Porter, Harvard Business Review, 2015 Let us say for the sake of argument that you can reduce a customer’s operating costs by 15% while simplifying their user interface and customer interactions so dramatically that their churn loss decreases by 5 percentage points. I can appreciate your team’s thrill and the sense of accomplishment that comes from learning that their new digital offer can provide significant cost savings and simultaneously offer the customer considerable revenue enhancements. Now comes the cold water. It is discouraging if a team then sees its offering rejected by the customer. This can happen if, say, a company outside your traditional set of competitors convinces the client that they can cut their costs by 23% and decrease their churn loss by 3 percentage points. The customer has decided that the significantly greater cost savings outweigh the upside potential from the slightly lower improvement in churn. As powerful as your solution is in absolute terms, it loses on a relative basis. Therefore, I cannot emphasize enough the critical role of competitive intelligence. Dollarizing your digital offer’s differentiation—the focus of this chapter—is essential to success. But it is impossible to achieve this without knowing what your competitors are offering and expressing your differentiation value in dollar terms. One of the most surprising findings of our ongoing research agenda over the years is the extent to which people underestimate or ignore competition because, for most of them, digital is still new. The reality is that many elements of the data value chain and its digital applications have been commoditized over the last three to five years. There is fierce competition in the digital space, and it will only increase with the current rate of innovation and growing M&A activities. From collaborating with executives, conducting expert interviews across industries, and reviewing well over 200 published DOI: 10.4324/9781003226192-7

49

S tepha n M . L i o z u

reports, I learned that underestimating competition has different dimensions. Decisionmakers underestimate the sources, the intensity, and the relevance of competition. The tendency to underestimate competition when developing novel offers is not new. Over the years, I have noticed this frequent problem as one of the main hurdles to leading product and service innovation projects in a broad cross-section of industries. Recently, though, I have seen the need to incorporate competitors’ moves into data monetization and digital innovation projects become even more pressing than before.

Competition redefined!

The growing digital tsunami has given customers many more transparent and available options. But I find that many companies, especially industrial natives, describe their data-enabled and digital offers to customers in absolute terms, as if the competition did not exist or the customer had not yet begun. This is not only surprising but also irrational. The constant message to your team should be that ‘we are not alone,’ because all your competitors are engaging in a digital transformation or at least claiming they are. Nonetheless, teams begin framing up their offer, analyzing their differentiation, and seeking their ‘wow’ factors as if every customer in the world was not scrutinizing the market for alternative offers. I have seen suppliers function as if there were no real competition. The reality is sobering. While some vendors are still in the initial stages of their digital growth journey, many large companies have been developing digital offers and monetizing data through connected devices and predictive maintenance for years; some have accumulated experience and built their learning curve for over a decade now. In fact, when you think about it, concepts of the internet of things or predictive maintenance have been explored and applied for over 20 years now! Who are you competing against when you propose a customer pilot and when you scale your solution? The answers may surprise you. Competition in digitally enabled products and services is not only intense and multifaceted but also structurally different from competition in traditional industries. I illustrate those differences in Figure 5.1. The customer can choose to do nothing or seek alternatives to your digital offer in one of three areas, all of which may appear as a better, next-best alternative. Let us briefly review these options: ■

■

50

Direct competitors. Customers can choose a direct competitor offering software and data, and this type of competition is stratified. It could range from giants such as Siemens, Honeywell, Cisco, Intel, HPE, and Rockwell to very nimble startups, small manufacturers, industrial distributors, and systems integrators. All have digital initiatives. All are searching for ways to apply artificial intelligence to resolve jobs to be done. All are pouring large sums of money into digital and making strategic acquisitions. Indirect competitors. This kind of competition is likewise stratified. They include the major global consulting firms (e.g., Deloitte and BCG) that sell billions of dollars worth of digital transformations, platforms, solutions, and go-to-market services. Indirect competitors also include smaller, niche consulting companies focused on business transformation and change management in the digital space. Indirect competitors also include the large digital business units of other manufacturers C hapter

5

Q ua n tifyi n g D igital Value

Customer “Does Nothing”

Direct Competitors

Manufacturers, Distributors, Suppliers or Start-ups

Competition

Customer Internal Solution Internal Digital Incubator Internal Consulting Group

Indirect Competitors Consultants (BCG Digital), Digital BU of Large Manufacturers (Bosch), GAFAM

Figure 5.1 The extensive nature of digital competition.

■

or conglomerates (e.g., GE Digital and Bosch Digital) that have begun to market their own expertise developed in-house. Finally, you must also carefully observe the moves of digital power players, such as Amazon or Google, as they continue to grow into the B2B space. Customer internal solution. Customers can often legitimately view a digital service as a make-or-buy decision. This turns the customer into a competitor. Companies often decide that they do not want to turn to outside help to get the job done. In the digital space, we increasingly see B2B companies set up their own internal digital incubator, sometimes in conjunction with a consulting group. So, you need to determine your differentiation from potential internal solutions as well. Industry experts concur that approximately 80% of companies today claim to have a digital transformation program underway. While this implies that 20% do not, these potential customers are not necessarily easy targets with limited knowledge of the competition. Rest assured that even these customers seek knowledge about alternative suppliers and available options, assess their own internal capabilities, and are at least in the process of framing their decisions.

Complicating this further is the convergence already underway. I fully expect many current indirect competitors to become direct competitors within the next five to ten years. All they need is hardware, either through partnership, joint venture, or acquisition, and they will be on their way thanks to their existing speed and scale. At one point, for example, Deloitte Digital offered turnkey IoT solutions within 90 days. How could Deloitte achieve results in such a brief time? Together with deep customer focus, the company has a large and deep pool of talent, it has scale, and it forges critical partnerships and coalitions to secure needed strategic resources and capabilities. This complexity and speed in digital competition mean that when you craft your value proposition and develop your dollarized value models, you will need up to four value propositions and customer value models, each with its own reference values (which is what you would compare your solutions against). You may have to understand and dollarize your differentiation versus (1) direct competitors, (2) consulting C hapter

5 

51

S tepha n M . L i o z u

companies, (3) a customer’s internal benchmark, or (4) in some cases, a generic ‘do nothing’ option. Let us complicate things even more! Let us take the example of predictive maintenance, which is not new and is growing at a steady rate across most industrial verticals. A vendor of digital solutions might face a ‘do nothing’ situation where a customer is not very sophisticated and digitally mature. The reality is that there are various levels of ‘do nothing.’ In that area, it is rare to find industrial accounts doing nothing to optimize maintenance costs. They might have something in place. So, you go from ‘do nothing’ to ‘might do something,’ as shown in Figure 5.2. The vendor’s marketing and sales team must identify and qualify the account and understand what the customer is currently doing. Competitors might certainly do this analysis. The list of differentiated value drivers and the quantification of value will vary based on what prospects are doing, are partially doing, or are not doing at all. They must also anticipate what competitors might offer in each situation.

What is your basis for true differentiation?

I have worked over many years with companies on understanding, quantifying, documenting, and communicating the superior value their offerings create for (and with) customers in business markets. Among the different value-based marketing and sales tools used, I refer here to the approaches and methods described in my 2016 book Dollarizing Differentiation Value. Among many commonly used tools, let us refer here to the well-known VRIO model, as it helps a firm to understand whether your digital offer can rely on one or more true point(s) of differentiation. The letters in the acronym VRIO, as we show in Figure 5.2, stand for valuable, rare, imitate, and organization. The essence of this process, and the steps that follow, is identifying your differentiators and then focusing on those that are valuable to you and your customers, which are rare and hard to imitate, and that you will be good at executing. Having a potential, meaningful competitive advantage boils down to how you answer the questions shown in Figure 5.3. The most common and deceiving situation I see is the company that develops something with an indisputably high value in

Vs. “Do Nothing”

Vs. Existing Preventative Maintenance Programs

Competition Vs. Existing Routinebased Predictive Maintenance Programs

Vs. IoT-based Predictive Maintenance Programs

Figure 5.2 The various levels of ‘do nothing.’

52

C hapter

5

Q ua n tifyi n g D igital Value Data Competitive Advantage Using The VRIO Model Is it Valuable to Customer in Dollars & Cents? Is it Rare to Find or Access? Is it costly to Imitate? Is the firm Organized to exploit it?

Figure 5.3 The VRIO model.

absolute terms. But, if that kind of solution is common, easy to imitate, and/or hard for the organization to scale and sustain, that immense value is not the soundest basis for a competitive advantage. When you are scrutinizing competition to determine your relative position, the same questions and criteria apply. For all the talk and bluster you may read or hear about digital ‘solutions,’ many competitors fall short because their organizations cannot make the offering sustainable. The VRIO model offers an understanding at the macro level, but the real understanding and action take place at the micro level. There are potentially hundreds of various aspects that a company could leverage to differentiate itself and create a competitive advantage. In each case, you would need to find out whether direct or indirect competitors or the customer is better or worse than you. To provide some structure and to facilitate that thought process, I group the potential areas of advantage into five categories: data-based differentiators, analytics-based differentiators, business model-based differentiators, value constellation differentiators, and pricing-based differentiators. ■

■

■

Data-based differentiators. A first set of differentiators refers to the nature of data itself. These differentiators may be based on your ability to access, store, clean, and aggregate strategic data over time, including your data, the customer’s, and data from external sources. Differentiation may flow from your ability to unite these different data sources under one roof. This explains why IBM acquired the Weather Company (former parent of The Weather Channel) and its extensive access to realtime weather data around the world. In a pilot, for example, the more data you have about an asset and its context, the better your scientific models can be. Then there is the question of integration: how well can you translate and incorporate data into one cohesive set? Analytics-based differentiators. These involve the applied research and science, the algorithms employed, and the mathematics leveraged for analyzing data. You may have deeper research into AI, more advanced applied science, better algorithms, and more creative predictive models than other companies. You might be better at identifying trends and patterns in data and therefore at training your machinelearning algorithms versus those of competitors. The faster it is trained, the faster you create value. Business model differentiators. All key building blocks of a digital business model can become the basis for differentiation. Successful firms often focus on a more attractive value proposition than their competition, such as by promising usagebased models, or outcome-based services that others cannot commit to. Companies also effectively differentiate themselves by focusing on specific customer groups

C hapter

5 

53

S tepha n M . L i o z u

that they serve better than others. Furthermore, firms combine resources and activities in smarter ways than peers, thus decreasing costs and improving the efficiency of operations. The next two differentiators can be part of your business model differentiators. Because of their widespread ramifications, you can also consider them a source of differentiation. ■

■

Value constellation differentiators. Developing and managing the right digital platform depends on partnering with the right players. A potential customer will assess you by the ‘friends’ you spend time with. Are you in the tight orbit of a Cisco, Microsoft, IBM, or Amazon, for example? Customers will also evaluate your access to other components and intelligent assets. Finally, they will judge you on your role in these partner networks. Are you a proven leader who can bring other, disparate suppliers together to solve a problem? This is another reason being first to market is advantageous. Pricing-based differentiators. Your pricing models may become a powerful source of differentiation. Choices in pricing, financing, and payment models, as well as service agreements, may set you apart from the competition. Does your company find new ways to share the value pool with customers? Do you innovate in gainsharing agreements with customers? What is a win-win for you, your customer, and potentially your partners?

I have listed only a few avenues for differentiation. The goal is to find the small set of unique points of difference for which your digital offer is truly and sustainably superior to your competitors (direct, indirect, and internal) in the spirit of the VRIO model. Most digital factories and incubators lack the skills, time, and resources needed to conduct deep competitive analysis. It is frightening to see companies pour millions into technology without having a process and budget to conduct deep competitive benchmarking to discover which elements will thoroughly differentiate them from direct or indirect competitors. If you are about to launch an IoT solution for example, how do you know if what you are doing is better than what your direct competitors are doing? Recall the first paragraph of this chapter. You might be able to claim that your dataenabled solution will save customers 15% in operational efficiencies. But you must know if your competitors are claiming that they can do a better job because of their unique design, relationships, and access to data. They have already committed to 18% in operational efficiency savings. They began publishing value case studies showing superior monetary impact. Choosing and using dollarization techniques

Dollarization is about translating your competitive advantage into financial benefits for customers. Approaches and techniques focused on value-based marketing and sales, especially in B2B market contexts, proliferated in the early 2000s.1 During the same time, professional buyers increasingly emphasized TCO in sourcing decisions rather than focusing discussions solely on price. Translating the unique points of difference of your digital offerings into financial benefits for your customers is a key to 54

C hapter

5

Q ua n tifyi n g D igital Value Dollarization Techniques TCO: Total Cost of Ownership TBO: Total Benefit of Ownership EVE® or CVM® Models LCC: Life Cycle Costing Analysis LTV: Customer Lifetime Value Analysis ROI: Customer ROI Calculation

Figure 5.4  Dollarization techniques.

commercializing your data resources and skills. My Dollarizing Differentiation Value likewise delves deeply into various dollarization techniques. Figure 5.4 enumerates some of the most common and powerful approaches, some of which we will illustrate (i.e., economic value estimation, or EVE) in greater depth. I highly recommend that you do your own research on the one that seems most applicable to your situation. Most digital providers aim at providing some type of business case including basic ROI calculations. This is predominant in the SaaS world for example. TCO has a long history and is common in outsourcing decisions. TBO is a newer approach that captures the revenue-generating benefits of a solution, not only the cost-savings aspect. The challenge these models pose is to set a sufficiently long-time horizon for your digital offering because under some models you might lose money or barely break even as a supplier over the short term. Most of your upside is in the medium or long term. The objective of EVE is to begin with a reference value and then add and subtract the values of your significant differentiators until you arrive at the net differentiation value your solution generates. I also refer to this as the value pool. This model stands out because it includes the negative aspects as well. Part of being honest and transparent with customers is acknowledging where your solution may have drawbacks or deficits relative to alternatives, even though your overall solution is superior on an aggregated basis. This laddering process is shown in Figure 5.5. Costs unique to doing business with you Positive Differentiation

The Unique Value You Provide the Customer

Reference Value

Figure 5.5 An EVE example. C hapter

5 

55

S tepha n M . L i o z u

EVE has an elegant simplicity, but it also has an additional advantage. As a straightforward and easy-to-understand software tool, it allows you to ensure a common system when you are evaluating many MVPs. When you have 75 or 100 or even more MVPs in testing, you cannot afford to let each team adopt its own model or its own way of applying a common model. This will lead to confusion and make decisions across offers difficult, if not impossible, because of the diminished comparability when it is time to allocate resources and select which MVP will go to the next step. Even if you develop your own model, putting it on an automated platform is essential. My experience with the framing process suggests that, just as you would choose a specific business model framework for your framing, you should also select one or two dollarization techniques to include in the framing process. By doing so, you will also industrialize your customer value modeling process and make it easy for MVP teams to learn one or two techniques, maximum.

Difficulties in dollarizing data-driven and digital offers

As straightforward as some methods (e.g., TCO, TBO) or tools (e.g., EVE) may be, I do not want to imply that dollarization itself is straightforward. There are many unknown variables in the digital world, and we have not come across resounding success stories. I also see many companies doing great customer value success stories, but they remain anecdotal. They have a challenging time generalizing value stories to a specific segment or vertical. So, you might encounter many difficulties in this process: ■

■

■

■

56

Lack of access to customer application data. Regardless of the level of your customer intimacy, some of your digital opportunities might require access to operational data from different stakeholders. If you cannot identify or access a data source, it will be hard to derive KPIs and determine their value. Lack of deep knowledge of competitors’ digital offerings. You may know your competitors’ traditional products and services in detail, but how well do you know what they are ‘cooking’ in their digital incubators? Beyond what you can find publicly, even the richest and most reliable competitive intelligence may fail to give you a complete picture. The same applies when you are doing competitive benchmarking of software solutions. Unless you have access to demos or the software itself, you really do not know what is under the hood. Your technical solution is not fully conceptualized yet. Digital pipeline development means that MVPs are in constant iteration. The more disruptive they are, the harder it can be to compare performance with that of competitors’ offers, and the harder it may be to see precisely how your concept drives value in day-to-day reality. The same goes for your competitors. They may jump the gun and make claims about their differentiators. But it may all be a game to gain time and plant their flag. Lack of test methods to evaluate performance in application. Some of the features and benefits of your digital opportunities and data might be so new that you cannot find an appropriate model for testing. Your team may need to develop entirely new methods and metrics for measuring and extracting value. That may be the trickiest situation. C hapter

5

Q ua n tifyi n g D igital Value ■

MVPs might be complex systems or solutions. Innovative digital offers frequently integrate multiple components from complex hardware and software elements to value-added services. Measuring the combined value of such complex integrated offers can be a challenge.

I highlight these difficulties to preface the next step: value modeling. This is where words and concepts and high-level estimates are transformed into numbers, such as ratios, fractions, or percentage differences. These are then expressed in terms of hard money. This is another topic covered in greater detail in Dollarizing Differentiation Value, so I summarize it in the next section.

Value modeling: Expressing relative value in money terms

This next step helps you calculate dollarized value in an extensive, elaborate way. The sequence is to walk your customers through the differentiators and the benefits, highlight the most critical and compelling benefits, turn them into ratios, and then express them in monetary terms. The process is inherently mathematical, but it is an art as well as a science. Figure 5.6 lists the steps. The science lies in the mathematics and the mechanism for processing and calculating the value, formulas, and data. The art lies in how you prioritize your customer benefits, plus the nature of the numbers and the stories you tell around them. You can express your benefit as a ratio or a fraction or devise a number that optimally blends art and science in a way that convinces the customer. This is not necessarily a skill that a team will possess, so accomplishing this requires training and practice. The process of dollarizing an MVP is difficult and requires effort and creative thinking. Another challenge is finding relevant reference points. If what you are doing is new, what are the credible and established references against which to measure yourself? If you do not have a customer-specific reference available, fortunately you can turn to other sources for credible guidance: ■

Published digital case studies and industry reports. They often provide credible references on which to base your calculations. There are many reports out there on the public web! Because of previous independent validation or usage, these numbers are often better than theoretical calculations, no matter how logical or well-derived the latter may be. I also suggest that you remain on the lookout for additional nuggets such as consulting reports, technical reports, case studies, YouTube videos, and so forth. Approach to Customer Value Modeling 1) Translate differentiators into customer benefits (using customer vocabulary, thinking process, mental frames).

2) Prioritize customer benefits based on segment (most compelling hook and last impression).

3) Turn these benefits into compelling facts & ratios (% faster, increase of 2 pts of yield, 1/3 more durability)

4) Dollarize these benefits into $, £, ¥ or Euros

(dramatic numbers, round numbers, compelling savings/gains).

5) Create your value story or value script.

Figure 5.6 The sequence of steps in customer value modeling. C hapter

5 

57

S tepha n M . L i o z u ■

■

■

■

Theoretical calculations based on technical assumptions. Your technical and R&D teams might have test data or theoretical models, based on traditional customer operations. In the absence of independent external sources, these numbers can still provide a solid reference basis. Academic research in the end-use application. To generate some empirical data, you can collaborate with a university research center and sponsor an academic study generating data, providing results produced by a third party. You can also collaborate with a PhD student who can set up experimental research and run trials with customers. Many companies choose this alternative for the areas of smart seeds and agricultural IoT. Whether you sponsor them or not, published academic papers can be used as credible sources of data. Do not forget that part of the responsibilities of PhD candidates is to present their research work at conferences. Make sure you attend their presentations! Comparisons from other similar contexts. In some cases, the WOW differentiation of your MVP may be entirely new in one context but similar to solutions companies have provided in other areas. You might have to scan outside your narrow industry for such analogs or to draw correlations. Data from a connected-car environment, for example, may apply to solutions in a rail transportation environment. You may sometimes need to make a leap of faith, but there is still an underlying logic. Make assumptions to be evaluated with customers. When nothing else works, you must work collaboratively with very friendly customers who have the same dilemma and who would welcome a joint solution. Develop your assumptions but involve customers in checking their robustness. The two main downsides are that this process can take time and may be subject to strict confidentiality, which limits its applicability to other customers.

What are the potential financial benefits of investing in digital initiatives? When culling data from a broad cross-section of consulting reports, industry analyses, and white papers, I find general agreement among diverse experts that the upsides in comparison with ‘doing nothing’ are substantial. For example, most consulting reports show a 50% reduction in unplanned downtime costs when investing in predictive maintenance solutions. That has become the norm for savings when a customer ‘does nothing.’ If you are working on an MVP and you do not know the reference value basis or a potential value difference to target and beat, you can use these credible sources as a first hook to convince your customers that the differentiation value in your solution can be substantial. Let us say that a digitally savvy customer who knows you from your traditional business is leaning toward conducting a pilot with you but also has a do nothing option on the table. The financial savings numbers should be significant enough to convince them to pursue collaboration. Then your task is to convince the customer to collaborate with you and not with a direct or indirect competitor. Keep in mind that everything—from the savings generated by an offer to the advantage of faster time to market—can be dollarized. These are the practical details of dollarization. The must-do tasks for dollarization

Throughout this chapter I have referred to data points, assumptions, and the need to evaluate and validate. This puts the onus on you to be as meticulous as possible in 58 

C hapter

5

Q ua n tifyi n g D igital Value

recording and documenting every step of the process. This begins a mix of the ‘art’ and ‘science’ of dollarization, as you put your solution or your MVP into context for the customer. Defining the context is analogous to mapping your value constellation. You need to have an overview of all touchpoints, all relevant sources and types of data, and all interdependencies. The best way to do this is to bring a team together across different disciplines and functions, from AI to data analysis to marketing research, sales, and customer support. Include competitive intelligence experts in your team. We cannot understate that last point. Someone needs to study all relevant competitors (direct, indirect, and internal; current and future) in the digital space to know exactly what they are promising or what they are capable of. You must stay focused on the simplest question that matters: what makes you better than someone else at solving the job to be done? If your best answer is the size of your team or the number of your pilots, you are on the wrong path. If your answer is currently only in absolute terms, you need to put it in a relative context for it to be meaningful. Differentiation is always relative. During the dollarization process you need to list all your assumptions, document all your calculations, and document all your data sources. Documentation and validation are key as you iterate until you have defined your value pool, the aggregate value of your dollarized differentiation. Once you have an initial number that you can support and document with confidence, that number goes to a process owner, who continues iterating and validating. The process owner is also responsible for ensuring that you always have a single current version of your model to avoid confusion or multiple parallel tracks. The ideal situation for your innovation is to have it listed on your company website with strong, clearly identified differentiators and a relevant ROI calculation versus different situations. Then your data sheets and success stories are fully quantified versus the competition. You are also able in the best-case scenario to let the prospect build their own solution and do their own value calculation using a professional web-based ROI calculator. That is what the best-in-class companies do.

Bio

Stephan M. Liozu (www​.stephanliozu​.com) is the Founder of Value Innoruption Advisors, a consulting boutique specializing in value-based pricing, industrial pricing, and digital and subscription-based pricing. Stephan holds a PhD in Management from Case Western Reserve University (2013), an MS in Innovation Management from Toulouse School of Management (2005), and an MBA in Marketing from Cleveland State University (1991). He has authored seven books: The Industrial Subscription Economy (2022), B2G Pricing (2020), Monetizing Data (2018), Value Mindset (2017), Dollarizing Differentiation Value (2016), The Pricing Journey (2015), and Pricing and Human Capital (2015). He has also co-edited five books: Pricing: The New CEO Imperative (2021), Pricing Implementation (2019), Pricing and the Salesforce (2015), The ROI of Pricing (2014), and Innovation in Pricing: Contemporary Theories and Best Practices (2012 and 2017). Stephan sits on the Advisory Board of the Professional Pricing Society.

C hapter

5 

59

S tepha n M . L i o z u

Key objectives 1. Learn about the need to deep dive into the differentiation of your digital offer. 2. Discover the VRIO model and how to extract your true differentiators. 3. Understand the various situations you might be facing with discovering your customer needs and who your competition might be. 4. Learn about the dollarization process for digital innovations.

Key summary points 1. The competition in digital is fierce. Companies embarking on digital transformations must have capabilities in the areas of competitive analysis, differentiation analysis, and customer value quantification. 2. There are four competitive situations facing digital providers in the marketplace: do nothing, direct competitors, indirect competitors, and customer internal solution. The analysis of differentiation must be conducted against these four situations. 3. The quantification of customer value is complicated in the digital world because of the intangibility of the offers, the lack of available and published references, the potential lack of cooperation from prospects and customers, and the lack of testing methods. There are, however, companies doing this very well after many years of testing and iterating. 4. The outcome of the analysis of differentiation and the analysis of economic value is the development of robust ROI calculations showing the benefits of digital solutions versus different competitive situations.

Key questions 1. What might be examples of true differentiations in digital platforms? 2. How do you find out more about what your competitors are working on and are testing in the marketplace? 3. What are the key success factors in conducting economic value estimation (EVE)? 4. How do you get better at obtaining customer data to evaluate the monetary impact of your digital solution?

Note 1 See, for example, James Anderson, Nirmala Kumar, and James Narus, Value Merchants: Demonstrating and Documenting Superior Value in Business Markets (Boston, MA: Harvard Business School Press, 2007).

60

C hapter

5

SECTION 2

Software and SubscriptionBased Pricing



C hapt e r 6

Price Increase for Discounted Customers in SaaS Pricing Research Description and Success Story Maciej Wilczyński

Understanding value creation versus price realization

There are hundreds of concepts and definitions in price management, to name a few: pocket and contribution margins, discounting policies, price signaling, monetization, dollarization, top-line growth, behavioral nudges, anchoring, BATNA, and so forth. However, of all these things, what matters is how you create value and how you get paid for it. In Hermann Simon’s words: ‘Price is what you pay. Value is what you get.’ Understanding what your product/service does for the customer, how it benefits the overall client needs and goals, and how it’s utilized are primary questions any company needs to answer. To better illustrate the concept, we can use the value creation versus price realization waterfall chart in Figure 6.1. Even the worst product creates a certain amount of quantifiable value for the customer (value created). It is influenced by the relationship with the client, your brand, product quality, and overall contribution to the client’s success. This is the starting point from which the value waterfall starts to decrease: the subjective perception of the value created by the company differs from the client’s perspective (perceived value). The unseen value is effectively where marketing and communication efforts fail. Another step is when customers decide how much money they can pay for the value (willingness to pay—WTP). Understanding this is a part of the vast majority of pricing projects and initiatives. As a pricing principle, clients’ WTP should get as close to perceived value as possible. Otherwise, it erodes the value created by our product. The next step in the waterfall is where companies fail to set the price to match the WTP. One of the most common pricing pitfalls is setting the price too low versus what customers can or want to pay (target price). This step is heavily researched through data and analytics by pricing managers, and with the proper knowledge and guidance, it’s relatively easy to optimize.

DOI: 10.4324/9781003226192-9

63

64

The value the customer believes your product/ service creates

Determined by how well you communicate the value delivered

The value your product/ service actually creates for a customer

Determined by the quality of product, relationship, etc.

Influenced by perceived value, ROI expectation, comparables, budget etc.,

The amount the customer would pay for the perceived value

Willingness-To-Pay

Figure 6.1 Value creation versus price realization waterfall, with description.

Perceived Value

Price setting failure Transactional discounting

Relies on understanding of willingness-to-pay, and granular price differentiation

The customer-specific price you set for your product/ service

Target Price

Value Creation vs. Price Realization Waterfall

Price: value erosion

Value Created

Unseen value

Dependent on price execution of sales force

The amount you actually get paid for your product/ service

Realized Price

M ACI E J W I LCZ Y Ń SK I

C hapter

6

P rice I n crease f o r D isco u n ted C ust o mers i n S aa S

Last but not least, price realization of a previously set price is a usual quick win in any pricing initiative. Transactional discounts, negotiations, promotions, and price execution of a salesforce create a substantial optimization opportunity for most companies (realized price). Price realization is the function that generates massive margin leakage. Optimizing this area should be one of the essential priorities of a person responsible for revenue management in the organization. Price realization is part of the value versus price waterfall, which is covered in the case study.

Case study introduction

Brand24 is a publicly traded company offering real-time social media monitoring and analytics. It is designed to keep track of online conversations about a brand and its products. Brand24 collects real-time social data from millions of sources around the web. Their web-based dashboard provides actionable customer insights, email alerts, influencer analysis, automated and customized PDF reports, infographics, and many more features. The tool allows for measuring critical metrics around buzz and sentiment. Over 30,000 brands widely use its international version worldwide to monitor their PR and marketing efforts. The company serves a few most prominent customer logos globally, including H&M, IKEA, Intel, Carlsberg, Discovery, Vichy, and Leroy Merlin; however, their main client focus is small and medium businesses that generate the majority of revenue. The company operates in a cloud SaaS model, in which clients choose the plan and purchase a license on an annual or monthly basis. Such a model is the next step of the software pricing revolution. From 1980 until the early 2000s, ‘per user’ license keys were the most popular, focusing heavily on headcount billing. An excellent example of this model was the Microsoft Office licensing scheme. While quite transparent and practical from a client perspective, it didn’t fully unleash the value for the customer. The model has evolved with the cloud revolution, become more complicated, and relied on more usage-based approaches. SaaS companies, starting from Salesforce, revolutionized how we work and pay for software products. The subscription model provided a predictable revenue (monthly recurring revenue) and created challenges we didn’t know of before. For instance, one of the critical problems in SaaS businesses nowadays is churn rate, also known as attrition or customer churn. It’s the rate at which customers stop doing business with an entity. Effectively, the company needs to overcome the revenue lost with the new business or active sales toward an engaged customer base. Such a situation requires an effective marketing and sales organization that possesses dynamic capabilities to deliver value to the market. Also, increasing software development costs due to salary increases and inflation, a more commoditized product market, and new player entrants created pricing pressure on a company board of directors. The prices didn’t increase for a while, and when they did, the optimization didn’t cover the current customer base (so-called grandfathering of old customers). On top of that, management hypothesized that the current pricing scheme neither captured nor signaled the value to the distinct buyer personas, including marketing managers, PR companies and agencies, data analysts, and content creators. Initially, the product communicated three plans, differentiated primarily with the

C hapter

6 

65

M ACI E J W I LCZ Y Ń SK I

monitored keywords and mentions of volume; some feature differences were minor for the overall picture considering total price differentiation. Knowing it all, Brand24 decided to organize a value-based optimization initiative including a price increase and re-engineering a current pricing model. To do so, the company assembled an internal team and hired an external consulting firm to assist in the process.

Revenue engine diagnostics

One of the first steps of the pricing engagement is to run a diagnostic of internal data. It’s important to mention that Brand24 invested a substantial amount of money and effort in building its internal data warehouse and client repository, so the firm was prepared to run sophisticated product analyses on the data. In this case, the company compiled a repository of all the studies done in the field previously and secured the platform access. It included a go-to-market strategy, internal analytics reports, costs, usage of various customers, and segmentation exercises. All of these were essential assets in the pricing optimization effort. They have increased the organizational confidence that the change is possible, and that the product team has all the resources and capabilities necessary to deliver the outcome. After the internal materials analysis, the project team assembled initial hypotheses, for example, misalignment of key billing metrics with customers’ business outcomes, unclear discounting policy, and potential to increase prices for existing customers. With this in mind, further diagnostics continued. This time it was supported with more advanced data warehouse analytics. Among many things, it included the calculation of features usage by different customers; segmentation of listing prices versus what the clients pay; and general unit economics of the company, for example, which customer segments are close to each other and which grow faster than the others. The revenue engine diagnostics discovered that over one-third of company customers paid below the catalog price. This situation resulted from a relatively liberal discounting policy, which effectively retained customers who wanted to leave the company. Despite securing the account volume, it heavily decreased the two most important KPIs: the monthly recurring revenue (MRR) and average revenue per user (ARPU). The discounting exercise allowed the team to estimate initial leakage sealing impact: 10 to 15% of the MRR would be retrieved if prices were higher for existing customers.

Impact modeling and risk reduction

Afterward, the team modeled various impact scenarios. Two initial assumptions were used: (a) the price elasticity for software products is relatively inelastic, and a 1% price increase won’t result in 1% of customers lost (later proved in primary research as well); and (b) the delta between the listing and realized price is the probability of losing a client if the account price increases. For example, if a client has a 10% discount, their overall churn probability is the same; if it’s higher than 100%, the team is sure the client will leave. The team has applied strict and conservative modeling to reduce the risk of decision failure. Four scenarios were developed:

66

C hapter

6

P rice I n crease f o r D isco u n ted C ust o mers i n S aa S

1. Baseline expected value (EV). Churn probability equals discount size; the company doesn’t lose any clients due to price increase; the overall MRR increases by 22%. 2. Moderately negative (EV+10%). Churn probability equals discount size + 10%; the company loses 14% of discounted customer base; the overall MRR increases by 13%. 3. Negative (EV+20%). Churn probability equals discount size +20%; the company loses 42% of discounted customer base; the overall MRR increases by 5%. 4. Highly negative (EV+30%). Churn probability equals discount size +30%; the company loses all discounted customers; the overall MRR decreases by 4%. As seen in the scenarios above, the company decided to pursue a substantial amount of ‘money left on the table.’ Many of the discounted customers were satisfied product users who had substantially higher WTP and budgets to cover the increased cost. However, no active margin-leakage action was taken before—primarily due to different growth-oriented priorities, not so uncommon in startup businesses. On top of that, when customer management and IT infrastructure costs were taken under consideration, the project team realized that some accounts realized a negative 1,758% profit margin versus the listed price (e.g., legacy accounts with $1 payments per month). While such heavily discounted customers were not a sizable volume businesswise, Brand24 could achieve a positive project ROI by canceling these accounts, not to mention the impact of the potential price increase. The team felt prepared to move toward the implementation stage with the data analytics results. While the theoretical models justified the price increase cause, other factors had to be considered, for example, billing periods, client status, relationships with the account, or human factors, that analytical models can’t capture.

Marking leakage implementation and project ROI

During project setup, an international product of Brand24 had ~$220,000 monthly recurring revenue. The revenue number was a baseline for any further impact calculation. This section covers how the customer support team prepared for the price increase. Diagnostics discovered a few hundred discounted accounts. These customers were long-listed and taken under profound observation by the customer success team to check for qualitative and quantitative information in the customer relations management platform. Among the many things analyzed, a few were most important: why the discount was given, the specifics of these users, how long they are subscribing to the product, what the platform usage is, whether there are any quality or relations concerns, and so forth. In the end, 321 accounts qualified for the price increase. The team minimized the overall project risk and began margin-leakage sealing from the most discounted accounts. These accounts were perfect for the pilot: if they were to accept the price increase, they would provide a substantial profit uplift, but if they were lost, the overall value to the company was not critical. Among 201 ‘high risk’ accounts, only 6% churned, and the majority accepted substantial price increases, sometimes from $9 to $99.

C hapter

6 

67

M ACI E J W I LCZ Y Ń SK I

Afterward, the team increased other accounts, including annual deals. The whole process lasted for over three months. After this period, it was possible to calculate the first realized impact. It’s important to mention that in a subscription SaaS business, the overall client value increases over time, so the price increase payback is asymmetric—the longer the client stays in the product, the more money the account brings in the long run. Of 321 accounts, 72 churned within three months, resulting in a ~5% average monthly churn rate, which is on par with the general organic attrition rate in the company. Effectively, it proves the hypothesis on price inelasticity. The increase didn’t affect the churn decision. The overall MRR increased by $4,000. Furthermore, average revenue per user for discounted accounts, one of the vital North Star metrics in the company, increased from $44 to $67. With this fast, organized, and highly focused team effort, Brand24 increased its revenue and received 559% annualized ROI for this part of the project. It’s important to mention that the company also changed its overall pricing for new customers, added a new enterprise plan, and substantially improved its price signaling to better match the best-in-class practices. However, these aspects were not covered in this chapter. The price optimization results were widely communicated to investors and the stock market. In its 2021 Q4 report, the company stated that there were three main success factors in its record-breaking financial results, all price-related: (1) a pricing scheme change for new clients; (2) revision of discounts and sealing of margin leakages (covered in this case study); and (3) Black Friday promotions. Overall, quarter-to-quarter ARPU increased by 4%, and MRR for the whole company increased by 6%. In annualized results, MRR increased by 21% year-on-year. The consequences impacted the company’s market capitalization. On the day of the announcement, the company’s stock price rose to 26.50 PLN (~$6.5). Compared with the Q3 results with a closing price of 21.00 PLN (~$5.15) on the announcement day, the company achieved a solid 26% return on equity.

Conclusions

Pricing initiatives generate the fastest and highest return on investment for revenue optimization efforts. They require solid knowledge and organizational confidence, for example, to raise prices for the existing customer base or to renegotiate deals that are not profitable. However, the example of Brand24 proves that proper managerial guidance and leadership, organized efforts, advanced analytics, and team collaboration can create a tangible impact for the company’s top-line financials and bring value to shareholders.

Bio

Maciej Wilczynski is a pricing expert and partner at Valueships, a consultancy boutique specializing in software, cloud, subscription, and digital businesses. They primarily advise B2B companies on pricing, revenue models, monetization, commercial strategy, analytics, portfolio management, retention, customer experience, and sales excellence. He gained his consulting experience at McKinsey & Company, where he led EMEA,

68

C hapter

6

P rice I n crease f o r D isco u n ted C ust o mers i n S aa S

part of Agile Insights solutions focused on digital marketing and customer insights. He is a lecturer and faculty member at the Wroclaw University of Economics, where he holds a PhD title in strategic management. He wrote his thesis on software-as-a-service companies’ pricing capabilities. He is the author of scientific and business publications.

Key objectives 1. How can a SaaS company transform its pricing with data analytics and primary research? 2. What are the critical steps necessary to ensure seamless pricing optimization in subscription models? 3. What is the impact and return on investment from value-based pricing efforts?

Key summary points 1. A value creation versus price realization waterfall helps plan and prioritize pricing efforts. 2. Margin-leakage sealing from a revision of discounted accounts provides the fastest and easiest ROI in terms of pricing efforts. 3. Product analytics and data repositories management are critical dynamic capabilities of modern software organizations; used right, they provide tremendous value in pricing optimization efforts. 4. Impact scenario creation and war-room analyses allow for setting the organization’s ambition and help with the risk mitigation of a potential price increase failure. 5. Pricing efforts strongly influence the company’s financial results, alongside their stock market performance and shareholders’ return on investment.

Key questions 1. Why was product analytics influential in Brand24’s overall pricing optimization success? 2. What is the reason for differences in initial impact calculation versus the realized one (over $40k MRR vs. $4k MRR)? 3. Use the Brand24 case study to outline three to five critical success factors important in the price realization project. Why do you think they’re essential?

C hapter

6 

69

C hapt e r 7

SaaS Pricing From Subscriptions to Usage-Based Pricing Models Andreas Hinterhuber

Pricing: An important, yet neglected activity

Pricing is a key driver of superior performance. Small changes in net selling prices have an immediate and substantial effect on profitability. Managers typically underestimate the role of price as a key driver of superior profits. Pricing is also, surprisingly, neglected: in its annual survey of chief marketing officers (CMOs), the American Marketing Association polls marketers on activities where marketing leads (Moorman, 2020). The results of the 2020 survey are depicted in Figure 7.1. Marketing typically does not lead pricing: in only 20% of companies is marketing responsible for pricing. In 80% of companies, either the responsibility for pricing is dispersed—between sales, finance, controlling, and marketing, with no clear accountabilities—or other functional units have responsibility for this function. A recent study (Liozu, 2019) tends to lend support to the former explanation: in the vast majority of companies, there is no centralized, formalized, specialized organizational unit with responsibility for pricing. In most companies, pricing falls between the cracks: everyone is responsible for pricing, which means that, in the end, nobody is. But managing prices and managing pricing is necessary—and beneficial for firm performance. A recent academic study finds that the formalization of pricing—measured via the existence of a pricing process, cross-functional collaboration, and the use of pricing tools—is positively correlated with relative firm performance (Liozu, 2021). Key tasks in pricing are price setting and price getting (Hinterhuber & Liozu, 2012). The combination of these tasks leads to the pricing capability grid (Hinterhuber & Liozu, 2012). Price setting refers to the different approaches that companies use to determine selling prices: cost-based pricing, competition-based pricing, and customer value-based pricing. Price getting refers to different abilities to actually obtain the price set in the first place: some companies are very good at realizing their list prices, via, for example, value communication, customer value quantification, or price controlling. Others are DOI: 10.4324/9781003226192-10

71

A n dreas H i n terhuber

Figure 7.1 Domains of influence of the marketing function (Moorman, 2020).

less effective, and prices thus erode as a result of poor negotiation, poor value communication, or weak price-realization capabilities. Both dimensions thus need to be examined.

Price-setting approaches in the software industry

Across all studies, less than 20% of companies practice value-based pricing, and the predominant approach to pricing is competition-based pricing (44%), followed by costbased pricing (Hinterhuber, 2008). The adoption of value-based pricing is higher in the software industry, where data availability is large and where products are specifically designed to improve quantified, customer-specific outcomes: a recent analysis of over 2,200 companies in the SaaS industry (Poyar, 2021) reveals the following: most companies set prices based on competition, gut feeling, or costs, whereas just 39% set prices based on customer value (see Figure 7.2). The SaaS industry thus leads other industries in the adoption of valuebased pricing, but there is surely a need to further improve price-setting capabilities. Steven Sinofsky (2014), a partner of Andreessen Horowitz, a venture capital company, comments: Nothing is more critical to a software-as-a-service (SaaS) business than pricing strategy. Pricing is the moment of truth for a new product. But far more often than not, I’ve observed new start-ups leaving ‘money on the table’ when it comes to pricing enterprise products. I’ve seen founders say their product saves hundreds of thousands of dollars—yet their product is priced as if it’s only saving thousands of dollars. 72

C hapter

7

S aa S P rici n g SaaS pricing survey (n=2268)

39%

27%

24%

10% Value-based pricing

Best judgement

Competitor-based

Cost-based

Figure 7.2 Price-setting practices in the software industry (Poyar, 2021).

The ability to capture from innovation in the software industry is limited by current pricing models: in its annual survey of software application providers, Flexera, an enterprise company, polls 334 software executives on the effectiveness of their pricing strategies and finds that only 60% of respondents rate their pricing strategy as effective or highly effective (Flexera, 2012). There is a significant gap between value creation and value capture in the industry, which may explain why software companies are seeking ways to implement subscription-based and usage-based pricing models. More on this in the sections that follow. A separate survey of over 700 managed service providers (MSPs) examines the relationship between pricing approach and software company growth (Kaseya, 2014). Companies practicing value-based pricing exhibit significantly higher rates of sales growth than companies that practice cost- or competition-based pricing (see Figure 7.3). These data confirm findings based on data from several cross-sectional studies: valuebased pricing approaches have a positive effect on company performance, whereas costbased pricing has a neutral and competition-based pricing a negative effect on company performance (Liozu & Hinterhuber, 2013, 2014).

MSP pricing survey (n=700)

Overall 2014 MMR > 10% growth

16% 11%

2014 MMR < 10% growth

25%

59%

21% 22%

68% 29%

Price/Market match

49% Cost-Based

Value-Based

Figure 7.3 SaaS pricing approaches and performance (Kaseya, 2014). C hapter

7 

73

A n dreas H i n terhuber Price getting in the software industry: Extreme levels of discounting

Price getting refers to a company’s ability to actually receive the list prices established in the price-setting phase. Across and within companies, abilities differ widely. My analysis of discounting practices based on a proprietary dataset from a large enterprise software company (Goldszmidt, 2004) leads to important insights. The average discount is 39%; the average discount, weighted by deal size, is a staggering 86%; and the average weighted discount excluding outliers (5%) is 66% (see Figure 7.4). This analysis of discounting practices in the enterprise software industry leads to the following conclusions: (1) discounting is very widespread, (2) discounting is widespread also among low-value deals where discounts are far too high: stated differently, the relationship between discounts and deal size is weak (r2: 34%); finally, (3) discounting on very large deals (deal size > US$ 100 million) is extreme (frequently above 90%), thus distorting average values. This analysis also reinforces that simple averages are misleading; thus, weighting (by size) and trimming (to exclude outliers) are necessary. The average, trimmed discount level of 66% in this dataset points to significant opportunities to improve profits via improvements in price getting.

Software pricing models

Pricing models in the software industry are rapidly changing—perpetual licensing is gradually being replaced by subscription-based pricing models. These are in turn being replaced by consumption-based pricing models. Finally, some companies are developing pricing models that link prices to business outcomes, leading to performance-based, risk-sharing pricing models. These latter arrangements are still rare (Hinterhuber, 2017) and are typically practiced in B2B. The shift is thus from perpetual licensing to subscriptions, to usage-based, and, finally, to performance-based, risk-sharing pricing arrangements (see Figure 7.5).

100%

Average weighted discount: 86%

Discount from list price (%)

80%

60%

Average weighted discount (excl. 5% outliers): 66%

40% Average discount: 39%

20% Case study Enterprise software

0%

0

100.000

200.000

300.000

400.000

500.000

Deal size (US thousand $)

Figure 7.4 Discounting in the enterprise software industry.

74 

C hapter

7

S aa S P rici n g Perpetual licensing with bundles

Bundle

Price Time

Usage based with freemium

Usage metric

Performance-based, risk sharing Price

Price

Price

Usage based, access allowance

Usage based

Usage metric

Performance-based, with penalty Price

Time

Usage metric

Subscription

Price

Price

Price

Perpetual licensing

Business impact

Business impact

Key changes: perpetual licensing to subscription, to usagebased pricing and to, finally, performance-based pricing models.

Figure 7.5 Software pricing models.

Perpetual licensing, subscriptions, and usage-based pricing models are all cases of nonlinear pricing models and can be seen as simple examples of two-part or threepart tariffs. A two-part tariff contains an access fee plus a per-unit usage fee. For subscriptions, of course, the price is entirely the access fee (and the per-unit usage fee is zero), whereas for consumption-based pricing models, the price is entirely the per-unit usage fee (and the access fee is zero). Three-part tariffs contain an access fee, an allowance for free units, and a fee for overage use based on consumption (see example ‘usage-based, access, allowance’ in Figure 7.5). Because of the complexity and relevance in B2B, for most software companies the key challenge is a move from subscription to usage-based or consumption-based pricing models. KeyBanc, an investment bank, recently completed a survey of pricing metrics of 284 SaaS companies (Spitz & Noily, 2019). The authors find that pricing is typically per user (subscription or perpetual licensing). Still, usage-based pricing models are practiced by 20% of companies (see Figure 7.6).

Best practices of usage-based pricing models

Many SaaS companies have changed their pricing models from perpetual licensing to subscriptions. The shift is significant but relatively straightforward: customers benefit because subscriptions allow real-time software updates; suppliers benefit because revenue streams are better aligned with costs. In many cases, net prices paid, customer satisfaction, and company profitability increase conjointly (Hinterhuber & Liozu, 2020). The pricing challenge essentially consists in converting the benefits of a perpetual usage C hapter

7 

75

A n dreas H i n terhuber

33%

Seats 20%

Usage/transactions 10%

Modules/functionality Employees

8%

Sites

7%

Database size

7%

Other

18%

SaaS pricing survey (n=284)

Usage-based pricing models are already widely adopted.

Figure 7.6 SaaS pricing metrics (Spitz & Noily, 2019).

into an equivalent number representing a periodic usage right. This is, despite internal and customer resistance, relatively straightforward. The next shift, from subscription to usage- or consumption-based pricing models, is a true Copernican shift. Everything changes—usage is, of course, impossible to predict, which causes problems for both suppliers (revenue forecasting) and customers (expense budgeting). Monitoring of usage becomes critical. In addition, suppliers need to define pricing metrics that are aligned with quantified customer benefits. Metrics are the basis for tracking the value customers receive and how they pay for it … Metrics are the units to which the price is applied. They define the terms of exchange—what exactly the buyer will receive per unit of price paid. (Nagle & Müller, 2018, pp. 53, 85) A first critical capability that software companies with usage-based pricing models need to master is the capability to quantify customer value: The value quantification capability refers to the ability to translate a firm’s competitive advantages into quantified, monetary customer benefits. The value quantification capability requires that the sales manager translates both quantitative customer benefits—revenue/gross margin increases, cost reductions, risk reductions, and capital expense savings—and qualitative customer benefits—such as ease of doing business, customer relationships, industry experience, brand value, emotional benefits or other process benefits—into one monetary value equating total customer benefits received. (Hinterhuber, 2017, p. 164) 76 

C hapter

7

S aa S P rici n g

Next, software companies need to identify usage- or consumption-based pricing metrics. Nagle and Müller (2018) suggest the following criteria for effective pricing metrics: (1) alignment with customer value across segments; (2) alignment with costs-to-serve; (3) easy measurement and enforcement; (4) advantage over competition; and (5) alignment with customer value within a given segment. Sample pricing metrics of consumption-based SaaS companies include ■ ■ ■ ■ ■

number of contacts/messages (HubSpot) number of users/documents (DocuSign) number of features/users (ZenDesk) number of tasks automated through the platform (Zapier) customer annual recurring revenues (Chargify).

Many of the pricing metrics currently applied by SaaS companies are input metrics— not output metrics. These metrics measure activities performed by users, but they do not reflect the economic benefits that customers experience as a result of using the software. Current metrics largely fail to track the ‘monetary value equating total customer benefits received’ (Hinterhuber, 2017, p. 164) that customers experience as a result of using the software. There is thus a substantial need for further research on effective usage or consumption-based pricing metrics. Value communication is important. Nagle & Müller (2018, p. 6) are clear: Forget what customers who have never used your product are initially willing to pay. Instead, understand what the value of the product could be for satisfied customers, communicate that value to the currently uninformed, and set prices accordingly. Low pricing is always a poor substitute for an inadequate marketing and sales effort. Finally, customer perceptions of value and price are not a given; they take place within a context. The literature on psychological and behavioral pricing (Dowling et al., 2020; Hinterhuber, 2015; Kienzler, 2017, 2018; Monroe, Rikala, & Somervuori, 2015) shows that managers can favorably influence customer perceptions of value and price without actually lowering a price.

Capabilities critical for implementing usage-based pricing models in SaaS

Research and practice on capabilities required for usage-based SaaS pricing models are still at an early stage. I outline four areas where managers need to develop individual and organizational capabilities further in order to implement usage-based pricing: ■

■

Value quantification. As outlined, many of the current SaaS pricing metrics are input/activity metrics. Managers thus need to develop value quantification capabilities to develop pricing metrics that are closely aligned with the total, incremental economic benefits that customers experience as a result of using the software. Customer experience management. Also in B2B, customers buy experiences. Customer experience is a multidimensional construct. Pine and Gilmore (2011)

C hapter

7 

77

A n dreas H i n terhuber

■

■

■

identify four dimensions of customer experience: (1) entertainment: amusing participants via their senses; (2) education: increasing knowledge or skills via active engagement of the mind or the body; (3) esthetic: passive immersion of participants in a visually appealing environment; and (4) escapist: complete immersion of participants in an emotionally appealing environment. Providing superior customer experiences is critical to increasing usage and motivating users to purchase additional software functionalities. Customer success management. This refers to the ‘proactive (versus reactive) relational engagement of customers to ensure the value potential of product offering is realized by the customer’ (Hochstein, Rangarajan, Mehta, & Kocher, 2020, p. 3). Customer success management is thus distinct from customer support and sales. The role is proactive, typically noncommercial, and designed to ensure that customers extract full value from a product. The main objective is, of course, to prevent cancellations—very easy in subscription or usage-based pricing models and, this is clear, impossible in perpetual licensing—and encourage the adoption of further functionalities and stimulate further use in the organization. Table 7.1 summarizes key activities and outlines differences between apparently similar roles such as sales and customer support (Susquehanna Growth Equity, 2019). New KPIs (key performance indicators). It is clear that traditional KPIs—such as ARR (annual recurring revenues) and ARR growth—are not meaningful for companies that implement consumption-based pricing metrics. Simply put, adoption is as important as a new subscription in the context of usage-based pricing. Traditional metrics such as ARR essentially measure new subscriptions—but not adoption. The search for effective KPIs that reflect the ability of SaaS companies to acquire, retain, and develop new customers is at a very early stage. Much of what we know

Table 7.1  Customer

success management (Susquehanna Growth Equity, 2019)

Customer success

Customer support

Sales

ProactiveReps typically work hard to engage with customers frequently and endeavor to reach them before a problem is discovered Customer-centricReps are responsible for the overall success of the customer’s project (hence, the relationship is ongoing)

ReactiveSupport quickly resolves customer issues

Proactive and reactiveReps may work on an inbound or outbound basis, driving conversations to close new deals

Source of revenueReps often have a renewal and/or upsell goal

78

Product-centricSupport Closing-centricReps are addresses only defects or incentivized to close deals shortcomings in the product ASAP, so the relationship or service delivered typically changes after the sale has been won or lost Source of costsSupport is Source of revenueReps viewed as a cost center typically have sales quotas and are directly responsible for revenue generation

C hapter

7

S aa S P rici n g

■

originates from thoughtful practitioners, such as Mark Roberge, then SVP of sales at HubSpot. At HubSpot, sales managers note that the leading indicator of churn is the use of fewer than five of the software’s 20 features during the first 60 days after purchase (Heuberger, Premo, Roberge, & Andersen, 2019). This insight reinforces the point that SaaS companies should adopt KPIs that track not only sales but actual usage by customers. New metrics. There is thus a need to complement traditional metrics of SaaS businesses with metrics that reflect customer use: gross revenue retention—the relationship between starting revenues minus downsell minus churn over starting revenues—is one such indicator. Other KPIs will need to be developed and validated.

Conclusion

In sum: pricing is important but frequently neglected. In software, value-based pricing is more widely adopted than in other industries, but cost- or competition-based pricing surprisingly still dominate. Discounting is a huge problem: the dataset used in this paper indicates that the weighted and trimmed average discount level is 66% in the enterprise software market. Pricing models are shifting from perpetual licensing to subscriptions, from subscriptions to consumption-/usage-based pricing models, and, in the future and in B2B, from usage-based pricing to performance-based, risk-sharing pricing models. SaaS companies that implement usage-based pricing models should develop four critical capabilities: value quantification, customer experience management, customer success management, and, finally, new KPIs. Much is, to be clear, still to be learned, and in many ways some of the available studies barely scratch the surface of what we should know in order to implement usage-based pricing models effectively. Research is at an early stage. Future research opportunities are numerous and can examine, for example, dynamic SaaS pricing, optimal SaaS pricing models for market penetration, or the role of artificial intelligence in building customer-specific pricing models (Li & Kumar, 2022).

Key objectives 1. Review current empirical pricing studies in the SaaS industry. 2. Illustrate the ongoing transition from perpetual licensing to subscriptions to usagebased pricing models to performance-based, risk-sharing pricing models. 3. Highlight critical capabilities to implement usage-/consumption-based pricing models in the SaaS industry.

Key summary points 1. Pricing has a substantial impact on software firm profitability but is frequently neglected. 2. About 40% of companies in the software industry set prices based primarily on customer value.

C hapter

7 

79

A n dreas H i n terhuber

3. Discounting is a huge problem: based on the dataset analyzed in this study, the average, weighted, and trimmed discount is above 60% of list price. Software companies thus have a very significant opportunity to increase profits by reducing discounting, especially on low-value deals. 4. The SaaS industry is currently transitioning from perpetual licensing to subscriptions, from subscriptions to consumption-/usage-based pricing models, and, in the future and in B2B, from usage-based pricing to performance-based, risk-sharing pricing models. 5. SaaS companies that implement usage-based pricing models should develop four critical capabilities: value quantification, customer experience management, customer success management, and new KPIs.

Key questions 1. What are currently the most widely used approaches for price setting in the SaaS industry? 2. In which respect are pricing models currently changing in the SaaS industry? Please briefly characterize different pricing models. 3. What organizational capabilities are critical to successfully implementing usage or consumption-based pricing models?

References Dowling, K., Guhl, D., Klapper, D., Spann, M., Stich, L., & Yegoryan, N. (2020). Behavioral biases in marketing. Journal of the Academy of Marketing Science, 48(3), 449–477. https:// doi​.org​/10​.1007​/s11747​- 019​- 00699​-x. Flexera. (2012). Key trends in software pricing and licensing survey: Software value perception gap. [Flexera Software white paper]. Schaumburg, IL: Flexera Software. https://resources​ .flexera​.com​/ web​/media​/documents​/ wp​- SoftSummit​-2012​- KeyTrendsSurvey​- Sof​t war​eVal​ uePe​rcep​tionGap​- 01252013​.pdf. Goldszmidt, T. (2004). Siebel case study: Pricing strategies in software. Presentation at the 2004 SoftSummit software pricing conference, Santa Clara, CA. Heuberger, M., Premo, R., Roberge, M., & Andersen, P. (2019). Six keys to customer success. Boston, MA: Boston Consulting Group. https://www​.bcg​.com​/publications​/2019​/six​-keys​-to​ -customer​-success. Hinterhuber, A. (2008). Customer value-based pricing strategies: Why companies resist. Journal of Business Strategy, 29(4), 41–50. https://doi​.org​/10​.1108​/02756660810887079. Hinterhuber, A. (2015). Violations of rational choice principles in pricing decisions. Industrial Marketing Management, 47, 65–74. https://doi​.org ​/10​.1016​/j​.indmarman​. 2015​.02 ​.006. Hinterhuber, A. (2017). Value quantification capabilities in industrial markets. Journal of Business Research, 76, 163–178. https://doi​.org​/10​.1016​/j​.jbusres​. 2016​.11​.019. Hinterhuber, A., & Liozu, S. (2012). Is it time to rethink your pricing strategy? MIT Sloan Management Review, 53(4), 69–77. https://sloanreview​.mit​.edu ​/article ​/is​-it​-time​-to​-rethink​ -your​-pricing​-strategy/. Hinterhuber, A., & Liozu, S. M. (2020). Introduction: Implementing pricing strategies. In A. Hinterhuber & S. M. Liozu (Eds.), Pricing strategy implementation: Translating pricing strategy into results (pp. 3–8). Abingdon, UK: Routledge.

80

C hapter

7

S aa S P rici n g Hochstein, B., Rangarajan, D., Mehta, N., & Kocher, D. (2020). An industry/academic perspective on customer success management. Journal of Service Research, 23(1), 3–7. https://doi​.org​/10​.1177​/1094670519896422. Kaseya. (2014). MSP global pricing survey: White paper. Miami, FL: Kaseya. Kienzler, M. (2017). Does managerial personality influence pricing practices under uncertainty? Journal of Product & Brand Management, 26(7), 771–784. https://doi​.org ​/10​.1108​/ JPBM​ -11​-2016​-1352. Kienzler, M. (2018). Value-based pricing and cognitive biases: An overview for business markets. Industrial Marketing Management, 68, 86–94. https://doi​.org​/10​.1016​/j​.indmarman​. 2017​ .09​.028. Li, B., & Kumar, S. (2022). Managing software-as-a-service: Pricing and operations. Production and Operations Management. Advanced online publication. https://doi​.org​/10​.1111​/poms​ .13729. Liozu, S. (2019). Penetration of the pricing function among global fortune 500 firms. Journal of Revenue and Pricing Management, 18(6), 421–428. https://doi​.org ​/10​.1057​/s41272​- 019​ -00209​-2. Liozu, S. (2021). The adoption of pricing from an organizational perspective and its impact on relative firm performance. Journal of Revenue and Pricing Management, 20(4), 1–13. https://doi​.org ​/10​.1057​/s41272​- 021​- 00303​- 4. Liozu, S., & Hinterhuber, A. (2013). Pricing orientation, pricing capabilities, and firm performance. Management Decision, 51(3), 594–614. https://doi​.org ​/10​.1108​/00251741311309670. Liozu, S., & Hinterhuber, A. (2014). Pricing capabilities: The design, development and validation of a scale. Management Decision, 52(1), 144–158. https://doi​.org​/10​.1108​/ MD​- 09​-2012​ -0683. Monroe, K. B., Rikala, V.-M., & Somervuori, O. (2015). Examining the application of behavioral price research in business-to-business markets. Industrial Marketing Management, 44, 17– 25. https://doi​.org ​/10​.1016​/j​.indmarman​. 2015​.02 ​.002. Moorman, C. (2020). The CMO survey—Highlights and insights report: American Marketing Association. https://cmosurvey​.org​/wp​- content​/uploads​/2020​/02​/ The​_CMO​_ Survey​ -Highlights​-and ​_ Insights ​_ Report​-Feb​-2020​.pdf. Nagle, T., & Müller, G. (2018). The strategy and tactics of pricing: A guide to growing more profitably (6th ed.). New York, NY: Routledge. Pine, B. J., & Gilmore, J. H. (2011). The experience economy: Updated edition. Boston, MA: Harvard Business Review Press. Poyar, K. (2021). SaaS pricing in 2021: Data on how 2268 SaaS companies price their products. Boston, MA: Openview Venture Partners. Sinofsky, S. (2014). The price is right: For early-stage SaaS companies, it needs to be. TechCrunch, May 16, 2014. https://techcrunch​.com ​/2014​/05​/16​/the​-price​-is​-right​-for​- early​-stage​-saas​ -companies​-it​-needs​-to​-be/. Spitz, D., & Noily, A. (2019). 10th annual SaaS survey results [KeyBanc Capital Markets White Paper]. Cleveland, OH. Susquehanna Growth Equity. (2019). Boosting revenue growth by scaling customer success [White Paper]. Bala Cynwyd, PA.

C hapter

7 

81

C hapt e r 8

The Digital Pricing Framework Best Practices in B2B Pricing and Offer Design Scott Miller

Portions of this chapter are reprinted with permission from the Professional Pricing Society’s White Paper ‘Best Practices in Pricing B2B Software & Digital Hardware Solutions: Offer Design in the Subscription Era’ (Miller, 2021).

Introduction

From an age of on-premise desktops, servers, and mainframes, to an era of connected devices, cloud computing, big data, and digital transformation, the world and technologies around us are changing at a seemingly rapid pace. These technological changes provide digital vendors not only the ability to create new value and ultimately improve their products and services, but also how such companies deliver, price, and monetize those products and services.

The complexity of pricing in today’s B2B world

There can be a high degree of complexity around pricing, monetization, and offer design in the world of B2B software and software-enabled hardware. From pricing new innovation (including AI) to XaaS, to managing large, multiyear deal opportunities, to selling into procurement arms, or even supporting clients on a variety of different software versions and delivery models (including on-premise and SaaS). It’s no wonder B2B pricing and offer design can be a confusing undertaking for product management and pricing teams. Enter a need for best practices in B2B pricing in the digital era.

The digital pricing framework

Frameworks and processes are not uncommon in the world of product development. Consider the likes of agile, waterfall, scrum, and v-model (Eriksonn, 2016). Even the product management discipline itself is governed by frameworks such as the software product DOI: 10.4324/9781003226192-11

83

S co tt M iller

management (SPM) framework (ISPMA, 2019). Such processes and methodologies are the cornerstones for delivering digital project success and positive outcomes. Pricing processes are no exception: they help establish best practices, consistency, and improved end-to-end pricing delivery across an organization that surpasses, from a growth and financial perspective, outcomes that would otherwise be achieved under more ad-hoc pricing approaches. Successfully developing a value-based segmented pricing approach involves using the right design inputs, the right company-wide price-value training, and the right approaches to executing pricing within B2B sales opportunities. It’s this full company-wide process of developing and delivering new value-based strategies in the marketplace that truly makes pricing a strategic capability for one’s organization (Figure 8.1).

Offer design

From strategy to value analysis to commercial structure development, to financial analysis and price stress testing—the offer-design phase ensures that product and pricing teams bring in the right inputs and analysis to drive the most favorable value-based segmented pricing and offer structures. A stress-tested commercial price list and discounting guidelines policy are the final outputs of this phase that ensures that both sales and product teams are armed with the right pricing tools that drive a successful commercial offer strategy.

Enablement

Whether it is internal sales teams, channel partners, customer success representatives, or even technical support staff—enablement is a critical activity for understanding how one’s solution is packaged, priced, and linked to the value story. New pricing strategies and business models are accompanied by ‘all-on-board’ organizational alignment

PRICEVALUE ANALYSIS

FINANCIAL ANALYSIS

STRATEGY

MONITOR, REVIEW, & ADJUST

CONTINOUS CYCLE OF PRICING IMPROVEMENT

COMMERCIAL STRUCTURE

PRICING ECOSYSTEM (PECO)

PREPARATION

TRAIN & ENABLE

DEAL & BID PROCESS GO-LIVE LAUNCH

SUPPORT & REINFORCE

OFFER DESIGN. Develop the right value-based offers, to the right customers, at the right price. ENABLEMENT. Drive the desired pricing strategy outcomes with enabled Sales and channel teams. EXECUTION. Best practice implementation & results tracking of pricing strategies to deliver the desired pricing outcomes. PRICING ECOSYSTEM (PECO). Laser-focused alignment of cross-department people, processes, & systems with new pricing structures.

Figure 8.1 The digital pricing framework©.

84

C hapter

8

T he D igital P rici n g F ramew o rk

programs and effective communications that bring transparency to new pricing structures, deliver new monetization opportunities, and awareness of and compliance with new guidelines and policies.

Execution

Applying an effective deal-evaluation process, assigning pricing negotiation policies, and applying pricing approval matrices avoids a ‘race to the bottom’ price approach when submitting best-and-final-offer (BAFO) in bids. Consistently monitoring winloss outcomes, optimal key performance indicators (KPIs), and insights from post-bid award reports are critical to act as an early warning system for ongoing readjustment to pricing and offer structures.

The pricing ecosystem

At the center of the framework is the pricing ecosystem (PECO); this includes all other people, processes, systems, incentive structures, and policies within an organization that are impacted by changes to revenue models, pricing, and offer structures. Crossfunctional teams need to be consulted and informed throughout the offer-design phase to fully understand the company-wide PECO implications and ensure that a well-coordinated plan drives to the desired milestones and outcomes.

The digital pricing framework subprocesses

Each major process within the digital pricing framework (offer design, enablement, and execution) is supported by underlying subprocesses (Figure 8.2) that bring rigor to the new pricing evaluation and decision processes, as well as ensuring an ‘all-on-board’ company alignment when implementing new pricing strategies. It is not uncommon when developing new pricing strategies to manage and coordinate a variety of subtasks and interconnected mini-projects across cross-functional teams. Coordinating company efforts is key to success when implementing new pricing strategies, and the subprocesses below can serve as both a ‘next steps’ checklist and a planning tool for product and pricing teams.

The road to value-based digital commercial offers

Product and pricing teams are often faced with the challenging question ‘How do we go about pricing our digital solution?’ Enter the offer-design process within the digital pricing framework. This process focuses heavily on applying value-based pricing concepts and tools to develop optimal and segmented pricing and offer structures. This stage consists of four subprocesses: (a) strategy; (b) price-value analysis; (c) commercial structures; and (d) financial analysis. In many ways, the four offer-design subprocesses can be analog to skydiving: ■

Strategy (10,000-foot pricing view) begins by gathering key inputs—strategic, industry, and competitive analysis that determines pricing guidance principles, product, and customer segmentation strategies.

C hapter

8 

85

86

5

4

3

2

1

Comparative Reference Point Sales Story

Customer Research

Intra-Portfolio Price-Value Analysis

Price-Value Analysis vs. Competition

Economic Value Analysis

Price-Value Analysis

8

7

6

5

4

3

2

1

Alignment of Global and M&A Pricing Structures

Channel and/or InterBU Pricing Model

Discounting Guidelines

Commercial Terms & Pricing Schedule

Maintenance, Support, Customer Success Pricing Models

Segmented Offers

Price, Packaging, Metrics, & Tiers

Revenue & Delivery Model(s)

Commercial Structure

6

5

4

3

2

1

Financial KPIs and Success Metrics

Revenue Recognition & Cash Flow

Customer ROI & Payback Analysis

Customer-Level Impact Analysis

Mock Deal Financial Analysis

Business Case

Financial Analysis

Figure 8.2 The digital pricing framework© subprocesses.

Channel & Channel Partner Strategy

3rd Party Partnerships (Teaming)

Inventory Capacity (Supply/Demand)

Monetization Strategy & Tactics

Roadmap & Versioning Strategy

Competitive Analysis

Segmentation Analysis

Market & Industry Analysis

Value Proposition Analysis

Link Price to Objectives & Strategic Plan

10

9

8

7

6

5

4

3

2

1

Strategy

Offer Design

Sales Toolkit(s) & Incentive Plans

Marketing Plan & Omnichannel Marketing Material

Communication Plan & Materials

Training Plan & Materials

Pricing Ecosystem (PECO) Planning:

Pricing Guidelines and Compliance Documents

Resources, Processes, & Systems across Cross-Functional teams to Support New Pricing Structures

7

6

Price List Documentation (Published / Unpublished)

5

4

3

2

1

Prepare

4

3

Channel Partner Training

Cross-Functional Team Training

Sales Training & Change Management

Train Digital Champion Leads

2

1

Train & Enable

3

2

1

Enablement

Reinforcement Communications

Customer Feedback Loop

Supporting Pricing Resources

Support & Reinforce

3 Go-Live Pricing Ecosystem (PECO)

Implement Migration Strategy

Launch Go-Live Communications

Resources, Processes, Systems, Compliance

2

1

Go-Live Launch

7

6

5

4

3

2

1

Promotions & Campaigns

Post-Tender “Change Request” Pricing

Contract Signing

Due Diligence, Risk, & Compliance

Deal Preparation & Submission

Price-to-Win Strategy

Bid / No-Bid Assessment

Deal & Bid Process

Execution

5

4

3

2

1

Win-Loss / VoC Research

Renewal Pricing Management

Level III Analytics: Optimization

Level II Analytics: Enhanced Insights, Cost-to-Serve

Level I Analytics: Sales, KPIS, Pricing Dashboards

Monitor, Review, & Adjust

S co tt M iller

Deal Review Process

C hapter

8

T he D igital P rici n g F ramew o rk ■

■

■

Price-value analysis (2,000-foot pricing view) seeks to determine approximate price positioning based on perceived value, client requirements, use cases, and willingness to pay across key segments. Commercial structures (500-foot pricing view) seek to create segmented price lists (packages, price metrics, and tiers) as well as discounting guidelines and approval matrices. The pricing outcomes need to link to the values story as quantified in the price-value analysis phase. Financial analysis (ground-level pricing view) is the final phase that involves conducting price stress testing as well as profitability and cash flow analysis that contributes to final decisions around commercial offers and pricing policies.

Following these four processes within the offer-design phase will ensure that your organization has conducted a rigorous and well-thought-out assessment of pricing strategies and tactics that sets the stage for your final go-to-market commercial offers. Strategy (the 10,000-foot view)

Strategy is a first step to gathering the right inputs that determine the go-forward shortand long-term pricing and offer a design approach. Not only does a strategic assessment help guide offer design, it also helps to drive product strategy decisions, including product positioning, new product introductions, product marketing, and value selling, as well as value innovation and portfolio investments (Gale & Swire, 2012). The strategy review process involves integrating and interpreting the implications of a variety of inputs from the following strategy-related categories. Product management teams will need to spend adequate time to gather, discuss, and interpret strategic insights and determine how this impacts the short-, medium-, and long-term go-to-market pricing, customer, and product strategies. From a pricing and offer-design perspective, this strategic analysis should be documented as a set of guiding principles that act as a vision when designing and selecting optimal offer structures. Although each of these strategy subprocesses, as defined in Figure 8.2, can easily be chapter topics on their own, monetization strategy and tactics are most definitely a key consideration for digital product and pricing teams. Key to a successful monetization strategy is growth … growth … and more growth. This growth strives for maximizing customer lifetime value (CLV) and needs to be strongly linked to the right value, to the right segments, and at the right price and offer structure. Various tactics (see Figure 8.3) can be applied to drive additional growth with new, current, and returning clients and are closely linked to future roadmap development and effective offers. The importance of strategic analysis can never be underestimated: it is the precursor to building sound pricing strategies, structures, and tactics. Ensure that your team reviews these on a quarterly basis to identify new opportunities and changes within a fast-moving market.

Price-value analysis (the 2,000-foot view)

Value-based pricing is defined as setting prices primarily to the perceived or estimated value of a product/service to the customer, rather than according to the cost of the C hapter

8 

87

88

Vertical Market Horizontal Markets Government Markets New Regions Teaming Arrangements

Niche Play APIs (v. Software) Hidden Assets (“Data new Oil”) White/Private Label

Pre-Cancelation Offer Promotions to Legacy (Canceled) Clients

Customization Consulting

Add New Module / Feature Set

Upgrade from OnPremise to SaaS Managed Services / BPO

Training

Negotiated Renewal

Services

Upgrade to New Version

Top-Ups / Overages

Auto Renewal

Renewals, Retention, & Winback

New Product Category

Cross-sell

Upgrade Product Tier

Upsell

Figure 8.3 B2B digital monetization strategy and tactics.

• • • •

New Markets - Solution Modifications

• • • • •

New Markets

• New Adopters • Direct Channel • Indirect Channels • Client Referral Program • From Competition • Same as above • Trade-in Program • RFP (renewal bids)

Current Market

New Acquisition

• • • •

Discounting Guidelines T&Cs Cost-to-Serve / Client P&L SKU rationalization

Managing “Net Price” & Margin Leakage

Beta/MVP/Pilot Program

Promotions

Ancillary Fees

• Price/Cross Elasticity • Price-Mix-Volume (PVM)

Price Optimization

Price Increases

Other Tactics

S co tt M iller

C hapter

8

T he D igital P rici n g F ramew o rk

product or historical prices. Linking price with value can be achieved using two useful software pricing tools: (a) economic value analysis and (b) price-value trade-off analysis. These tools will help establish the higher-level price-positioning views that will serve as the price targets for the underlying roll-up of the price structure (packaging, metrics, and tiers). Economic value analysis. Case example. A product team within a large global services firm was developing a solution that used artificial intelligence (AI) to identify, organize, extract, and report (IOER) keywords and clauses within large documents. The product team, however, was struggling to determine an ideal pricing model beyond guesswork around willingness to pay. In their case, economic value analysis was an ideal starting-point pricing tool of choice. The team first identified ‘how’ their solution improved their clients’ pain points by quantifying their client’s KPIs, which included 10 full-time employees (FTE), conducting IOER for over 5,000 documents per year, averaging 150 pages per document, at 1.5 minutes per page. They then devised realistic and achievable changes to those KPIs as a result of using their AI solution. Most notably, scanning time could be decreased by a very conservative 75% (from 1.5 to 0.38 minutes). This savings in IOER time could drastically help clients reduce the need for manual scanning FTE labor costs (or alternatively, a reallocation of those FTEs to better value-add activities within their organization), resulting in as much as $280,000 in FTE cost savings per year. This financial benefit could now be used as a reference point in their pricing discussions: ‘The annual savings we agreed you could achieve per year is $280K. At our subscription price of $42K per year, you are still retaining 85% of those annual financial benefits.’ Price-value trade-off analysis. Also known as best value trade-off evaluation (Watson, 2015) in US federal government procurement circles, this tool is particularly useful not only to quantify and link the price and value relationship but also to understand the client buying-decision drivers, and how your digital offering compares against next best alternatives: these alternatives can include competition, intra-portfolio offers (cannibalization), or even the potential for clients to develop their own in-house solutions. This analysis seeks to quantify both price and value using a weighted scoring approach, factoring in the trade-off that a client segment makes between price and value. One major advantage of using this tool for larger B2B and business-to-government (B2G) client opportunities is that it also mirrors the commonly used scoring approach used by procurement arms as part of their request-for-proposal (RFP) assessment to determine a vendor contract award. Final scores are determined based on the tradeoff between price and value (a reflection of price sensitivity, which can differ across segments). For example: in Figure 8.4, Software Company D won a procurement opportunity despite having a price premium of $2.45 million compared with the other vendors. Software Company A ($1.53 million) contested the bid (‘we were the better price’) but lost the protest as a result of the bid being considered a ‘trade-off’ type evaluation (20% price and 80% value). In this case, Software Company D scored the highest 6.6 total score for a win.

C hapter

8

89

S co tt M iller

Overall price-value scorecard: (A)

(B)

(A)x20% + (B)x80%

Good

Price Score (A) 6.2

Value Score (B) 6.4

(A)x20%6.4 + (B)x80%

$1,210,000 Price

Acceptable Value

6.7 Price Score

Value 5.0Score

5.3 Total Score

$943,000 $1,530,000

Marginal Good

7.5 6.2

4.8 6.4

5.3 6.4

$2,450,000 $1,210,000

Outstanding Acceptable

5.2 6.7

6.9 5.0

6.6 5.3

Price

Value

Software Company A

$1,530,000

Software Company B Software Company CA Software Software Company Company DB

Total Score

Software Company C

$943,000

Marginal

7.5

4.8

5.3

Software Company D

$2,450,000

Outstanding

5.2

6.9

6.6

Software Company A

Software Company B

Software

Software Company D

Software Company C 3.6

Software Company D 5.8

Value score breakdown: Technical Merit

Weighting Weighting

(i) Technical and Functional

Scoring

Company C Scoring

Technical Merit Software solution

25%

Software Company A 4.3

Software Company B 4.0

(i) Technical and Functional Implementation

10%

8.1

7.0

6.0

8.1

25% 10%

8.1 4.3

6.0 4.0

5.0 3.6

8.1 5.8

10%

8.1

7.0

6.0

8.1

10% 10%

7.5 8.1

3.0 6.0

4.0 5.0

7.5 8.1

Past performance (ii) Experience and Past Performance

10%

5.6

4.0

5.0

7.5

(iii) Management Approach Organization experience

10%

7.5

3.0

4.0

7.5

5% 10%

6.7 5.6

5.0 4.0

5.0 6.0

7.5 6.7

(iii) Personnel Management Approach

5%

9.0

6.0

5.0

4.5

Oral & Solution Demo Presentation Approach

5%

6.7

5.0

6.0

6.7

5% 5%

8.0 9.0

7.0 6.0

4.0 5.0

8.0 4.5

Business scenarios (i) Oral Presentation

5% 5%

6.0 8.0

5.0 7.0

6.0 4.0

8.0 8.0

Ability todemonstration satisfy goals (ii) Solution

5%

6.0

6.0

4.0

6.0

Navigation / Ease-of-Use Business scenarios

5% 5%

6.0 6.0

6.0 5.0

8.0 6.0

8.0 8.0

Effectiveness of goals team Ability to satisfy

5% 5%

6.0 6.0

4.0 6.0

4.0 4.0

6.0 6.0

100% 5%

6.4 6.0

5.0 6.0

4.8 8.0

6.9 8.0

Hosting Software solution (ii) Experience and Past Performance Implementation Organization experience Hosting

Approach Past performance

(i) Oral Presentation Personnel (ii) Solution demonstration Oral & Solution Demo Presentation

TOTAL VALUE SCORE (100%) Navigation / Ease-of-Use Effectiveness of team

TOTAL VALUE SCORE (100%)

5%

6.0

4.0

4.0

6.0

100%

6.4

5.0

4.8

6.9

Figure 8.4 Price-value trade-off analysis.

In many cases, product teams can leverage value scores to provide better direction around product roadmap prioritization (i.e., focus on what is important to clients).

Commercial structure (the 500-foot view)

This process focuses on creating well-defined pricing and offer structures that integrates the 2,000-foot view price-positioning findings from our previous price-value analysis. During this process, teams will need to assess (a) optimal packaging formats that support pricing and monetization objectives, (b) ideal metrics that address differing use cases and requirements across segments, and (c) tiering that takes into consideration volume discounting and/or price levels. The goal of this phase is to develop an indicative commercial price list for sales teams alongside guidelines and policies to manage allowable negotiation and price floors. Packaging

Similar to restaurant menus, digital solutions (hardware, software, and services) can often be sold as either an à-la-carte approach or more as a bundling approach—these 90

C hapter

8

T he D igital P rici n g F ramew o rk

bundles can be used to target different segments with differing needs (value) and differing price sensitivities. It will also be important to assess which package best aligns with the original guiding principles as defined in the strategy phase. Functional packages, for example, may be preferred in those cases where clients integrate a mixture of different vendors into a larger system—for example, functional packages are a common occurrence within large ERPs whereby each solution of finance, payroll, HR, and cash management can be purchased as a standalone with integrations into a larger cohesive system. Alternatively, modular packages may prove to be more beneficial as a packaging type during a mature life-cycle phase that enables monetizing a legacy client base with new innovations and feature sets as ongoing addons; modules are often not sold as a standalone functional solution but instead rely on a base platform in order to provide enhanced functionality (Figure 8.5). Decisions around packaging types are closely linked to the short and long-term monetizing strategies that seek to maximize revenue and growth. All too often, software vendors will ‘build it first’ and then realize after the fact they are undermonetizing as a result of using a suboptimal packaging format. It will be critical for product and pricing teams to assess how best to price and package their software offering at the beginning of the product development stage to ensure all roadmap efforts are closely aligned with an optimized pricing strategy.

Metrics

A software metric is a standard unit of measure that links a fee structure to six possible software dimensions: access-based (who is accessing); architecture-based (resources being accessed); content-based (what is being accessed); usage-based (how much and how often it is being accessed); and outcome- and revenue-based metrics. In a more simplistic software pricing environment such as B2C, one metric can serve as the basis for determining the pricing strategy (e.g., per-user fee). However, where B2B solutions are concerned, a greater degree of solution complexity and range of varying client use cases often calls for the use of multiple underlying metrics that roll up to determine the overall license or subscription fee. Commonly used licensing and subscription metrics include those shown in Figure 8.6. ‘Which metrics are the best metrics for our digital solution?’ There is no magic bullet answer—but there is a process for evaluating the options. Every B2B solution is unique and requires an assessment to evaluate potential options while weighing the respective pros and cons. During this metric evaluation phase, product and pricing teams should be asking some key questions: Do the metrics represent value delivered? Are they scalable and provide incremental monetization opportunities? Is it sellable and well understood by the client? Can the metrics be measured, tracked, and billed?

One-SizeFits-All

GoodBetter-Best

Functional

Modular

Customized

Hybrid

(ex: Modular with G-B-B)

Figure 8.5 Digital solution packaging types. C hapter

8 

91

92

Figure 8.6 Common software metric types with examples.

% of cost savings Per closed/won Loan

Outcome-based

Workflows Workflow type 1 Workflow type 2…

Customization Per custom reports/pages Per custom dashboards Per KLOC (lines of code)

User/Device Fees Named User Named User (by profile) Named User (limited) Daily Active User (DAU) Concurrent User Per Connected Device …

Location-Based Fees Country/Region fee Localization fee

Environment Fees # interfaces Server license Hosting license Per Instance Per connection/node Storage (Per GB / TB)

Architecture-based

Platform Fees Platform access fee Platform (non-prod / test)

Access-based

% of transaction fee % of customer revenue

Revenue Share

Advanced Features Advanced reports Advanced models Enhanced data

Insights By dataset type By report type By model type

Actual Content Content library#1 Content library#2… 3rd party content

Content-based

Proxy for Usage: Company size (Revenue) #FTEs Budget size #Customers #accounts, # branches

Transaction ranges Reports accessed ranges

(e.g., “between X and Y volumes”)

Usage Ranges

Per Usage Per transaction fee Per resource unit Per message/call (APIs) Per batch job v. on demand

Usage-Based

S co tt M iller

C hapter

8

T he D igital P rici n g F ramew o rk

Case study. One software company, that specialized in helping US health care insurance providers detect insurance claim errors and fraud using AI, changed their licensing metrics from number of insured members (a good proxy for usage and health insurance company size) to a per transaction metric (i.e., number of insurance claims scanned using their AI solution). This new metric was a much better reflection of their solution’s value proposition, resulting in +30% revenue for both new and renewing opportunities. Tiering

Tiering is a methodology that creates variation in pricing and/or packaging value that targets a defined client segment. These clients can vary in terms of size (small, medium, large), usage volumes (low to high), requirement needs (low to high feature requirements), and/or price sensitivity. The most common tiering approaches can include the following: 1. Volume discount matrices. A starting point for determining volume discounts is to understand where you want to price your smallest client tier (ceiling prices) and lowest allowable price to your largest client tier (price floor); this approach can help to establish your overall discount slope across tiers. Volume discount matrices are useful to standardize predefined price concessions with clients based on volumes/usages. Be mindful that there is a respective link to other sales negotiation practices as well in order to mitigate ‘discounting on top of discounting’ scenarios. 2. Good-better-best bundles. This tiering approach serves both as a packaging type and a tiering methodology. Good-better-best can address several strategies including offensive plays aimed at generating new growth and revenue, defensive plays meant to counter or forestall moves by competitors, and behavioral plays that draw on principles of consumer psychology, whatever the competitive landscape (Mohammed, 2018). 3. Price levels. This tiering approach involves using defined categories of client segments that receive differences in pricing. For example, a software firm selling to municipal governments might create six price levels, each consisting of ranges of number of inhabitants (Level A = 1 to 25,000; Level B = 25,001 to 50,000; etc.). Price levels are often useful when targeting a clearly defined market segment and provide simplicity in a price structure that is easy to sell and communicate with clients.

Indicative commercial price list (pre-financial analysis phase)

Creating the indicative (pre-financial analysis) price list brings together all the offerdesign analyses as determined from strategy, price-value analysis, and commercial structures (packaging, metrics, and tiers). Commercial price lists can be used in external communications (published to clients, GSA price lists) or used for internal price purposes only (deal pricing calculator tool). It is also important for product and pricing teams to clearly define the standard term and conditions (T&Cs), pricing schedule, and renewal options associated with offers (Figure 8.7). C hapter

8 

93

S co tt M iller

Total Contract Value (5 Years) Total Contract Value (TCV)

$7,365,000

Contract Fee Summary One-Time Fees

Implementaon Fees System Support & Maintenance Hardware Soware Training (onsite) Training (remote, online) Training documentaon Other Total One-Time Fees

$2,790,000

$305,000 $65,000 $30,000 $3,190,000

On-Going Fees

SaaS Subscripon Customer Success Premium Total On-Going Fees

$800,000 $35,000 $835,000

SaaS Subscripon Pricing Schedule 5-Year Term Commitment Year Annual Subscripon

Year 1 Year 2 Year 3 Year 4 Year 5

Metrics

$800,000 $800,000 $800,000 $800,000 $800,000

6,400 FTE 6,400 FTE 6,400 FTE 6,400 FTE 6,400 FTE

$832,000 $832,000 $832,000 $865,280 $865,280

6,400 FTE 6,400 FTE 6,400 FTE 6,400 FTE 6,400 FTE

$125

per addional FTE above 6400 baseline

5-year commitment non-cancelable

*FTE = Full-Time Employee

Renewal Opons

Year 6 Year 7 Year 8 Year 9 Year 10

Renewal Opon +4% + 3 years Renewal Opon +4% + 2 years

Overages

Figure 8.7 B2B digital solution pricing schedule (example). Financial analysis (ground-level view)

The last subprocess within the offer-design phase involves conducting financial analysis and price stress testing to determine price structure breaking points, potential for pricing and offer structure rework, and even the possibility of re-evaluating the larger product and segmentation strategies. This testing of model inputs and financial views will need to be done across a variety of different possible scenarios to determine where a price model may have some unforeseen nuances, improvement areas, or even high-risk areas that will require rework. These risks can ultimately influence adjustments to one’s pricing structure and offer design, identify which profitable client segments to target (and unprofitable segments to avoid), prioritize internal activities to mitigate certain costs, or even highlight a need to change the product strategy itself. Various tools can be applied to analyze and stress-test price structures. For example, once an indicative pricing structure has been determined, it can be useful to conduct mock deal financial analysis as well as customer-level impact analysis: 1. Mock deal financial analysis. Mock deal analysis should be conducted across small, medium, large, and super-sized client-type profiles. Finance teams can help link these models with direct and indirect cost allocation assumptions. Be sure to test extremes for possible metric outcomes that can ‘break’ a pricing structure. In many cases, for example, mock deals can often show low or negative profitability to smaller client segments, identifying a need to rework an offer and the respective volume curve, or to re-evaluate a low cost-to-serve commercial offer as a whole that is more profitable with the smallest client segment. 2. Customer-level impact analysis (‘gainers vs. decliners’). This analytical approach can help a product team understand how historical clients would have been priced under the new price structures. Are the new pricing structures driving to the desired objectives? Which clients would see the largest gain in pricing? Which clients would see the largest decline in pricing? This analysis is often useful to help tweak pricing and offer structures across client types to reach the desired objective. 94

C hapter

8

T he D igital P rici n g F ramew o rk

And in those cases that will involve migrating legacy clients to the new pricing structures, gainers versus decliners helps sales teams plan for potentially challenging price increase conversations.

Conclusion

With a solid foundational understanding of B2B industry pricing practices and by adopting processes and tools defined within the digital pricing framework, your organization can embark on its journey toward developing and implementing optimal valuebased and segmented offer structures that not only compete in the marketplace but also deliver material improvements to your organization’s recurring revenue and market performance.

Bio

Scott Miller, CPA, CMA, is Founder and President of Miller Advisors Inc (www​.miller​ -advisors​.com), a B2B pricing and commercialization consultancy with a specialty in software and digital solutions. Over the past 20 years, Scott has held various roles as head of global pricing with $10 billion technology companies and has conducted over 100 pricing initiatives as a senior consultant. Scott is also a Fellow of the International Software Product Management Association (www​.ispma​.org), an educator with the Professional Pricing Society, a published author, and speaker on best-in-class pricing practices. Scott can be reached at scott​@miller​-advisors​.​com.

Key objectives 1. Introduction to the digital pricing framework. 2. Provide an overview of the pricing and offer design process for B2B digital solutions. 3. Highlight industry examples that apply pricing tools to link pricing to the digital solution value story. 4. Overview of common commercial structures for B2B digital solutions. 5. Overview of price stress testing tools for proposed commercial offer structures.

Key summary points 1. Pricing for B2B digital solutions can be treated as an end-to-end process (via the digital pricing framework) that includes three core processes: offer design, sales enablement, and execution. 2. The offer-design phase is a staged value-based process that includes defining strategy, conducting price-value analysis, developing commercial structures, and conducting financial analysis.

C hapter

8 

95

S co tt M iller

3. Economic value and price-value trade-off are two useful pricing tools to link pricing to a quantified value story. 4. Commercial structures for B2B include aspects of ‘packaging, metrics, and tiers’ as well as discounting guidelines, terms and conditions, and renewal options. 5. Financial analysis includes stress testing a proposed commercial structure using tools that include mock deal analysis and customer-level impact analysis.

Key questions 1. What processes can help ensure that product teams adopt best-in-class value-based pricing and offer design for their digital solutions? 2. What pricing tools can be used to improve how to price digital solutions, and link that pricing to the ‘value story’? 3. What key considerations are required to develop a final commercial structure for B2B digital solutions?

References Eriksonn, U. (2016, May 19). The A to Z guide to the software testing process. https://reqtest​ .com ​/testing​-blog ​/the​-a​-to​-z​-guide​-to​-the​-software​-testing​-process/. Gale, B., & Swire, D. J. (2012). Implementing strategic B2B pricing: Constructing value benchmarks. Journal of Revenue and Pricing Management, 11(1), 40–53. ISPMA. (2019). https://ispma​.org ​/framework/. Miller, S. (2021, September). Best practices in pricing B2B software & digital hardware solutions: Offer design in the subscription era. Professional Pricing Society. https://bit​.ly​ /32XT8x7. Mohammed, R. (2018, September). The good-better-best approach to pricing. Harvard Business Review. https://hbr​.org ​/2018​/09​/the​-good​-better​-best​-approach​-to​-pricing. Watson, K. (2015). LPTA versus tradeoff: How procurement methods can impact contract performance. Monterey, CA: Naval Postgraduate School.

96

C hapter

8

C hapt e r 9

Tapping into the Subscriber Psychology with Good/Better/Best Is There an Optimal Ratio between Tiers? Gaurav Sonpar and Michael Mansard

Executive summary

The rise of subscriptions has positioned services, and the recurring revenue they generate, front and center. And services are a fantastic win-win for companies and their customers in terms of value. But the thing about value is that you must put a number on it. That is where subscriptions become a challenge. While pricing and packaging is one of the most powerful mechanisms to capture new subscribers, it’s also the area that requires the most notable change in mindset. To successfully monetize subscription-based products and services, businesses need to completely move away from the cost-plus paradigm. This new world is about value— from the customer’s perspective. Many organizations taking the plunge into subscriptions tend to stick with what has worked in the past, specifically, a linear product-centric approach, and cost-plus pricing. Instead of reimagining value metrics and packaging tiers from the subscriber perspective, they instead obsess over the price point. Now is the time for that to change. To that end, we created this chapter on the art and science of good/better/best (GBB) pricing and packaging for subscription companies. Leveraging numerous client engagements and industry-leading practices, we have developed a 3D pricing and packaging (P&P) framework to help companies launch and refine a GBB-based packaging model. Along the way, we have embedded real-life examples, specific guidance, and benchmarks. Advanced readers will also be delighted to check out the ‘Design’ section, where we reveal unique benchmarks for GBB pricing. This section leverages data points from leading players to articulate guardrails for price differentials among the tiers.

The shift to usership is accelerating

The rise of the subscription economy and the shift to consumption-based models means that customers today value access to services over ownership of products. And while DOI: 10.4324/9781003226192-12

97

G aurav S o n par a n d M ichael M a n sard

this phenomenon was once limited to media and tech businesses, it has now extended across various industry sectors. Companies that successfully manage this shift to usership are reaping the benefits. The most recent subscription economy index report found that subscription businesses experienced faster growth rates compared with the S&P 500 in 2021, with 16.2% and 12% revenue growth, respectively (Zuora, 2022). 70% of business leaders say subscription business models will be key to their prospects in the years ahead. (Global Banking & Finance Review, n.d.) When purchasing subscriptions, buyers weigh their options to make the initial transaction, but they also have expectations after the ‘sale’—and their expectations are constantly evolving. The customer’s relationship with the company is no longer just a moment in time; it’s an entire series of interactions, including future transactions and renewal decisions. In effect, the linear decision-making process has turned into a living, breathing, dynamic subscriber-centric journey. In a subscription model, the customer decision-making process follows a distinctive pattern: 1. 2. 3. 4.

identifying business or individual needs that are met by the offering gathering information about the offering from direct and indirect sources evaluating options based on key considerations (like price and willingness to pay) eventually (hopefully!) subscribing to the service

But the purchase process itself is just the beginning. Customers expect ongoing value beyond the initial transaction, with post-purchase interactions and experiences playing a critical role. In many ways, this is good news because it costs less to keep existing customers than to acquire new ones. That is why market leaders focus on improving customer lifetime value (CLV). The shift from product-centric to subscriber-centric

one-time transactions > sold through recurring subscription and consumption growth through unit margins > growth over customer lifetime consumer disposes of product (no end-of-life opportunities) > product returns (drives circularity with new end-of-life opportunities) cost-plus pricing and packaging > value-based pricing and packaging Linear, product-centric model

One-time transaction where the consumer buys a product without a longer-term commitment (Figure 9.1​). Dynamic, subscriber-centric model

An agreement with the consumer that facilitates regular delivery or long-term use of a service and/or product (Figure 9.2). 98

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

PRODUCT

CHANNELS

Figure 9.1 Linear, product-centric model.

SUBSCRIBER EXPERIENCE

CHANNELS Figure 9.2 Dynamic, subscriber-centric model.

C hapter

9 

99

G aurav S o n par a n d M ichael M a n sard Pricing plays a critical role in the customer’s decision-making process

The single most important decision in evaluating a business is pricing power. If you have the power to raise prices without losing business to a competitor, you have a very good business. And if you must have a prayer session before raising the price by 10 percent, then you have a terrible business. Warren Buffett Pricing is much more than just a number. It’s a representation of the value that customers perceive from your goods and services. Yet too many businesses set prices without even considering the customer’s perspective. This oversight can have costly consequences. It could result in unsuccessful subscription launches for companies looking to make the transition to consumption-based models. For companies who already offer subscriptions, pricing mistakes could hold them back from scaling up or reaching revenue targets. In fact, McKinsey (Wilton & Khanna, 2018) estimates that packaging mistakes translate to a loss of 5 to 15 percentage points of revenue growth per year for those companies. Pricing and packaging mistakes are a problem because they directly impact customer decision-making. A price that is beyond customers’ willingness to pay sends a message that your company’s offering is out of reach. A low price may lead customers to conclude that your offering is of low value or low quality. The solution is to align pricing and packaging with the target segments for your business. We call this customer-centric model value-based pricing. It’s worth noting that value-based pricing does not mean value for money. Customers are prepared to pay more than most businesses assume they will, provided clear and structured options are presented to them. However, their willingness to pay higher prices directly correlates to their needs, whether specific (and immediate) pain points have been solved, and whether packaging has been thoughtfully designed to aid the decision-making process. In a traditional transactional business, making a pricing mistake is unfavorable, but it’s not the end of the world. But in a recurring business, it’s a catastrophe that repeats over and over for the life of the subscription—with devastating downstream effects on the CLV. The question then becomes: how do you use packaging to communicate value to customers, while also creating meaningful choices? Pricing and packaging is more than just a means for your own business to succeed; it’s also a competitive tool that helps differentiate your offering. Now let’s look at some of the different packaging models used in business today. Recurring revenue packaging models

There’s no single right way to design pricing and packaging. Packaging is simply the bundling of features and functionalities to formulate a unique value proposition for the target audience. Depending on the industry, the competitive landscape, the level of subscription maturity, and the company’s strategy, the calculation will vary. Figure 9.3 shows examples of packaging models being used by different leaders in the subscription economy. 10 0 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

Figure 9.3 Pricing and packing strategies. Source: Simon Kucher & Partners, with enrichments from Subscribed Strategy Group.

Choosing the best model for your business comes down to who your customers are and what they need. For example, GoPro targets a monolithic customer type (consumers and content creators looking to produce amazing, professional-quality video) who are looking for a simple transaction—'a buy it and use it’ package. Whereas AWS serves large enterprise customers who expect solutions that have been completely tailor-made for their needs, the opposite of GoPro. For organizations looking to launch new offerings with multiple customer targets and packaging options, GBB is often an effective approach. Most companies need a package that is flexible enough to satisfy a wide range of customer types, allow upgrades and downgrades during the customer journey, and clearly communicate the value proposition at the point of sale. That is why we recommend the good/better/best model. Why good/better/best?

What makes the good/better/best (GBB) packaging model so compelling is its simplicity. GBB is driven by a well-studied behavioral mechanism observed in decision theory: asymmetric dominance. This concept, also called the decoy effect, is best illustrated by the beer trial. The decoy effect in action: The beer trial

Despite the rational nature of comparing numbers, price decisions usually have a strong intuitive component. People’s perception of prices can be influenced by C hapter

9 

10 1

G aurav S o n par a n d M ichael M a n sard

Figure 9.4 The beer trial.

introducing a cognitive bias in the decision-making process. Behavioral economist William Poundstone (2011) highlighted this cognitive bias in a series of trials with a group of business undergraduates (see ‘The beer trial’). The beer trial demonstrated people’s psychological inclination to avoid extreme options in favor of intermediate or middle options (Figure 9.4). Those who chose the middle-priced option of the three described their decision as a ‘safe’ or a ‘compromise’ choice. The cheapest beer might taste terrible, and the most expensive might be a ripoff, but one in the middle of the pack ought to be okay. Today, GBB pricing can be seen everywhere, from SaaS to media to insurance companies. It allows companies to target as much of the market as possible by diversifying their offerings to customers of different sizes and budgets. It simplifies the decisionmaking process for customers while paving the way for future upgrades to ‘better’ or ‘best’ options. And, when building a new market category, the different, progressive tiers of GBB provide flexibility to adjust to uncertainty. The decoy effect has even proved effective in commoditized markets. For example, AllState experienced significant uplift for its insurance product by repackaging its offering with a GBB model (nicely documented in the famous Harvard Business Review article ‘The Good-Better-Best Approach to Pricing’; Mohammed, 2018). 10 2 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est The beer trial​

Examples of good/better/best (Mailchimp, n.d.) provides a good example of valuebased GBB price points. The tiered pricing plans address a wide range of customer segments, from businesses starting out to those on a growth trajectory, to mature businesses already collaborating with many contacts (Figure 9.5).​ Netflix offers GBB tiers based on video quality. At each successive tier, the screen resolution improves, which provides a meaningful differentiation for subscribers and helps reach a wide range of customer targets (Figure 9.6). Considerations when adopting GBB pricing Benefits ■

■

■

■

■

■

■

Watchouts

Expand market reach. Attract a wider range of subscribers by offering choices targeted to different customer segments. Grow revenue pools. Capture more of the total addressable market (TAM) and increase conversion rates by meeting a wider range of customer demands.

■

■

Predictable revenue. More accurate revenue forecasting and planning to manage growth.

■

Diversified pricing strategies. Flex and align tiered packaging models with pricing strategies, e.g., set low base pricing for an effective penetration strategy.

■

Clear upgrade path. Subscribers easily see and understand the upgrade rationale for the next-best tier.

■

Maximize CLV. By ring-fencing features, companies can allow customer downgrade—it’s better to downsell than to churn. Name the game. Naming each tier supports your marketing narrative, especially when building a new category.

‘Paradox of choice.’ Providing prospects with too many options can trigger decision fatigue or choice overload, resulting in choice deferral. Packaging and optimization. Ring-fencing tiers (based on features, usage, etc.) and pricing is challenging and requires continuous testing and iteration to find the right balance. Apple-to-pear comparison. Subscribers are unable to make a price-to-value comparison if features vary too much across tiers, resulting in decision fatigue. Managing subscriber expectations. Some subscribers will always be unhappy about not getting everything at their price point. Suboptimal value capture in some contexts. With a broad user base, some customers will use the product or service in a myriad of unexpected ways, which is why Zoom created industry-specific packaging and pricing for education and health care while offering GBB for enterprise customers.

To summarize, tiered GBB packaging enables companies to position offerings to a wider spectrum of customers, resulting in better revenue monetization opportunities. But pricing, launching, and managing these tiers require a sizeable effort to meet a wide range of customer demands. When it comes to GBB, companies will need to put in the work to make it work.

Finding the right balance: How to build a GBB package

GBB is an effective strategy for addressing multiple customer segments because it offers differentiated, progressive service tiers at increasing price points. But to be effective, C hapter

9 

10 3

Figure 9.5  Examples of good-better-best.

G aurav S o n par a n d M ichael M a n sard

10 4 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

Figure 9.6  Examples of good-better-best.

GBB tiers must incorporate meaningful trade-offs between price and perceived value. This means finding the right balance of features, services, quantity offered, usage, customization, and so forth. It’s paramount that the distribution across the tiers matches target customer segment(s) and is aligned with the revenue strategy. Finding the right balance isn’t easy. For example, if you load the base ‘good’ tier with extra features that meet the requirements of a premium segment (with higher WTP), you will undermine the perceived value of the higher tiers. Your customer pool will skew toward the lower tier, adversely impacting your overall revenue numbers. On the other hand, if your ‘good’ version is too light, or the top-tier ‘best’ version is overloaded with high-value features, you won’t provide a meaningful choice for customers no matter what the price differential is, because the ‘best’ tier will be seen as the only quality option. So how do you ‘get it right’ when it comes to GBB tiers? To help you navigate this process, the Zuora Subscribed Strategy Group (SSG) has developed the 3D P&P framework (Figures 9.7 and 9.8).

Differentiated • Pricing • Bundling

Untapped customer "head"

Not focused on any specific persona

Single Price Point

VS.

Untapped customer "long tail"

FEATURES/QUANITY (Perceived Subscriber Value)

Tier 1 PRICE (Willingness to pay)

PRICE (Willingness to pay)

Persona 1

Persona 2 Tier 2 Persona 3 Tier 3 FEATURES/QUANITY (Perceived Subscriber Value)

Figure 9.7  One-size-fits-all/Tiered packaging. C hapter

9 

10 5

G aurav S o n par a n d M ichael M a n sard 1.

DEFINE

2.

ITERATE

DESIGN

3.

ITERATE

DEPLOY

Figure 9.8  Define, design, deploy.

Introducing ABC Ltd

To help illustrate the 3D P&P framework, we use the case study of a fictional company called ABC Ltd. ABC offers workspace business solutions for SMBs, including emails, video conferencing, collaboration in real time, cloud-based file sharing, and storage, along with other applications. ABC Ltd has designed an innovative workspace collaboration platform. They want to set up an effective and differentiated pricing and packaging strategy using the GBB model. As we move forward, we will see how ABC builds and deploys its tiers. Define target audience, willingness to pay, and value metrics Define your target audience

It’s essential to figure out who your customers are—not who you assume they are. You obviously have a sense of who buys your products. But do you really understand how prices influence their purchasing decisions? In a subscription model, understanding customers means gaining visibility beyond the point of purchase. You need to see the subscriber journey from prospect to signup, what the drivers are for subscribing, and what barriers may prevent a subscriber from considering your offering. To get started, you’ll need to define your entire subscriber pool, that is, the addressable market. Then divide that market into smaller segments. Each of these segments will be represented by a subscriber persona, which will form the foundation for each pricing tier. We recommend the use of multiple sources of data to understand customers’ motivations, attitudes, behaviors, challenges, and desired outcomes. Qualitative research is a wonderful place to begin: customer interviews, focus groups, insights from sales or customer-facing teams, and ethnographic research are all excellent sources. Use this data as the building blocks to create customer descriptions, including expected (‘table stakes’) versus higher value (‘premium’) features and services. Once you’ve sketched out your customer personas, embellish them with quantitative research using market insights, web analytics (demographics, interests, buying patterns, etc.), and internal client data. It’s also important to explore the commercial viability of your customers according to core subscription model metrics like customer acquisition cost (CAC), lifetime value (LTV), and willingness to pay (WTP). We will do a deeper dive into WTP in a moment!

10 6 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

Based on the research and analysis, ABC Ltd identified three target subscriber personas: SME Bill, Mid-market Emma, and Corporate John.

ABC Ltd example

Persona 1 SME Bill

Persona 2 Mid-market Emma

Persona 3 Corporate John

About

SME Bill runs a small team on a tight budget. He needs solid functionality to keep his staff from getting overloaded without breaking the bank.

Mid-market Emma runs a department at a midsize company that is growing rapidly. She needs a solution that meets her needs now, but with the potential to seamlessly upgrade later.

Goals

Upgrade from his Manage rapid growth homegrown workspace system Needs professional-grade Needs to manage a larger software team

Corporate John works for a multinational enterprise dealing with increased security concerns and new acquisitions. He has been allocated a substantial budget and needs a top-ofthe-line solution. Looking for a turnkey solution

Motivations

Frustrations

Tight budget

Features and Must have preferences ■ Ad-free business email (Must have vs. Nice to ■ Calendar sharing have) ■ Video and voice conferencing ■

Messenger

Doc sharing and collaboration Nice to have ■

■ ■

Digital whiteboards Support advanced doc types

Optimal price $3.5 point Indifference $4.5 price point Range of $3–5 acceptable prices C hapter

9 

Does not want to have to start all over again if the company grows Must have ■

■

Phishing and spam protection for emails Cloud storage for highresolution docs and files

Needs top-notch security and a collaboration platform to support productive ways of working Does not have time for poor customer service Must have ■ ■

Secure emails Archive and search emails and docs

■ High video and Video conference with a sound quality large audience across the Nice to have globe ■ Unlimited storage Nice to have ■ Record meeting ■ Meeting recording participation ■ Invite guests in chat rooms ■

Enhanced security features $7.5

$12.5

$9

$15

$7–10

$12–15

■

10 7

G aurav S o n par a n d M ichael M a n sard Define subscribers’ willingness to pay

WTP is the maximum price a customer is willing to pay for a product or service. Typically, WTP represents a figure but in some cases it could be a range. Assessing WTP is key to driving value-based instead of cost-plus pricing because it incorporates the customer’s perception of the product or service. Many different models have been developed to assess WTP, including conjoint, Gabor– Granger, monadic, or Van Westendorp’s price sensitivity meter (PSM). We typically recommend PSM because it provides an effective, easy-to-use, data-driven, survey-based approach. It does not provide a precise price point, but it helps define an optimal range of acceptable price preferences based on how subscribers value the product or service. The PSM approach asks four price-related questions to determine the maximum amount that a consumer is willing to pay and how high the price must be set for subscribers to still value the product/service. 1. At what price is the product/service so expensive that you wouldn’t consider buying it? (Too expensive) 2. At what price is the product/service priced so low that you would feel the quality couldn’t be very good? (Too cheap) 3. At what price is the product/service starting to get expensive, but you’d still consider buying it? (Expensive/high side) 4. At what price is the product/service a bargain—a great buy for the money? (Cheap/ good value) (Figure 9.9) Expensive

Too expensive

Cheap

1. Indifference price point Normal price point

Too cheap

2. Point of marginal cheapness (PMC) Lower end of the range

3. Point of marginal expensiveness (PME) Upper end of the range 4. Optimal price point

Define value metrics

Peter Drucker famously wrote in his classic, The Practice of Management, ‘What gets measured, gets managed.’ But according to a 2021 survey by OpenView Partners, only two in five SaaS companies (39%) use a metrics-based approach to set prices. ‘The rest make a judgement call (27%), copy from competitors (24%), or take a cost-plus approach (10%)’ (Poyar, 2021). As a result, most SaaS offerings are not priced according to the value perceived by subscribers. This lack of visibility can adversely impact the overall revenue potential of your offering. The most direct way to understand the value of your product for subscribers is to somehow quantify how much the product helps their business. In other words, you need to link the perceived subscriber value with a unit of measurement. This metric should be a useful tool for both the company and the customer in terms of evaluating the effectiveness of the product. These value metrics should be easy for the customer to understand when they are perusing your pricing tiers. They should clearly communicate the value creation associated with the product or service and demonstrate how that value will grow with 10 8 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

%

Acceptable price range

Cheap

Too cheap

1.

Too expensive

3.

2.

PROPORTION OF RESPONDENTS

Expensive

4.

PRICE LEVEL

$

Figure 9.9  PSM graph.

customers’ usage over time. Effective metrics we’ve seen in our practice include number of active users (Slack, Monday); emails sent (MailChimp); storage used (AWS, Azure); or number of hosted domains (GoDaddy, Hostgator). Unit metrics can be functional, output-based, or outcome-based. Outcome-based metrics are the Holy Grail, but they are often impractical and hard to measure. Think of a SaaS procurement solution that enables customers to streamline and optimize expenses—an outcome-based metric could be ‘$ expenditures under management.’ The problem is that as your product successfully generates cost efficiencies, your company will end up capturing a smaller portion of the expenditure. Another option could be ‘$ saved expenditures,’ which seems attractive at first, but makes it extremely hard to assess the exact outcome given all potential external factors impacting savings. Both scenarios create potential asymmetry between you and the customer. As a result, we recommend value metrics that are SMART: specific, measurable, attainable, relevant, and time-bound. Output-based or functional metrics are usually aligned with SMART principles, as customers are accustomed to them, and they can be easily articulated. Outcome-based metrics should be reserved until the offerings—and the customers—mature. ABC Ltd chose ‘number of users’ as an easy-to-understand functional metric and a good proxy to assess value. They chose this metric instead of ‘cloud storage in gigabytes’ or ‘number of projects’ (output-based), or ‘employee productivity’ (outcome-based), C hapter

9 

10 9

G aurav S o n par a n d M ichael M a n sard

as these are difficult parameters to assess, for both the company and customers. The number of users shows that employees with a customer’s company are engaged with the product and find it useful, thus demonstrating value. However, ABC Ltd also identified certain output-based parameters, like the average number of participants per meeting and average cloud storage requirements, which are relevant metrics to create segmentation between tiers.

ABC Ltd example

Persona 1 SME Bill About

Goals Motivations

Frustrations Features and preferences (Must have vs. Nice to have)

Persona 3 Corporate John

SME Bill runs a Mid-market Emma runs Corporate John works small team on a department at a for a multinational a tight budget. midsize company that enterprise He needs solid is growing rapidly. She dealing with functionality to needs a solution that increased security keep his staff from meets her needs now, concerns and new getting overloaded, but with the potential acquisitions. He has without breaking to seamlessly upgrade been allocated a the bank. later. substantial budget and needs a top-ofthe-line solution. Upgrade from his Manage rapid growth Looking for a turnkey homegrown solution workspace system Needs professionalNeeds to manage a larger Needs top-notch grade software team security and a collaboration platform to support productive ways of working Tight budget Does not want to have to Does not have time start all over again if for poor customer the company grows service Must have Must have Must have ■

■ ■

■

Ad-free business email Calendar sharing

Messenger

Doc sharing and collaboration Nice to have ■

■

■

■

Video and voice conferencing

■

110 

Persona 2 Mid-market Emma

Digital whiteboards Support advanced doc types

Phishing and spam protection for emails Cloud storage for high-resolution docs and files

Video conference with a large audience across the globe Nice to have ■

■ ■

■

Meeting recording

■ ■

Secure emails Archive and search emails and docs

High video and sound quality Nice to have ■

■ ■

Unlimited storage Record meeting participation

Invite guests in chat rooms Enhanced security features C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est Optimal price point (Market entry or penetration) Indifference price point (Revenue maximization) Range of acceptable prices (Profit maximization or skimming) Value metrics

$3.5

$7.5

$12.5

$4.5

$9

$15

$3–5

$7–10

$12–15

# of users

# of users

# of users

Design tiers to address the needs of your customer segments

The subscriber personas, WTP, and value metrics become the foundation for the GBB tiers. Use these factors to build a perceived value matrix (see Figure 9.11). A feature with high perceived value but low WTP will be table stakes (good), whereas a highvalue feature with a high WTP will represent a differentiated offering (better/best) (Figure 9.10). On the other hand, a low-value feature with a high WTP represents a premium addon, which is an option that one can subscribe to, alongside one of the offering tiers. If these features were included in a tier, a small number of subscribers would extract too much value for too little price, while others would extract no value from them and would feel that they were being ripped off. Finally, low value and low WTP features are throwaways, suitable for a freemium or free trial offer. You also need to consider market strategy when evaluating features. You may value features differently based on whether you are pursuing market penetration, revenue growth, skimming, or customer retention (see the ‘Deploy’ section for more information on strategy-specific pricing). This will require some trade-offs. For example, if the objective is to achieve high growth, there will be a balance between high lifetime value due to high WTP and the cost of acquiring the persona. Execution is also important. A well-designed pricing page with a user-friendly layout and clearly laid out pricing is an opportunity for companies to drive their pricing and packaging narrative. Examples of good/better/best. Zoom offers optional add-on plans for audio conferencing, large meetings, premier support, and so forth (Figure 9.11). Company ABC Ltd identified three target personas based on surveys, valued features, and WTP. After ring-fencing the features within each of the three personas, a relative preference matrix emerged. Persona 1 (SME Bill) managing a small team, needs only basic table stake features (workspace essentials). C hapter

9 

111

G aurav S o n par a n d M ichael M a n sard

Figure 9.10 Design tiers tied to subscriber persona.

Persona 3 (Corporate John), managing multicountry, cross-team alignments, with a need for more stringent security rules, is ready to pay a premium (workspace premium) for enhanced security and attendance tracking. The workspace standard tier fulfills all the necessary requirements for Persona 2 (Mid-market Emma). Note the simple (self-explanatory) naming convention for each tier, which conveys a clear narrative and avoids any ambiguity or confusion. This along with the pop-out recommendations for the best value for money or the most popular tier 11 2 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

Figure 9.11 Optional add-on plans.

guides the subscriber to make informed decisions and helps generate more clickthroughs on that page. Workspace Essentials ($4.5/user/month) (Good)

Business Email and calendar

■

■

C hapter

Customer and secure business email

Workspace Standard ($8.5/user/month) (Better)

Workspace Premium ($13.5/user/month) (Best)

Everything in Essentials, plus

Everything in Standard, plus ■ Smart search

Shared calendars

9 

11 3

G aurav S o n par a n d M ichael M a n sard Video and voice conferencing

■

100 participants

■

150 participants

■

■

Digital whiteboarding

■

Noise cancellation

■

■

Meeting recording

■

Secure cloud storage Team messenger Document sharing and collaboration Security, management, and support

■ ■ ■ ■

50GB per user Support 100+ file types Team messenger Document, notes, slides

■

Polling and breakout rooms 1TB per user

■

Shared drive for teams Guest access Custom branding for documents and form templates

■ •

■

Writing assistance Two-step verification

■

TB per user

■

Endpoint management

■

Enhanced support

■

Standard support

■

■

■

■

500 participants Attendance tracking

5TB per user

Digital vault to retain, archive, and search data Advanced endpoint management

Let’s look at how ABC Ltd applied the price variability coefficient (see ‘Spotlight: GBB data-driven guardrails’ for more on this topic). ABC determined that the Essentials tier should be $6/user/month. The Standard and Premium tiers were $10 and $18, respectively. Based on the price variability coefficient, the good-best ratio falls within market standards, while the good-better is too low. They can decide to increase the price point for Standard to $12/user/month or assess whether their packaging needs further finetuning based on their strategy.

Deploy pricing strategy based on your organizational maturity and market positioning Deploy in-sync with your go-to-market strategy

Pricing and packaging is an iterative process. Deployment may differ based on market conditions and business maturities. A useful way to tailor the three GBB tiers is with a ratio that represents the anticipated revenue for each tier. Penetration or revenue monetization play (WIN). Minimize adoption friction of your offering to successfully acquire customers. For this play, the revenue ratio across GBB tiers may range from 50:40:10 to 70:20:10, as you offer more value-added features at a lower price point. Skimming or profit maximization play (GROW). Build customer stickiness and focus on land-and-expand strategy. Pivot the customer from comparison with competition mentality to incremental value across various tiers. For this play, the ideal GBB revenue ratio is 20:65:15. The objective is to maximize profit before competitors enter the market. Typically, price skimming applies to new,

114 

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

innovative products. With time, as the product becomes less novel and more accessible, the price steadily declines. Retention play (RETAIN). Minimize churn among existing customers, consolidate market position, and/or counter the threat from competition. In this play, the GBB ratio should be close to 35:60:5. Deploy an agile monetization approach for continuous price iteration

Going live with pricing tiers is the beginning, not the end, of the pricing process. As the market evolves over time, so does subscriber buying behavior. Hence, it’s essential to keep pace with the changing subscriber value perception over time. This necessitates an agile approach, with pricing catalog reviews on a continuous basis. As a best practice, market-leading SaaS companies typically review their tiers every 90 to 180 days. This does not mean a full revamp of the pricing tiers. It could mean adding new features, offering premium features for lower tiers (biphasic monetization; Kanellakis, 2021), or creating new tiers (Figure 9.12). Continue iterating—with your customers

The 3D P&P framework is a simple yet effective tool for navigating the sea changes that your business will encounter when moving to subscription pricing and packaging. Packaging needs to entice new customers but also encourage retention: the lifeblood of subscription businesses. Especially during the launch of new subscription offerings, this means that companies need flexibility to capture a wide range of target audiences while also building space for their subscribers to upgrade or downgrade down the line. When combined with a solid definition of your target customers, an understanding of what they value, and what they are willing to pay, you should be off to a solid start with GBB pricing. Obviously, there will be exceptions to the guardrails, and companies should always expect to have to iterate and refine their product and pricing propositions based on customer feedback after launch. And this is a key point: you need to always stay in beta with your pricing! Customers change, the market changes, and your company’s capabilities change. But that’s the strength of the subscription model: it gives your business the built-in flexibility to stay agile over time. If you maintain a laser focus on customers, your pricing and packaging should become a strong revenue driver for your business. Bios

Gaurav Sonpar is a Business Strategist at Zuora, a leading Subscription Economy evangelist. He advises clients on effective go-to-market motions, pricing, packaging, and ­customer-experience strategies. He has worked with leading organizations across multiple industry sectors to deliver high-impact strategies to launch new business models and accelerate business growth. He is also a member of Zuora’s Subscribed Institute and collaborates with leading academic institutions and industry peers to conduct both qualitative and quantitative research on business transformation and maturity in the subscription economy. Prior to Zuora, Gaurav spent 14 years as a Strategy and

C hapter

9 

11 5

116 

PERCEIVED VALUE

PAST

Differentiated

Throwaway (Limited or One-TimeUse)

Table Stakes (Basic Featur+es)

Table Stakes (Basic Features)

Throwaway (Limited or One-Tim e Use)

Table Stakes

Enhancer (LimDitiefdfeorrentiated One-Time U seF)eatures) (Basic Features) (Add-On or (Premium Optional Features)

PAST

Enhan+cer

Optional Features) Throwaway

+

+

FUTURE

TIME DIMENSION

Differentiated Better/Best (Premium Features)

$

$

$ Better/Best

Add-Ons (Add-On or Better/Be(Psrtemium Features)

WILLINGNESS TO PAY

-

Enhancer

(Add-On or Optional Features)

Add-Ons

Add-Ons

+

-

+

FUTURE

PERCEIVED VALUE

WILLINGNESS TO PAY

$

Figure 9.12 GBB value matrix (value vs. WTP vs. time.

-

+

TIME DIMENSION

T0

T 0+1

T 0+N

G aurav S o n par a n d M ichael M a n sard

Good

Good

Good

C hapter

9

S ubscriber P sych o l o gy with G o o d / B etter / B est

Transformation Consultant with a range of professional services companies specializing in IT services and consulting such as Accenture, Cognizant, and Tata Consultancy Services. Gaurav holds a Bachelor of Engineering degree and an MBA from Lancaster University. Michael Mansard is a seasoned Subscription Economy business strategist. Since he joined Zuora in 2015, he has been accompanying more than 300 companies globally and across industries in their business model transformation. Those companies range from fast-growing startups (Doctolib, Treated​.c​om) to large enterprises (Philips, Siemens, Assa Abloy, Michelin, St Gobain, Ubisoft, Renault-Nissan). Leveraging his 14-year experience at Deloitte Consulting, SAP, and as a startup mentor, he has developed an original multidisciplinary profile in sales and marketing, finance, and IT. He currently serves as Principal Director, Business Transformation and Subscription Strategy within Zuora’s Chief Revenue Officer’s group. Michael is also the Subscribed Institute’s EMEA Chair. As such, he recently authored several thought leadership pieces—such as ‘Industry 4.0: An Executive Playbook for Business Model Transformation,’ ‘Subscription Economy Maturity Model,’ and ‘A New Formula for Growth: The Financial Services Industry and the Subscription Economy.’ He is also the co-creator of INSEAD’s Subscription Business Bootcamp, an elective for Global Executive MBAs as well as an executive education program. As an investor and advisor, Michael is working very closely with eight hypergrowth startups, in the fields of FinTech, FoodTech, MarTech, MedTech, and PropTech— which, in common, are all subscription-based businesses.

Key objectives 1. Inform on the main differences of a subscription business model. 2. Explain the best practice to develop a good/better/best model. 3. Propose the primary research methods that can be used to develop such models.

Key summary points 1. Subscription business models and offers flip the traditional transactional model upside down. You must put the subscriber at the center of the business model and manage the offer dynamically as the subscriber needs to evolve. 2. GBB is the most used packaging method in subscription. It provides subscribers with the best choices and makes it easier to choose offers. However, it needs to be designed based on science and best practices and not on gut feeling. 3. Good packaging and pricing practices go together. Research must fill the process with good customer insights. That is the start of a good packaging strategy. Then the pricing and monetization strategy is all-encompassing with packaging at its heart.

C hapter

9 

117

G aurav S o n par a n d M ichael M a n sard

Key questions 1. How do you get customer insights to develop packaging models? 2. How does GBB compare with other packaging options such as open bundles or functional bundles? 3. Why are most SaaS companies not leveraging the full power pricing and packaging science?

References Global Banking & Finance Review. (n.d.). How to make money from membership economics. https://www​ . glo​ b alb ​ a nki ​ n gan​ d finance ​ . com ​ / how​ - to ​ - make ​ - money​ - from ​ - membership​ -economics/. Kanellakis, S. (2021). What belongs in your basic bundle? Harvard Business Review, July 20. https://hbr​.org ​/2021​/07​/what​-belongs​-in​-your​-basic​-bundle. Mailchimp. (n.d.). Compare marketing plans. https://mailchimp​.com ​/pricing ​/marketing ​/ compare​-plans/. Mohammed, R. (2018). The good-better-best approach to pricing. Harvard Business Review, September–October. https://hbr​.org ​/2018​/09​/the​-good​-better​-best​-approach​-to​-pricing. Poundstone, W. (2011). Priceless: The myth of fair value (and how to take advantage of it). Hill and Wang. Poyar, K. (2021, January 19). Pricing insights from 2,200 SaaS companies. https:// openviewpartners​.com ​/ blog ​/saas​-pricing​-insights/#​.YikTLkDMLIV. Wilton, J., & Khanna, M. (2018, December 3). Pricing pitfalls: The four most common packaging mistakes. McKinsey & Company. https://get​.fuelbymckinsey​.com ​/article ​/pricing​ -pitfalls​-the​-four​-most​- common​-packaging​-mistakes/. Zuora. (2022, February 16). Zuora subscription economy index finds subscription businesses have grown 4.6× faster than the S&P 500 in the last decade, enduring beyond pandemic surge. https://www​.zuora​.com​/press​-release​/zuora​-subscription​- economy​-index​-finds​-subscription​ -businesses​-have​- grown-4–6x-fa​ster-​t han-​t he-s​p -500 ​-in-t​he-la​st-de​cade-​endur​i ng-b​eyond​ -pand​emic-​surge​/.

118 

C hapter

9

C hapt e r 1 0

Value-Based Pricing of SmartProduct-Service Offerings in the Manufacturing Industry Tobias Leiting, Calvin Rix, Regina Schrank, and Lennard Holst

Introduction

Digitalization offers the potential to increase the flexibility and productivity of production processes and the quality and performance of products and services in the manufacturing industry to a new level, thereby contributing significantly to the gross value added and competitiveness of manufacturing companies (Kohtamäki et al., 2020; Zheng et al., 2019). This can inspire manufacturing companies around the globe to optimize their production processes, products, and services, thus strengthening their international competitiveness (Zheng et al., 2019). Besides the optimization of internal value-creation processes, a real-time-based connection between two companies also enables new types of smart-product-service offerings with unique value propositions for external value-creation processes (Illner et al., 2019). For this purpose, manufacturers of machinery and equipment (suppliers) are expanding their offerings for producing companies (customers) by adding smart-product-service offerings to their existing product and service portfolios. Examples of these offerings are applications for the data-based visualization of machine status (e.g., dashboards) or analysis systems for predicting breakdowns (e.g., condition monitoring). These offerings are no longer sold to the customer as a one-off but are instead offered through continuous service integration by subscription (Mansard & Cagin, 2019). Despite the predicted potential and great euphoria, revenues from these new smartproduct-service offerings have so far fallen short of expectations. Successes are evident at best in improved processes but hardly in new, revenue-relevant business areas (Accenture, 2020). One reason for this is that the established manufacturing industry is strongly focused on the product-oriented business around physical products and services. In these businesses, the cost-oriented pricing approach dominates, in which the costs of providing services are relayed with a markup to the customer (Frohmann, 2018). However, this model does not align with smart-product-service offerings, as the cost structure is different because of high fixed development costs and low scaling costs. Moreover, subscription business models lead to stronger customer integration DOI: 10.4324/9781003226192-13

11 9

T o bias L eiti n g et al .

and cooperation (Tzuo & Weisert, 2018). The focus of value creation shifts from inhouse production toward the usage phase of the smart-product-service offering to the customer. These require a more value-based pricing approach (Liozu & Ulaga, 2018). This leads to novel potentials and challenges in pricing. ■

■

■

■

Selling value propositions. Customers no longer buy tangible physical products or services but instead the added value provided by the smart-product-service offering during the usage phase. Delivering an individually fitting value proposition requires an in-depth knowledge of a customer’s value-creation activities (Stoppel & Roth, 2017). Data-driven value quantification. Because the customer value from the provision of a smart-product-service is individual, the quantification of the value is a challenge. This requires measurable indicators enabling objective evaluation of the customer value based on the data collected (Macdonald et al., 2016). Design of value-based pricing models. Access to the customer’s operating data opens the use of innovative value-based pricing models that refer to customer data. Thus, new pricing models offer the potential for participative sharing of the additional value created. This can ensure a long-term and collaborative partnership between customers and suppliers (Frohmann, 2018). Definition of the subscription-based price metric. The offer is charged on a recurring subscription basis with periodic multidimensional flat and variable fees instead of a fixed one-off payment. The price points of the price components must be defined such that the added value from the smart-product-service offering is shared between customer and supplier (Mansard & Cagin, 2019).

Framework for pricing smart-product-service offerings

The first step toward achieving successful pricing for smart-product-service offerings is to provide a customer with a tailored offering and then harmonize this with the pricing model (Liozu & Ulaga, 2018). For this purpose, we introduce a systematic framework that structures the decision variables and recommendations for actions in elements (Figure 10.1). The results based on this framework were developed and applied in an innovation project involving more than 30 case studies from the manufacturing industry (Center Smart Services, 2022). Step 1 is the smart-product-service-system design, in which four possible archetypes with different pricing features are structured to determine how the smart-productservice is offered. This is followed by step 2, determining benefits and value. Based on individual requirements and characteristics, a selection can be made between value determination methods that can be used either in the usage phase or before providing the solution. The price model design in step 3 aims at selecting and designing the appropriate pricing model. This leads to a complex decision-making situation between various price models exploiting the possible price potential and assuming risks. Step 4 details the selected price models according to the pricing metric design. This design considers the characteristics and requirements of the different price components and points. Furthermore, the payment intervals and the contract duration are defined based on recommendations. 120

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g

1

Smart-Product-Service -System Design

Smart-product-service archetypes

2

Value Determination

Value determination before and during usage

3

Price Model Design

Fixed-fee and variable price models

4

Price Metric Design

Price components and amounts

Figure 10.1 Framework for pricing smart-product-service offerings. Smart-product-service-system design

For the effective design of the smart-product-service system, a spectrum of possible offering archetypes is needed. Based on a case study analysis (Center Smart Services, 2022), four different archetypes of smart-product-service offerings can be identified, which differ in terms of the reference basis as well as the goals for the expansion of the digital business. The archetypes range from ‘data product’ to ‘smart product’ to ‘digital product’ to ‘X-as-a-Service’ offerings (Figure 10.2) (Krotova, 2020). Interviews in the case studies indicate that the customer value from the smart-product-service offering increases from data product to X-as-a-Service offerings. The first archetype of smart-product-service offerings, ‘data product,’ consists of valuable data obtained within the production or from the production ecosystem. The raw data are aggregated and monetized as a data product through analytics services for external customers. The data have an information value for customers, which results on the one hand from the data quality and on the other hand from the degree of the resulting analytical maturity (Amann et al., 2020). Smart products are offerings of a connected physical primary product with digital and software-based features. The digital features create differentiation factors for the product with new functions, enable real-time connectivity to the customer, and provide access to valuable data from the customer’s production environment. Digital add-ons are increasingly becoming a basic requirement in the manufacturing industry. Hence, smart product enhancement primarily aims to secure revenue for the sold products, which is the reference basis for the price. In addition, smart products create the infrastructure to offer further smart-product-service offerings (Lünnemann et al., 2019). Digital products are independent, digital service offerings with an independent value proposition for the customer. Digital products represent digitally saved information in an unbound form that is offered to customers on-premises or via online systems. The services are purchased because of their function, as distinct from a data product. A strength is that digital products can process data and interact with smart products. They offer the potential to scale rapidly through digital platforms. Therefore, digital products that provide added value for many customers provide high market potential. The customer is mainly interested in the application of the added-value software, not in the underlying program code. Furthermore, the provision of a digital product is not necessarily subject to the principles of industrial services, for example, the integration of an external factor (Kampker et al., 2018; Stich et al., 2019). X-as-a-Service offerings are integrated solutions consisting of smart products, industrial services, and digital products that are individually tailored to solve a customer’s problem and integrated into the working environment of the customer as a holistic

C hapter

1 0 

1 21

Reference basis

Monetizing valuable data

Offer of data as a product

Data product

Differentiator & collecting data

Product

Digital Product

Offerof a product with the addition of digital elements

Smart product

Figure 10.2 Archetypes for the pricing of smart-product-service offerings.

Goals

122 Digital product

Establish an independent digital business

Product

Digital Product

Stand. alone offer of a digial product

X-as-a-service

Solution-oriented value proposition through integrated, data-driven service offerings

Product

Service

Digital Product

Integrated solution of product, service and digital product

T o bias L eiti n g et al .

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g

system. A customer no longer buys specific products or services but instead specifies their requirements. The provider configures a solution that is geared to these requirements. At the heart of this is the interaction between provider and customer in the usage phase and the associated understanding of value, which emphasizes ‘value-inuse’ (Grönroos & Helle, 2010; Lah & Wood, 2016; Mansard & Cagin, 2019).

Value determination

A deep understanding, also called ‘customer intimacy,’ is an important prerequisite for successful data monetization (Govindarajan, 2020; Liozu & Ulaga, 2018). A provider must understand the business model of the customer, meaning how they earn money, to make the customer’s business more effective and efficient. Therefore, value determination requires that the provider also understands the ‘why’ and the ‘how’ of the value proposition for the customer and how this can be measured using key performance indicators (Schuh et al., 2019). Calculating the specific added value from a smart-product-service offering represents the most relevant success factor in price enforcement toward the customer (Simon & Fassnacht, 2018). Depending on the situation, the type of smart-product-service offered, and individual requirements, a distinction can be made between different types of value determination (Figure 10.3). Before the usage phase, the supplier can choose between a standardized and an individual value determination. The standardized value assessment can be carried out based on a pilot customer value calculation. For this purpose, a smart-product-service is implemented for an indicative pilot customer. Additionally, key performance indicators in combination with interviews are used to measure the added value for the pilot customer. The emerging value serves as a reference point for all other customers with similar characteristics. A benefit is that the assessment is simple and inexpensive to perform and at the same time highly scalable. However, it is not necessarily precise, because the realized or perceived value for different customers can vary significantly (Macdonald et al., 2016). In addition, the resulting differences in willingness to pay cannot be exploited by different value-based price characteristics. The individual value assessment differentiates individual customers in terms of different properties based on a catalog of different customer attributes like risk, value proposition, or customer size. This allows the exploitation of different levels of willingness to pay but also complicates the process and requires specific expertise for the respective context. In both methods before use, however, the determination of value only represents a prediction prior to use. That means that there can be a discrepancy between the realized value-inuse for the customer and the predicted value due to missing feedback loops. The measurement of the realized value during the usage phase (value-in-use) is superior to the value prognosis because of the direct feedback to the value-creation activities. However, it requires simultaneously more effort (Simon & Fassnacht, 2018). Accordingly, value measurement enables individual price differentiation based on the value to the customer. A qualitative evaluation of the value-in-use can be achieved with a conjoint analysis—by a survey-based assessment of the perceived value of product features and performance. However, this approach is subjective and susceptible to manipulation through individual bias of surveyed actors (Klarmann et al., 2011). By contrast, the data-based value calculation calculates key performance indicators

C hapter

1 0 

123

124

During use

Before use

Value measurement

Value assessment

Figure 10.3 The procedures of value assessment and value measurement.

Value-in-use

Time of value determination

Value proposition

Quantitative

Qualitative

Individual

Standardized

Differentiating Elaborate Knowledge required Prognosis Differentiating Elaborate Knowledge required Susceptible to manipulation Differentiating Automatable Knowledge required Data required

Conjoint analysis

Data-based value calculation

Non-differentiating Simple Scalable Prognosis

Individual value assessment

Pilot customer value calculation

T o bias L eiti n g et al .

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g

with recorded data of the customer’s production. These are aggregated automatically through real-time data exchange. Since this measurement has the lowest strategic and hypothetical bias risks, it is considered superior to X-as-a-Service approaches. Problems exist in the practical implementation, as the necessary knowledge to anchor and scale the data basis for pricing is often missing (Klarmann et al., 2011).

Price model design

The price model represents a logic between the smart-product-service offering and a monetary amount paid through a defined set of price components. For the price model design, a distinction can be made between product-centered and customer-centered price models. These differ in terms of the quantifiable reference variables, the value proposition, and the orientation of the value perspective (Figure 10.4). In product-centered models, the focus is mainly on the features and functionalities of the product or service, so the actual usage phase is not considered. The price model is a one-time transaction. The reference is geared almost exclusively to the costs of the provider, and, apart from the conventional product warranty, no further risks are assumed by the supplier. These models are established within manufacturing industries, but this model is just oriented toward the bottom line of a price interval because of the cost focus. A sophisticated evaluation of willingness to pay is not required (Krotova et al., 2019). In addition to the product-centered approach, four customer-oriented price models can be distinguished with a focus on customer value (Stoppel & Roth, 2017). The availability-oriented pricing model (e.g., a flat rate) is the least risky for the supplier. The supplier guarantees continuous availability of a smart-product-service offer for a fixed fee and is financially responsible for the development and implementation of the offering. Furthermore, the provider takes over the costs for potential repair and maintenance services, such as in the event of a failure. The availability-oriented model is uncomplicated, needs no direct data linkage to the operation data of the customer, and is easy to scale. However, the supplier has no direct incentive to improve performance through the pricing model. The next pricing model is the use-oriented model. The intensity of the use of a service is the basis for assessing the price (e.g., pay-per-use). As a result, the success of the supplier depends on the usage behavior of the customer. This means that market, process, and capacity risks are also taken over by the supplier: for example, if the demand for the products of the customer is low or due to inefficient and failure-prone processes. The advantage of this model is the direct link between customer success and price, which creates incentives for the high performance of the customer through the smart-product-service. Moreover, the data for the reference variable (e.g., running time of a machine) are rather standardizable and clearly recordable across many customers. To prevent the customer from perceiving harm from high usage, the model can be also combined with, for example, usage tier levels or reduced prices for increased usage. In addition, a minimum purchase volume can reduce the risk for the supplier. The result-oriented price model (e.g., pay-per-part) offers the potential to exploit even higher added-value potentials than the previous models, but risks about the efficiency and effectiveness of production are also assumed. These models are recommended, especially if the supplier has a suitably positive influence on these risks through a high

C hapter

1 0 

125

126

Product-centered

Intensity of use (e.g. frequency of use, time of use) Outcome (e.g. number of units produced, number of usable units) Economic variable (e.g. cost or profit improvement)

Use-oriented (e.g. pay-per-use)

Result-oriented (e.g. pay-per-output)

Success-oriented (e.g. pay-per-success)

+ + + +

Availability level (e.g.: compliance with uptime/h, compliance with tolerances)

Availability-oriented (e.g. flatrate)

+ Value risk + Profit increase + Improvement

+ Efficiency + Effectiveness

+ Market risk + Process risk + Capacity risk

Solution offer Investment risk Availability risk Quality risk

+ Product/Service + Digital additional service (stand-alone/indirectly via primary product)

Cost-Plus (Effort of the provider (labor, material, production assets)

Value proposition

One-time payment (e.g. transactional purchase)

Quantifiable reference variable

Figure 10.4 Overview of different price models (adapted from Stoppel & Roth, 2017).

Customer-centered

Price model

Customer value

Supplier costs

Orientation

T o bias L eiti n g et al .

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g

level of integration in the value creation of the customer: for example, in the case that X-as-a-Service solutions are offered. The advantage of this model is the great alignment of interests of customer and provider on an optimal production result. This was not yet prevalent in the previous models. Suppliers and customers benefit from optimized value creation, which is achieved through high customer performance. However, for the reference variable, this model requires comprehensive access of the supplier to the production data of the customer, and heterogeneous customer products require, in some cases, individual definitions as price variables. The success-oriented price models take the improvement of business indicators of a customer as reference variables (e.g., pay-per-cost-decrease or pay-per-profit-increase). Monetary improvements of a customer by smart-product-service offerings are shared directly between the two parties. In this way, a supplier is linked more closely to the success of the customer and, in the worst case, might end up with no monetary payout if the customer’s objectives are not achieved. In addition, the model requires a high level of transparency about a customer’s key business data by the provider. Typically, provider and customer need to define a baseline for measuring improvement. The advantages of the model, analogous to the previous variant, are the high-value-creation potential through optimized customer performance and strong customer loyalty through successful implementation of the smart-product-service.

Price metric design

Within the price metric design, the operative formula for the price calculation for the provided smart-product-service is derived. In the manufacturing industry, a price formula often consists of several price components because of the complexity of the smartproduct-service offering. A differentiation between three types of price components can be made: a fixed one-off payment, a fixed subscription fee, and a variable subscription fee (Frohmann, 2018). The diverse price components differ on several elements (Figure 10.5). For the one-off payment (which is often cost-oriented), there is a fixed amount of money ($) to be paid before the service is provided. The payment often serves to cover high costs and risks, especially if expensive products are provided within an X-as-a-Service offering to the customer. For both types of subscription fees, the payment intervals must be defined first. They may range from weeks to months to years. Shorter intervals lead to greater flexibility but also to increasing processing efforts. For the fixed subscription fee, a reference unit must also be selected to which the fee relates (e.g., user accounts, measurement points, or several machines). For this unit, an amount to be paid ($f) needs to be determined. This results in periodic fixed payment amounts for a constant number of reference units. This price component relates to the availability-oriented price model and often serves as a single price component for a digital product and as a payment baseline for complex solution-oriented offerings. The variable subscription fee is linked to the use-, result-, and success-oriented pricing models. This component is linked in each interval to a varying value of a datadriven reference base (e.g., production hours or produced parts) (x n). For each unit of this data-based reference variable, an amount ($v) to be paid must be determined. A unit amount is to be selected in such a way that the added value is divided between the customer and the supplier analogous to the fixed fee. Variable subscription fees

C hapter

1 0 

127

128

Price components

t

$f

$f

$f

$v*X1

t $v*X2

$v*Xn

Synchronous charging of the service (after subscription interval)

t

Synchronous charging of the service (before or after a subscription interval)

$

Figure 10.5 Price components for smart-product-service offerings.

Variable subscription fee

Fixed subscription fee

One-off payment

Advanced charging of the service

Time

Time

Time

T o bias L eiti n g et al .

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g

are often implemented as a multidimensional model in combination with a fixed fee (Frohmann, 2018).

Four pricing patterns for smart-product-service offerings

1. Data product. The data product can be bundled with additional functions and services into an integrated product that exceeds the actual content of the data product. Additional functions such as a dashboard can help to sort, understand, and better present data. The value assessment for the price should be based on the potential added value by the knowledge gained for business decisions of the customer. Even if compliance with the described quality criteria is obligatory for the sale of a data product, the added value for the customer is characterized by the value-determining attributes (Wells & Chiang, 2017). The evaluation is primarily based on the respective increase in value for the customer, which is reflected in time, labor, or money savings; a higher ROI; or reduced risk. The price metric is either a one-time payment or a recurring fixed-fee payment (e.g., for the continuous updating of data). One example of an offering of a data product is the Telematics Data API provided by Krone Trailer GmbH. This offering provides the customer access to their telematics data and information within their trailer fleet via an API. Customers can choose between three package levels: basic, premium, and premium dialogue, which allow access to different types of data. Thereby, they address different customer demands and value levels. With the premium data package, further functions such as temperatures are analyzed in addition to the GPS localization of the trailer. A customer pays a fixed subscription fee per month for the availability-oriented model per trailer. The customer therefore gets access to the processed sensor data, enabling them to obtain transparency about their own trailer fleet via a dashboard accessible on the telematics portal or in their own system (Krone, 2022). 2. Smart product. The smart product is typically sold to customers through a one-off payment, and the additional value of the digital features is priced indirectly into the price of the product. Since smart products enable a lock-in effect and data access as well as the possibility to offer additional services such as data products or digital products, suppliers often decide to provide these components to customers for a low cost or for free. Nevertheless, the additional functions should be clearly listed and the added value communicated to the customers and, if possible, quantified. Attention should be paid to ensure that customers do not become accustomed to cost-free digital services and do not attach any value to them. An example from the B2B industry is Atlas Copco with the provision of a smart torque driver for production lines. From the smart product features, the customer receives the opportunity for intelligent tightening strategies and the realization of higher process control. The prime purpose of this component is to improve the product’s functionalities and therefore provide differentiating features. The value for the customer is higher uptime, a reduction in defects, flexibility in production, and higher productivity compared with similar offerings. Therefore, premium prices can be enforced for a one-off payment price metric (Atlas Copco, 2022). 3. Digital product. Digital products are offered either individually or in bundles to a customer. The product portfolio should be as simple and standardized as possible C hapter

1 0 

129

T o bias L eiti n g et al .

for high scaling potential. To address different willingness to pay and customer needs, versioning with modular features can be used. Clear modules of digital products that are aligned with customer requirements allow price differentiation based on performance factors for the customer. The added value of a digital product can be assessed by pilot customers. To achieve a quick scaling, moderate amounts of added value should be skimmed from the customers. To enable the scalability of digital products through platform sales, a fixed subscription fee and a simple and clear reference unit within an availability-oriented price model are recommended. One example of a digital product is the Alsense Cloud Services from Danfoss. These digital products are offered through an IoT cloud. To address different customer requirements, the digital products are offered in four different version levels with a range of functions that increase in value from level to level. The lowest-priced ‘transparency package’ offers the function to visualize the measurement data. The ‘stabilization package’ builds on this, with a status that can be monitored and alarms that can be set. Subsequently is the ‘optimization package’ for adjusting controlled variables automatically and benchmarks. The ‘Best-in-Class package’ also includes a permanent consulting service provided by experts. The offered digital product portfolio creates new value propositions for a customer of supermarket refrigeration equipment in terms of operational efficiency, energy efficiency, and sales performance. The monetary value (loss of refrigeration, food waste, energy savings) was identified for a supermarket as a pilot customer. The digital products are each offered at a flat rate per supermarket within a monthly fixed fee (Danfoss, 2022). 4. X-as-a-Service. X-as-a-Service offerings are individual solutions tailored to customers’ requirements with customized pricing. To provide the solution effectively, the solution and pricing need to be customized based on modular components that are configured to the requirements. In principle, all customer-oriented price models are feasible. In the selection, the potential for value creation for customers and provider risks are important factors. Whereas the availability- and usage-oriented models are more suitable for service offers with little potential for improvement within the usage phase, the resultand success-oriented price models create great incentives for continuous improvement of the solution offered. Simultaneously, the level of risk assumed also increases. In general, the customer’s business model should be analyzed before the solution is offered to determine whether it is suitable for the intended pricing model. An indicator here is the growth ambitions and opportunities of a performance increase for the potential customer. If it is not economically viable for the supplier to assume the necessary risks, a corresponding pricing model should not be chosen. Furthermore, the price metric of an X-as-a-Service offering often consists of a combination of multiple price components, such as fixed and variable subscription fees. One example of such a service offering is the Heidelberg subscription offered by Heidelberger Druckmaschinen. In this pay-per-outcome model, a printing press, services, consumables, the IoT system Prinect, and consulting and trainings are provided to the customer as an integrated solution via a subscription contract as a print-asa-service offering. For this purpose, Heidelberger signs a five-year contract with the customer. The contract consists of a result-oriented price model with two price components. As a fixed subscription fee, an individually agreed minimum quantity of charged 130

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g

prints is defined, often by considering previous printing performance. Based on this as a variable fee, a pay-per-print scheme is applied, whereby the customer is charged a certain price for each printed sheet when exceeding the minimum quantity (Heidelberger Druckmaschinen, 2022).

Conclusions

In manufacturing, the ambitious monetization aspirations of smart-product-service offerings often fall short of the high expectations. One problem is that the pricing in the industry is not customer-oriented but product-oriented. Thus, the potential of digitalization is not fully exploited. In the context of pricing, four central challenges hinder the establishment of customer-oriented data-driven product-service offerings: (1) selling value propositions; (2) data-driven quantification of value; (3) design of value-driven pricing models; and (4) definition of subscription-based price metrics. To structure the pricing for smart-product-service offerings promisingly, a framework with four specific elements has been developed. To address the value propositions properly, this chapter presented four archetypes for offering smart-product-service systems. In order to quantify customer value, methods for a prior assessment and a value-in-use measurement are pointed out. Furthermore, the design and selection of the appropriate pricing models are explained based on the price potential and the risk assumption capability. As a last step of the framework, the definition of the pricing metric is discussed, highlighting price components. For practical application, recommendations for action are given within four patterns on the basis of case studies from practice. These enable a clear and user-friendly abstraction of value-based pricing by abstracting these patterns for any smart-product-service offering in the manufacturing industry.

Bio

Tobias Leiting, MSc, Calvin Rix, MSc, Regina Schrank, MLitt, and Dr.-Ing. Lennard Holst are project managers in the service management department at the Institute for Industrial Management (FIR) at RWTH Aachen University. Their research and consultancy activities focus on the transformation of manufacturing companies from producers to suppliers of smart-product-service systems by enabling companies and business units to design, market, and efficiently deliver offerings for their external and internal customers. For further information, please visit www​.fir​.rwth​-aachen​.de​/en/ or contact Lennard​.Holst​@ fir​.rwth​-aache​n​.de.

Key objectives 1. Introduce a framework for addressing the challenges of value-based pricing for smartproduct-service offerings in the manufacturing industry. 2. Highlight the possible degrees of freedom for designing tailored pricing for smart-product-service offerings in the manufacturing industry. 3. Provide practical, case-based pricing patterns for implementing value-based pricing in the manufacturing industry.

C hapter

1 0 

1 31

T o bias L eiti n g et al .

Key summary points 1. 2. 3. 4. 5.

Design smart-product-service offerings based on the classification of four archetypes. Highlight the difference between value determination before and during usage phase. Empower value-based price models. Define the right price metrics and transferring these into contracts. Understand how the smart-product-service offerings can be priced in practice in the manufacturing industry.

Key questions 1. What are the characteristics of the four archetypes of smart-product-services in the manufacturing industry? 2. How can the customer value from a smart-product-service offering be evaluated? 3. What are value-based price model types and what are their reference variables?

References Accenture. (2020). Top500-Studie Deutschland Weltmarktführer von morgen. https:// www​. accenture​.com/​_ acnmedia​/ PDF​-115​/ Top500​- Studie​- Deutschland​-Weltmarktf​%C3​ %BChrer​-von​-morgen​.pdf. Amann, K., Petzold, J., & Westerkamp, M. (Eds.). (2020). Management und Controlling: Instrumente—Organisation—Ziele—Digitalisierung (3rd ed.). Springer. https://link​ .springer​.com ​/content ​/pdf ​/10​.1007​%2F978​-3​- 658​-28795​-5​.pdf; https://doi​.org​/10​.1007​ /978​-3​- 658​-28795​-5. Atlas Copco. (2022). MicroTorque transducerized smart electric screwdriver. https:// www​. atlascopco​ . com ​ /en ​ - us ​ / itba ​ / industry​ - solutions ​ /electronics0 ​ / low​ - torque ​ - handheld​ -transducerized​-screwdriver. Center Smart Services. (2022). Pricing of digital products: How to build a value-based pricing model. https://center​-smart​-services​.com ​/en ​/research ​/pricing​-of​-digital​-products/. Danfoss. (2022). Alsense IoT Cloud-Plattform Kontakt. Frohmann, F. (2018). Digitales pricing: Strategische Preisbildung in der digitalen Wirtschaft mit dem 3-Level-Modell. Springer Gabler. https://doi​.org​/10​.1007​/978​-3​- 658​-22573​-5. Govindarajan, V. (2020). Who will win the industrial internet? Industrial incumbents and digital natives both have a chance. In HHBR insights series: Monopolies and tech giants (pp. 87–92). Harvard Business Review Press. Grönroos, C., & Helle, P. (2010). Adopting a service logic in manufacturing. Journal of Service Management, 21(5), 564–590. https://doi​.org​/10​.1108​/09564231011079057. Heidelberger Druckmaschinen. (2022). Subscription smart und subscription plus. https://www​ .heidelberg​.com ​/global ​/de ​/services ​_ and ​_ consumables ​/print ​_ site ​_ contracts ​_1 ​/subscription​ _agreements​/subscription ​_1​.jsp. Illner, B., Konjusic, R., Lässig, R., Lorenz, M., & Petzke, A. (2019). Leitfaden zur Kommerzialisierung von digitalen Produkten und Services. Frankfurt am Main. https://

132

C hapter

1 0

Value - B ased P rici n g i n S mart M a n ufacturi n g media​-publications​.bcg​.com ​/ Digital​- Mechnical​- Engineering​- German​- Report​- May​-2019​ .pdf. Kampker, A., Husmann, M., Harland, T., Jussen, P., & Steinbauer, M. (2018). Six principles for successful data-driven service innovation in industrial companies. In 2018 IEEE international conference on engineering, technology and innovation. Klarmann, M., Miller, K., & Hofstetter, R. (2011). Methoden der Preisfindung auf B2B-Märkten. In C. Homburg (Ed.), Preismanagement auf Business-to-Business Märkten: Preisstrategie— Preisbestimmung—Preisdurchsetzung (1st ed., pp. 155–178). Gabler. Kohtamäki, M., Praida, V., Patel, P., & Gebauer, H. (2020). The relationship between digitalization and servitization: The role of servitization in capturing the financial potential of digitalization. Technological Forecasting and Social Change, 151(February), Article 119804. Krone. (2022). Datenschnittstellen (API). https://www​.krone​-trailer​.com ​/service ​/ krone​ -telematics​/datenschnittstellen​-api/. Krotova, A. (2020). Der Weg zu datengetriebenen Geschäftsmodellen: Eine modellbasierte Analyse. Köln. https://www​.iwkoeln​.de​/studien ​/manuel​-fritsch​-alevtina​-krotova​-der​-weg​-zu​ -datengetriebenen​-geschaeftsmodellen​.html. Krotova, A., Rusche, C., & Spiekermann, M. (2019). Die ökonomische Bewertung von Daten: Verfahren, Beispiele und Anwendungen. IW-Analysen: Vol. 129. Institut der deutschen Wirtschaft. Lah, T., & Wood, J. B. (2016). Technology-as-a-service playbook: How to grow a profitable subscription business. Point B Inc. http://gbv​.eblib​.com​/patron​/ FullRecord​.aspx​?p​= 4528279. Liozu, S. M., & Ulaga, W. (2018). Monetizing data: A practical roadmap for framing, pricing & selling your B2B digital offers. Value Innoruption Advisors Publishing. Lünnemann, P., Wang, W. M., & Lindow, K. (2019). Smart industrial products: Smarte Produkte und ihr Einfluss auf Geschäftsmodelle, Zusammenarbeit, Portfolios und Infrastrukturen. Studie Frauenhofer IPK. Macdonald, E. K., Kleinaltenkamp, M., & Wilson, H. N. (2016). How business customers judge solutions: Solution quality and value in use. Journal of Marketing, 80(3), 96–120. https://doi​ .org​/10​.1509​/jm​.15​.0109. Mansard, M., & Cagin, J.-M. (2019). Reaping the recurring benefits of industry 4.0: A manufacturing executive playbook for business model transformation towards new revenue streams. Zuora. Schuh, G., Frank, J., Jussen, P., Rix, C., & Harland, T. (2019). Monetizing industry 4.0: Design principles for subscription business in the manufacturing industry. In 2019 IEEE international conference on engineering, technology and innovation (ICE/ITMC) (pp. 1–9). https://doi​.org ​/10​.1109​/ ICE ​. 2019​.8792607. Simon, H., & Fassnacht, M. (2018). Price management: Strategy, analysis, decision, implementation. Springer International Publishing. Stich, V., Schumann, J. H., Beverungen, D., Gudergan, G., & Jussen, P. (2019). Digitale Dienstleistungsinnnovationen: Smart Services agil und kundenorientiert entwickeln. Springer. Stoppel, E., & Roth, S. (2017). The conceptualization of pricing schemes: From product-centric to cusstomer-centric value approaches. Journal of Revenue and Pricing Management, 16(1), 76–90. Tzuo, T., & Weisert, G. (2018). Subscribed: Why the subscription model will be your company’s future—And what to do about it. Portfolio/Penguin. Wells, A. R., & Chiang, K. W. (2017). Monetizing your data: A guide to turning data into profit-driving strategies and solutions. John Wiley & Sons. Zheng, P., Wang, Z., Chen, C.-H., & Khoo, L. P. (2019). A survey of smart product-service systems: Key aspects, challenges and future perspectives. Advanced Engineering Informatics, 42(C), 100973.

C hapter

1 0 

133

C hapt e r 1 1

Price Sensitivity Meter and Conjoint Analysis as Tools for Setting Your Industrial Subscription Pricing Maciej Wilczyn´ski and Matt Johnston

Part 1: How to run a price sensitivity meter exercise

Maciej Wilczyński

Choosing the right pricing strategy

One of the key questions asked during product development efforts is how should we price it? The easiest is to apply a cost-based approach and add a healthy markup to the product and distribution costs—it’s enough to pay the bills, but we’re below the bottom line if the cost structure shifts. Another idea is to see what the competition is doing within the pricing space. With that, it’s easy to position your product as more expensive or as cheaper, depending on the overall brand strategy. However, there’s a drawback here. If you’re using competitor-based pricing, then it’s not your pricing strategy at the end of the day, but your peers’ strategy. You probably wouldn’t outsource product development efforts or sales to your competitors, so why do it with your core business model? One strategy is simply the best but also the hardest to achieve for most companies: value-based pricing. This strategy sets the price based on how the customer perceives the value rather than on how much the product costs or the price of competitive products. To get it right, companies need to understand how much customers can pay to receive the benefits. Some try to estimate it based on previous sales, and some tend to focus on interviews, while the vast majority rely on a guess or gut feeling, as they spend too little time on it. For instance, a subscription company’s average time on pricing improvements is less than 10 hours a year. In other words, we focus more on choosing the right toilet paper than on choosing the right pricing strategy. To get it right, it’s important to start from the willingness-to-pay concept. This essentially means how much your customer can pay for your product or service you offer. There are ways of doing it through interviews, as described in other sections

DOI: 10.4324/9781003226192-14

135

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

of this book, or you can try getting it through Van Westendorp’s price sensitivity meter.

Preparing the price sensitivity meter survey

Van Westendorp’s price sensitivity meter (PSM) has been with us since 1976, and it’s one of the best techniques for determining the optimal price point. Market researchers widely use it, and it’s one of the most effective tools available. It’s also one of the easiest ways to discover the customer’s willingness to pay. In general, you need to ask these four survey questions to run the Van Westendorp PSM test: ■ ■ ■ ■

At what price would it be so low that you begin to question this product’s quality? At what price do you think this product begins to be a bargain? At what price does this product begin to seem expensive? At what price is this product too expensive?

There are discussions around the order or exact phrasing of the questions, but ‘from cheapest to most expensive’ tends to be prevalent in the literature. When it comes to sampling and its size, we need to focus on our goal. It’s best to receive answers from the market, but current customers could also work if you don’t have access to a panel or a research agency. The good thing is that ~100 survey responses are enough to drive meaningful results in B2B industries. Some research suggests it’s 40 to 60, but reaching the above three-digit threshold allows for creating more segments and identifying different buyer personas. Naturally, for B2C products, we need to boost the sample size to at least 300 to 400. Keep in mind that we want to move quickly and make data-driven business decisions at the end of the day, so full statistical significance is not our goal. According to my experience, it’s ~80% accurate, which is more than enough.

Plotting the data

Once you have all the responses, it’s essential to plot them correctly. On the x-axis, you need to plot the price points, ideally in thresholds, closest to your hypothetical price. I recommend focusing on two decimal places; similarly, it’s always good to round to fives and tens for more expensive products. On the y-axis, it’s critical to plot the cumulative number of responders. It’s needed later, as it’s essential to calculate the price elasticity and potential sales lost with the price increase. Once done, it’s possible to display it on a chart, as in Figure 11.1. Each question asked corresponds to a line on the graph. If you follow the order proposed, ■ ■ ■ ■

136

the first question corresponds to the ‘too cheap’ line the second corresponds to the ‘not expensive’ line the third corresponds to the ‘not cheap’ line the fourth corresponds to the ‘too expensive’ line

C hapter

1 1

P S M a n d Co n j o i n t A n alysis # of respondents (cumulative %)

Van Westendorp Price Sensitivity Meter

100% 90% 80%

Not Expensive

Not Cheap

Too Cheap

Too Expensive

IPP

70% 60% 50% 40%

PMC

PME

30% 20%

OPP

0%

$0 $1 $2 $3 $4 $5 $6 $7 $8 $9 $10 $11 $12 $13 $14 $15 $16 $17 $18 $19 $20 $24 $25 $28 $30 $35 $40 $50 $60 $70 $80 $90 $100 $150 $200 $250 $300

10%

Acceptable Price Range

Price

Figure 11.1 Price sensitivity meter interpretation.

Once it’s done, you can interpret the data: ■ ■

The intersection of ‘too cheap’ and ‘not cheap’ is called the PMC, or the point of marginal cheapness. Its opposite, the intersection of ‘not expensive’ and ‘too expensive,’ is the PME or the point of marginal expensiveness.

These two are critical because this is precisely our price range. In other words, if you price the product between them, it’s already good, but you can plot it further to discover the exact price point. ■

■

The intersection of ‘too cheap’ and ‘too expensive’ is our OPP, or optimum price point. It’s the price at which the number of respondents who consider the product too cheap equals the number who find it too expensive. You can interpret this as a market equilibrium that you’ve just discovered. Another point worth mentioning is the intersection of ‘not expensive’ and ‘not cheap,’ called the IPP, or the indifference price point. It’s the price that usually represents the median price or very often the leading brand on the market. If you combine the PSM with other survey questions, this may become useful.

This already provides a lot of knowledge, but there’s more we can gain from the data set. If you take only the ‘too cheap’ and ‘too expensive’ lines and begin plotting them to the OPP, you can better understand the percentage of sales lost within the process. This answers the fundamental question ‘what happens if I increase/lower the prices?’ Also, it allows us to pick the right strategy: for example, if you want to focus on having the highest market share. C hapter

1 1 

137

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

It’s possible to focus on a penetration price. On the other hand, if the curve is flat, we can easily recognize how much money we’re leaving on the table. It’s an excellent hint to identify a customer’s lifetime value-maximizing price. Suppose you’ve gathered more than 100 surveys and asked other questions relevant to your business. You can now run some cross-tabs to identify different buyer personas. Doing so may lead to a complete redefinition of your current pricing scheme.

How to take your PSM even further

There are other things you can do with pricing research. One of them is the Newton– Miller–Smith extension, which adds two questions to the mix: ■ ■

At the (‘not expensive’ price listed by the responder), how likely are you to purchase this product in the next six months? At the (‘not cheap’ price listed by the responder), how likely are you to purchase this product in the next six months?

Both questions are rated on a five-point likelihood scale. If you want, you can also use an NPS-like 0–10 scale, but keep in mind that it’s widely known in the industry, so you don’t want to bias your responders. Afterward, apply these scores to create the two straightforward graphs illustrated in Figure 11.2. Compared with the traditional PSM, this allows us to better understand the price elasticity of demand, which translates to ‘what happens if I increase my price by x%?’ You’re no longer guessing; you’re doing science. On top of that, because you have the data, you can multiply the products sold by 1,000 units to obtain the revenue curve, which shows the revenue-maximizing price. Sometimes it’s different than the OPP derived using the standard approach, but that’s why we’re doing the extended analysis.

Predicted revenue (for 1000 units sold)

% of customers willing to pay

Price Elasticity vs Demand 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% $0

$4.500 $4.000 $3.500 $3.000 $2.500 $2.000 $1.500 $1.000 $500 $0 $0

$5

$10

$15

$20

$25

$30

$35

$40

$45 $50 Price level

$40

$45 $50 Price level

Predicted revenue vs. price level

Revenue-maximizing price around $15 $5

$10

$15

$20

$25

$30

$35

Figure 11.2 Applying the Newton–Miller–Smith PSM extension.

138

C hapter

1 1

P S M a n d Co n j o i n t A n alysis

The biggest drawback of this analysis is that you need to use more advanced tools to run your survey. There’s a need for question-and-answer piping, so free tools probably won’t handle it. And although this isn’t a substantial cost, it’s always good to come prepared.

Pros and cons of PSM analyses

In general, PSM analyses are excellent for identifying your products’ price points because they’re relatively easy to perform, provide actionable information, and allow you to segment the data by customer. They also work great in B2B industries, as customers are more aware of their needs. On the other hand, it doesn’t take costs into consideration, so we need to cross it with internal data to see whether we retrieve a healthy profit margin. Also, it works amazingly well for new product development, but competitive pricing tends to be a challenge because the questions don’t consider other products. Keep in mind that you need to put an analyst in place, so if you don’t have capabilities within the organization, hiring a pricing research consultant/agency may help. It’s powerful knowledge that requires some effort and analytical data-driven thinking. Unfortunately, there are no quick growth hacks or A/B tests. The benefits of knowing your clients’ willingness to pay can improve profitability by up to 12%, so investing in pricing provides a decent ROI.

Bio

Maciej Wilczyński is a Co-founder and CEO of Valueships (https://valueships​.com), a consulting boutique for subscription businesses, serving mostly SaaS/D2C. Valueships solves their clients’ acquisition, retention, and monetization problems through data analytics and research. Wilczyński is an ex-McKinsey & Company marketing and sales consultant and has worked with top Fortune 500 companies in industries such as software, banking, telco, insurance, and retail/e-commerce. He is currently finishing his PhD in strategic management with a focus on pricing capabilities of SaaS companies. He is the author of multiple publications, an MBA/postgraduate lecturer, and guest speaker. He is a co-founder of Stanversity (https://industry​.stanversity​.com/)—a university-lecturers platform connecting the best scientists with universities in need of new study courses, mostly on the postgraduate and MBA levels. He is excited about tech, B2B pricing, monetization, digital marketing, customer insights, and quantitative and qualitative research. Connect via Facebook: http://fb​.com​/wilczynski24; Twitter @ wilczynski24, at LinkedIn www​.linkedin​.com ​/in​/wilczynskim/, and via email: maciej​ @valueships​.​com

Part 2: A practical guide to conjoint analysis: Injecting more confidence into your strategic pricing decisions

Matt Johnston Market research and pricing are both fascinating disciplines that go, to a certain degree, hand in hand. One of the most trusted and scientifically advanced methodologies C hapter

1 1 

139

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

to measure willingness to pay is conjoint analysis—the holy grail of market research. In this guide I share practical knowledge and best practices to motivate companies, large and small, to benefit from the power of conjoint analysis and boost your research success.

Setting the scene: Why willingness to pay matters

The goal of every business is to charge the maximum willingness to pay for each of your customer segments. Too many companies leave substantial money on the table by simply not charging what the client thinks their product or service is worth monetarily. While company costs can help identify the minimum price floor, what’s the price ceiling in each segment? In some markets, we can rely on competitive pricing data, which might be limited to substitute offers. However, the best way to determine what your customers are willing to pay is to ask them! How you determine willingness to pay is critical. In my experience, companies find it difficult to appreciate that customers think in terms of value, not price. In other words, nobody cares about the costs or features of your product or service—they care solely about the value it creates for them. The upside: conjoint analysis can effectively help you here!

Optimizing pricing strategy: Use cases for conjoint analysis

Conjoint analysis helps you to focus on key value drivers of your offers and present them in a way that a customer can subconsciously decide which offer is most valuable. Finally, it allows you to add a price tag to this perception of value. This is because conjoint analysis presents products in a way that simulates real-world product comparisons and asks respondents to make realistic trade-off decisions. In a decision scenario, people don’t like to choose unless there’s something with which to compare or contrast. By asking your customers repeatedly to choose from a variety of offers with a unique set of specifications and prices, you can gather and process information about customer preferences and price elasticity using statistical algorithms. Eventually, conjoint analysis will help you answer your burning pricing questions: ■ ■ ■ ■ ■ ■ ■ ■ ■

14 0 

Determining customer willingness to pay for a product or service at an aggregate, segment, and individual level. De-risking new pricing/charging models (servitization, monthly fee, up-front plus monthly fee, once-off fee). De-risking new price promotions (discounts, bundling, rebates, etc.). Understanding your product’s price/value perception versus competing brands. Measuring the impact of bundling versus add-on on revenue. De-risking new price-setting. Determining price elasticity/sensitivity at an aggregate, segment, or individual level. Understanding customer reactions to various pricing T&Cs. Evaluating competitive pricing threats (promotion, new low-cost challenger/ entrant).

C hapter

1 1

P S M a n d Co n j o i n t A n alysis

As you can see, there is a myriad of pricing use cases suitable for conjoint analysis. Most commonly brands conduct a conjoint analysis before a product launch as insurance against nasty surprises. It’s easy to reduce a price when you get it wrong, but almost impossible to increase it. Therefore, smart brands ensure that the price they launch with is what the customer is willing to pay and that the associated charging model/structure is how the customer is willing to pay.

Your research success: What’s needed to conduct a conjoint analysis

As mentioned, a conjoint analysis attempts to simulate a real purchase decision as closely as possible. The following best practices should always be considered: 1. Who should you target? The target audience must represent the end customer as closely as possible. Setting screen-out questions and using online respondent panel provider profiles can facilitate capturing accurate target samples. 2. What should you test? To ensure the integrity of your results, the products/services must be direct substitutes and/or competing products (apples to apples!). 3. Agree on the burning question: it can be a challenge or opportunity that needs to be addressed with all stakeholders before designing the survey. 4. Never test more than eight attributes in one study. Any more will be too much for respondents to digest and could compromise results. 5. ‘None of these,’ or a derivative of it, should feature as an option on every choice card. Because conjoint analysis tries to simulate a real purchase decision, respondents should be given the option to walk out of the store or leave the webpage if they don’t like any of the product options on offer on a choice card. 6. For consumer conjoint studies, to avoid respondent fatigue and the resulting risk to result integrity, we recommend using no more than 12 choice cards in a study. 7. If you’re introducing new features, products, or pricing models, we recommend that you help respondents familiarize themselves with them using text or visual explanations in the survey’s introductory message. A special challenge for industrial companies is often the available respondent pool. This can be approached in two ways. First, limit the survey’s complexity—specifically the number of concepts and attributes to be tested. Second, increase the number of choice cards. The number of choice cards needed has an inverse relationship to the number of respondents. If the pool of respondents is limited, industrial conjoint studies can accommodate up to 20 choice cards.

A practical example from EPIC Conjoint: Testing the market acceptance of IoT smart tracker devices

A classic use case for conjoint analysis is product testing before a market launch. The leading brands invest millions of dollars to find the best combination of technical features that make a great product, which ultimately can conquer market share and

C hapter

1 1 

141

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

increase sales numbers. Using conjoint analysis, we can understand the effect of product and pricing decisions and test thousands of different feature combinations. In the following, we take a brief look at an IoT smart tracker device survey that was launched in July 2020 in Germany—highlighting the key survey steps. 1. Identify your product’s peer group, against which consumers will match your offer. In this case, we test one product—an IoT Smart Tracker—that could be offered by different providers at different prices. 2. Identify the key attributes that matter most for your customers—and on which they base their purchasing decision. These can be technical specifications (the product itself), or softer value drivers (e.g., brand of the provider). Finally, we attach two prices: one monthly fee and one initial fee. Also, we include a free test period as a potential value driver. 3. Ask questions that you need to have answered in order to distinguish customer segments (segmentation questions), to have laser precision on your target audience (screen-out questions), or to ensure a specific quota in your answers (e.g., male-tofemale ratio). 4. Identify the target audience! A crucial step in every survey is to ask the people who bought or might buy your product. Having a dedicated respondent panel provider and a wide variety of methods to distribute your survey make it significantly easier to reach the target audience. 5. Launch the survey. Choice cards for a choice-based conjoint analysis run by EPIC Conjoint typically look like the example in Figure 11.3. Respondents choose their favorite offer from the screen, while within the attributes (e.g., brand) the specific attribute levels (e.g., Vodafone or O2) will change—along with the two prices. 6. Analyze the results and identify the key learnings! At EPIC Conjoint, all the questions you add to the questionnaire can be actively used to filter the results, so you can replicate your customer segment. Hence, differences in preferences and willingness to pay among different buyer groups are easily identifiable.

Relative feature-level preference

Bringing all attributes and attribute levels to one linear chart makes differences in preferences to the target sample easily identifiable. For example, look at how the perception of the attribute ‘brand’ shows two groups of preferred providers: Telekom, Vodafone, and 1&1 are generally preferred, and the other tested providers are less preferred (Figure 11.3). These findings can be identified for all other attributes, too (Figure 11.4).

Relative feature importance

Knowing what really drives your customer’s purchasing decision, and what the significant value drivers are, is key for every comprehensive pricing strategy. Taking the learnings of relative feature preferences, we can calculate the importance of each attribute. In our IoT tracking device survey, we identify a clearly price-driven product decision: price itself drives 72.1% of respondents’ product decisions; brand (16.5%) is only onefourth as important (Figure 11.5).

14 2 

C hapter

1 1

P S M a n d Co n j o i n t A n alysis

Figure 11.3  Choice cards. Ideal product

A great starting point for every product developer is to know your target audience’s most desired combination of attributes. For IoT tracker devices, the attribute combination shown in Figure 11.6 is the one most voted for.

Price elasticity

How price sensitive are your customers—and how does price sensitivity change between different customer segments or to different product categories? Do you know C hapter

1 1 

143

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

Figure 11.4 Relative preference level (whole panel, 251 respondents).

the answer? Conjoint analysis can determine this by deriving a precise estimate from the trade-off decisions of your customers. In this survey, the respondent panel had a price elasticity of −0.55 to −0.57. respectively (Figure 11.7).

Powerful market-share and price-optimization simulators

The beauty of conjoint analysis is that based on what you learn about preferences and price elasticities, you can roll up your sleeves and simulate what-if decisions about optimal prices for your product, or market-share changes triggered by your (or your competitors’) product decisions.

Figure 11.5 Relative feature importance (whole panel, 251 respondents).

14 4 

C hapter

1 1

P S M a n d Co n j o i n t A n alysis

Figure 11.6  Ideal product.

Going into further detail would take an extra chapter but look at our price optimizer (Figure 11.8). It tells you the take rate and the revenue—or profit-maximizing price—for your product, launched on a market consisting of your target audience. You can easily and precisely simulate what happens to your take rate and optimal price if you change attributes (e.g., offering a free test period). Here, conjoint analysis gives you a unique tool to identify willingness to pay—not only for your product as a whole but also for different product attributes!

The five things to remember for a powerful conjoint study ■ ■

■

Conjoint analysis must not be rocket science. Rely on the powerful combination of user-friendly conjoint software and professional guidance by industry experts. Conjoint analysis can give businesses of any size a competitive advantage by providing precise information on your own—and your competitors’—market perception and willingness to pay. Guarantee statistical significance of results by asking enough survey respondents. If the number of available respondents is lower than approximately 300, try to decrease the complexity of the survey and increase the number of choice cards slightly.

C hapter

1 1 

145

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

Figure 11.7  Price elasticities.

In every survey, answer the two major questions: ■

■

14 6 

Who are you targeting? Be clear about who your target audience is, make sure you can reach them during the survey distribution phase, and ensure that you give enough context in the survey to enable the respondent to give precise answers. What are you targeting them with? Conjoint analysis is an excellent opportunity to think about your offer not only technical feature terms—but especially regarding the value you provide. Identifying key value drivers and using them as attributes in a conjoint analysis is a perfect starting point!

C hapter

1 1

P S M a n d Co n j o i n t A n alysis

Figure 11.8  Price optimizer. Bio

Matt Johnston is the Founder and CEO of EPIC Conjoint. He is a seasoned commercial marketing professional with over 20 years of pricing, product, and segmentation experience. Matt has an extensive background in telecommunications as former Head of Pricing at Telefónica Ireland and Ooredoo Group Qatar. Visit www​.epicconjoint​.com for more information and to contact Matt.

Key objectives 1. Learn how pricing research techniques can be applied to digital innovations. 2. Learn the content, method, and outcome of a price sensitivity study. 3. Learn the content, method, and outcome of a conjoint analysis study.

Key summary points 1. Too many companies develop their digital offers and their pricing based on cost or gut feeling. They must embrace advanced research methods to improve their pricing power. 2. PSM and conjoint analysis have been greatly simplified and democratized. Results can be achieved in days and with a minimal budget. In addition, the complexity has been removed from a design perspective as software development has also improved. 3. Both research techniques allow for the calculation of deep pricing insights that should enable the power of digital innovations. There is no reason why marketers and digital business leaders should not embrace pricing research in the future.

C hapter

1 1 

147

M A C I E J W I L C Z Y N´ S K I A N D M AT T J O H N S T ON

Key questions 1. Should leaders rely on quantitative research methods alone? What happens if there are only a few customers available to research? 2. What are the key success factors in designing and implementing these two research techniques? 3. How can companies include these research techniques in their digital innovation development process?

14 8 

C hapter

1 1

SECTION 3

The Value and Pricing of Data



C hapt e r 1 2

Overcoming Real-World Challenges in B2B Digital Pricing Transformation Lalit Wadhwa

Introduction

When organizations launch their digital transformation journey, a pricing transformation from a cost-based model to a value-based model is generally considered to be lowhanging fruit. Organizations see pricing optimization as an opportunity to generate quick value, both in revenue and margin upside. Most successful B2B case studies show that the margin upside can be between 2 and 4 percentage points, and the realization of this value can be achieved within a few quarters. Yet despite investing in digital pricing technologies, many organizations find progress to be frustratingly slow or, worse, give up on their digital pricing transformation ambitions and revert to cost-based models. These organizations generally conclude that either the technology doesn’t work as they assumed or that digital pricing initiatives are out of step with their go-to-market framework. This chapter investigates potential stress and failure points across the entire spectrum in B2B digital pricing initiatives. Identifying these issues early in the design and implementation cycles and preparing to counter them with a suite of options is also covered. While this list may not be exhaustive, it does include a practitioner’s insight into unique and unexpected challenges. These challenges originate in factors such as the organization’s technology choices, technology skills, digital transformation roadmap, project versus product-based approach, build versus buy decisions, current and future state pricing workflows, the industry ecosystem, competitive positioning, leadership resolve, and organizational culture. The list is long, but awareness and careful attention to these factors significantly improves long-term success probability. Each of the following sections in this chapter identifies a challenge, investigates the impact, and proposes potential solutions. Sections are not arranged in any particular order, although readers will quickly be able to evaluate that solutions to certain challenges must precede others given their foundational nature. These sections can also serve as a guide to determining organizational readiness when proceeding with digital pricing transformation. When used as a readiness guide, it is best to bring in DOI: 10.4324/9781003226192-16

151

L alit Wadhwa

perspectives from a broad range of stakeholders including sales, product, procurement, pricing, marketing, finance, IT, and L&D (learning and development). Digitalization of pricing impacts existing practices and downstream workflows for each of these stakeholders in different ways, and each stakeholder needs to prepare for technology enablement as they transition from the current state to the future state. The team

As a part of an organization’s digital transformation roadmap, a digital pricing transformation is generally driven by pricing or finance teams. Irrespective of which stakeholder drives the transformation, two key questions must be addressed head-on. 1. Do all stakeholders have representation in this transformative initiative, do they understand their role, and have they identified the right talent that would contribute to the pricing transformation? 2. Who is best positioned to lead the pricing transformation? Over and above the usual attributes of successful teams that deliver transformative initiatives, every team member must be comfortable with leveraging data and technology to accelerate customer-centricity and organizational productivity in the future state. Over the course of the transformation, this team will encounter multitudes of dataand technology-related challenges and must have the skill to carefully understand the impact of these on long-term success. Any shortcuts adopted by this team will end up showing as technical and process debt in the solution. While it is virtually impossible to end up with a zero-debt solution, understanding their long-term impact is important. When it comes to who is best positioned to lead the digital pricing transformation, two key attributes need attention. The first is familiarity with an agile approach. The need to deliver digital pricing transformation as a series of iterative outcomes is foundational to success. Attempting this through a traditional waterfall methodology can create major impediments to adoption and results. This is discussed in greater detail in the following section. The second is the need to understand value-creation opportunities in a digital pricing workflow to some degree of granularity. While this doesn’t mean that those without the above-mentioned attributes will not be able to deliver pricing transformation, it is almost certain that outcomes will be suboptimized and that adoption will be challenging.

Project or product?

While there is enough literature available in the public domain that outlines the difference between the two approaches, one of them significantly boosts the probability of long-term success, while the other almost always is a recipe for frustration and insufficient returns. If you guessed ‘product’ as the desired approach, you are right but in the minority. In the B2B world, organizations are used to taking up large initiatives, packaging them as projects to achieve specific outcomes, completing those initiatives, disbanding project teams, and moving on to the next initiative. In a digital solution that is tightly coupled with internal and external stakeholders (such as customers, vendors, and business partners), everything must evolve continuously to maintain relevance. 152

C hapter

1 2

B 2 B D igital P rici n g T ra n sf o rmati o n

Evolution of the solution is driven by data that comes in as product telemetry, user surveys, competitive pressures, technical advancements, and newer data attributes. Digital products thrive when backed by an agile development approach. In these products, the product owner has a deep understanding of the client’s shifting needs and maintains product relevancy by continuously iterating around the needs of internal and external customers and accordingly adjusting UI, back-end integrations, data sources, data structures, algorithms, models, APIs, and product telemetry. In a waterfall approach, there is no efficient mechanism to manage the product’s evolution, leading to rapid degradation of the user experience and value creation. Additionally, this approach rarely maintains the original team, and critical knowledge gets dispersed or lost. Successful implementations of digital pricing transformation usually involve identifying the product owner upfront at the start of the transformation journey, instead of as an afterthought. In many cases, the product owner also ends up being the best person to lead this transformation. Data

The foundation of all digital products is data. Most organizations believe, rightly so, that they have plenty of data that can be leveraged to digitalize their pricing. What rarely gets discussed are critical topics like data quality, data governance, data architecture, data silos, master data, and third-party data. Traditional organizations have generally neglected these topics since the monetization of data has not been a priority. Worse still, there’s limited appreciation of the need to address these topics using combinations of technology and processes. The problem gets compounded if the organization lacks contemporary technology skills to identify, build, and manage efficient data architectures and data pipelines. The topics discussed in the paragraph above are nontrivial issues. They cannot be resolved overnight. It is also impractical to mandate that digital pricing transformation cannot be kicked off until these topics are adequately addressed. However, beginning work on these topics as early as is feasible is key, and iterative goals must be defined and achieved. Another issue to be aware of is that in anything related to data, business and IT have deep interdependence. As an example, defining a data architecture without a good understanding of current and future business-use cases will almost always lead to inferior outcomes. ‘Data’ is where most of the technical and process debt gets created during a digital pricing transformation. The best way out of this challenge is to clearly document the to-be-state that allows delivery of known use cases and iteratively invest in it. The ability to monetize a pricing transformation is largely dependent on getting this right. One final recommendation related to building any data product is to have representation from the legal or compliance team. Data products must comply with the organization’s data policy as well as with any regulatory framework(s). In many pricing transformations, this is an afterthought that can lead to significant rework.

Technology and technical skills

Most organizations end up making technology decisions based on the availability and cost of IT skills. By definition, then, certain use cases cannot be operationalized efficiently since the technology selection and technical design are inconsistent with those C hapter

1 2 

1 53

L alit Wadhwa

use cases. To illustrate this, consider a use case that involves exposing pricing to external third-party aggregators through APIs and charging for that access. If the technology selection does not easily support the deployment of secure APIs with monitoring, this use case must wait its turn. Another example could be a use case for updating the pricing model every few hours. If the data architecture and technology selection do not support efficient data pipelines and model engineering, this use case would need to be parked for another day. Understanding the technical constraints surrounding the design and deployment of the solution is key to building a product roadmap. While there are numerous technology-related decisions that organizations must evaluate when implementing their pricing transformation roadmap, three areas have a disproportionate impact on future success. These are infrastructure, integrations (internal and external), and user interfaces. Using a tech stack that adequately supports decisions in these areas can help reduce technical debt. While not mandatory or critical, when technology decisions are led by those who have a high-level understanding of the industry and ecosystem in which the organization functions, outcomes are generally better. These leaders understand that technology is an enabler of achieving stated business objectives and must not be evaluated in isolation. Algorithms and models

Algorithmic and dynamic pricing use data from a multitude of sources. In general, every new data source added to the algorithm brings in new signals and refines the pricing prediction slightly. Most B2B organizations have multiple internal data sources (master data, ERP, CRM, WMS, web commerce) as well as dozens of external data sources (industry data, aggregator data, trend data, customer financial data, competitive data) that can deliver incremental gains toward accurate value-based pricing. When following a product approach, each data source can be added iteratively to create more robust pricing models. With few exceptions, B2B organizations are perfectly happy working with algorithmic pricing that is refreshed daily. Pricing models are usually built using some combination of algorithms. As an example, basic pricing models can be built using a combination of clustering and decision trees, while complex pricing models could use reinforcement learning (RL) and neural networks. Ideally, organizations must spend some time experimenting with different algorithms and multiple models and identify several that meet their pricing transformation goals. Mature digital pricing transformations leverage continuous A/B testing. More on that in a later section. While algorithms and models are the lifeblood of digital pricing innovation, it is easy to go overboard. Complex models can end up being black boxes, as well as computing intensive. This can make long-term adoption trickier, and the general wisdom is that organizations should begin with more parsimonious models and iteratively graduate to complex models. Again, an agile-led product approach serves the cause well. Finally, it is important to be aware that most models built using ML may someday be regulated in some fashion. Models that impact economic or social equity are definitely candidates for regulation in the long run but there is new academic literature available in the public domain that outlines how algorithmic pricing in the B2B domain can lead to anticompetitive practices under certain conditions. This is another good reason to have representation from legal or compliance teams when digitalizing pricing. 154

C hapter

1 2

B 2 B D igital P rici n g T ra n sf o rmati o n Testing and auditability

Most organizations are adept at running unit, performance, regression, and resilience tests before deploying deterministic software code. However, models with predictive power need additional attention and testing, and many times organizations come to this realization when the model output begins drifting after the pricing model has been deployed. Thus, observability and monitoring are key components of the solution; without these, drift cannot be identified or measured. Model testing in machine learning is a component of the rapidly growing field of MLOps (machine learning operations) and needs serious attention and specialist skills. There are three primary goals of model testing: robustness against noise and drift, interpretability of output, and reproducibility of output. Organizations wishing to deploy complex dynamic pricing models must simultaneously ensure that their modeltesting skills mature rapidly. Digital pricing models are not a high-stakes application (unlike, say, health care diagnosis or credit scoring), and thus the focus on auditability of the solution is limited. Simultaneously, model explainability and auditability have seen rapid innovation in the past few years. When organizations are aware of the need for auditability and launch their digital pricing journey using simpler models, most concerns in this area can be adequately addressed. Integrations

A pricing platform cannot function in isolation without being integrated into the data flow and pricing workflow within the organizational ecosystem. Automation of data pipelines that feed the pricing model, model refresh based on specific criteria and triggers, and pushing the model output to relevant systems (such as the CPQ platform, CRM, or an e-commerce back end) are critical. In practice, most organizations understand and execute quite well on internal integrations. What generally gets little or no attention are potential integrations with the external ecosystem. Examples of these integrations include those with marketplaces, industry data aggregators, and customers. Without solving this adequately, an end-to-end digital pricing transformation that improves client experience doesn’t get addressed. A best practice is to ensure that from an architectural perspective, the pricing model can be ‘served’ and is best discussed right up front during the system-design phase. Continuous experimentation (A/B testing)

A/B testing of pricing is a controversial idea, with plenty of opinion for and against it. Organizations in the initial stages of their pricing transformation are generally better off not considering it. Once the power of pricing digitalization and the potential drawbacks of A/B testing of pricing are well understood in the organization, this can be a powerful tool for optimizing product quantity, customer count, and margin concurrently. Interpreting results of A/B testing does need skills to ensure that only statistically significant results are considered. Conclusions

This is not a complete list of challenges. The focus of this chapter has been to unearth challenges that are either underrated or ignored but that have a large impact on the C hapter

1 2 

1 55

L alit Wadhwa

pricing transformation journey. Indeed, readers may point out, rightly, that organizational culture is a larger predictor of success when undertaking digital transformation. But pricing transformations do fail even when the organization’s heart is in the right place. Successful digitalization of pricing depends on bringing together clear goals, adaptive culture, quality data, technology skills, customer-centricity, and some persistence.

Bio

Lalit Wadhwa is an accomplished engineer with over 25 years of wide-ranging experience in modern technologies. He is a recognized thought leader and frequent speaker on a wide range of important technology topics, including cloud adoption, data science, data products, AI, and predictive analytics and digital services. As Executive Vice President and Chief Technology Officer, Lalit focuses on the use of advanced technical tools and methods in Encora’s client services. He leads Encora’s technology practices areas, in support of the company’s sustained growth. His prior responsibilities included building the company’s data, advanced analytics, and digital capabilities across the supply chain and customer chain. Lalit holds an engineering degree from the Delhi College of Engineering, specializations from the University of California and Johns Hopkins University, and over 20 technology-related certifications. LinkedIn: www​.linkedin​.com​/in​/ lalitwadhwa/

Key objectives 1. Understanding unique challenges and accelerators, technical and cultural, associated with B2B digital pricing transformation. 2. Potential solutions to challenges that ensure alignment with the overall digital transformation framework of the organization. 3. Best practices that accelerate deployment and adoption of pricing methodologies as a part of the organization’s digital transformation roadmap.

Key summary points 1. Sustained pricing transformation with digital technologies is very rewarding from a margin and revenue perspective but is more complex than the low-hanging fruit that it is generally seen as. 2. Digital pricing transformation impacts several stakeholders. Having all stakeholders participate and contribute to the journey is a best practice. 3. Organizations underestimate the power of their data and overestimate the quality of their data. Without addressing both, pricing transformation can be underwhelming.

156

C hapter

1 2

B 2 B D igital P rici n g T ra n sf o rmati o n

4. There is a high correlation between use cases, technology choices, and customer experience. Shifting one also shifts the other two. 5. Taking a product approach to pricing digitalization greatly improves the probability of sustained success.

Key questions 1. What are some of the challenges born of organizational culture that you can think of? 2. What could be some of the key differences between digital pricing transformations taken up by B2B versus B2C organizations? 3. Do you think dynamic pricing is a fair practice? Evaluate this from the seller’s as well as the buyer’s perspective.

C hapter

1 2 

157

C hapt e r 1 3

Holistic Approach to Market Segmentation of Industrial Smart Services What Is the True Value of Data? Luis Prato

Introduction

Markets are more complex than ever. Original equipment manufacturers (OEMs) can merge their manufactured products with sensor technology (Cenamor et al., 2017; Luz Martín-Peña et al., 2018; Opresnik & Taisch, 2015), providing an array of advantages for the collection of performance data from their installed base. OEMs can use newly developed digital capabilities (Porter & Heppelmann, 2014, 2015) to convert performance data into insights, thus replacing traditional service offerings with advanced services (Lightfoot et al., 2013). The term industrial smart services (Allmendinger & Lombreglia, 2005) describes a new generation of services resulting from the combination of data collected from an installed base or connected products and information from other additional field sources. Thus, firms are beginning to consider performancebased relationships more effectively, through the delivery of guaranteed production outputs, machine uptime, or product quality, based on those data insights. Recent research shows how leading European OEMs aim to more than double their service revenues from industrial smart services by 2024 (see Figure 13.1). The service categories with higher growth potential approach the optimization of maintenance from the current installed base by being remote maintenance, and predictive maintenance is the major driver for this growth. Moreover, more and more operational efficiency services and other digital and training services are filling the gap in teams’ digital capabilities, enabling this shift to happen. However, there remain several barriers to the full adoption of industrial smart services, even though the future state of industrial markets seems bright. This rapid pace of change poses challenges for OEMs in some areas. For instance, the monetization of these offerings and their market reception remain unchanged, although OEMs have deployed highly valuable investments in technology, service capabilities, and R&D to accelerate their transition to industrial smart services. Further, the process of gathering performance data from the installed base is not inexpensive, and the incremental growth of in-house digital capabilities still falls short of penetrating untapped markets. DOI: 10.4324/9781003226192-17

159

L uis P rat o Growth projected to more than double by 2024

2024

53%

Service categories affected by data analytics

Other digital and training services

47%

47% 26% 95%

Operational efficiency

63% 95%

Remote Maintenance 2019

80%

20% Predictive Maintenance

Traditional services

79%

Industrial Smart services

100% 42% 2024

2019

Figure 13.1 Industrial smart services: A critical driver of future market segmentation approaches. Source: Bain & Company 2019 Service Circle Survey (N = 22).

Executives in the C-suite overseeing these service businesses are more impatient than ever due to a lack of sales growth of industrial smart services, despite being shown commitments. Over the past four years, I have interviewed more than 100 business leaders facing similar challenges, and the results do not seem to be different. OEMs from diverse industry sectors, sizes, and geographical locations face great reluctance from industrial markets to accept these novel services during the purchasing process. It seems that OEMs have not defined market segments that have sufficient appetite for industrial smart services. One reason is that how these markets are segmented appears to be underdeveloped, as the ‘buyer purchasing criteria’ in these organizations tilt more toward product-based business models than toward service-based businesses (Van der Valk & Rozemeijer, 2009). In particular, the buying functions of firms are considered barriers to selling services (Stoll et al., 2020) because tools and mechanisms for buying products are employed when these firms purchase services (Stoll et al., 2021). As a consequence, the resulting market segments do not recognize the value of newly developed digital capabilities (i.e., capturing, connectivity, transformation, storage, analytics) embedded in such industrial smart services (West et al., 2021). Second, the mismatch of goals between buyers and OEMs remains a source of disagreement for the development of reciprocal and symmetrical performance-based relationships. Mostly, either the performance orientation of these services is not well understood by buying firms or performance as a concept has other connotations. Thus, the potential of industrial smart services to penetrate such markets is jeopardized by a variety of difficulties. This burning platform suggests that the ability to identify target market segments is critical for OEMs while selling their industrial smart services. For this reason, understanding a buyer’s view of market segmentation entails that OEMs pay more attention in times of turbulence such as in the post-COVID-19 era. The conclusion: industrial natives need alternative market-segmentation frameworks that consider the value provided by their digital capabilities and the performance orientation of such services to better position these service offerings. 16 0 

C hapter

1 3

S mart S ervice a n d D ata Value Industrial market segmentation

Market segmentation is a marketing technique that involves conducting strong market research with customers to define smaller and more refined segments. Over the past 40 years, different segmentation models have been used to identify customer segments. Although these models are widely similar, the variables triggering market clusters differ slightly. The main segmentation criteria in business-to-consumer (B2C) markets are triggered by geographic, demographic, psychographic, and behavioral variables related to consumers (Kotler, 2000); market segmentation in business-to-business (B2B) settings, however, still lacks a generally accepted and validated way to achieve this goal (Beane & Ennis, 1987). The task is even more complex in industrial markets because consumer products serve only one purpose whereas industrial products have multiple applications. Moreover, the ultimate purchasing decision in industrial markets is made by a group of different stakeholders representing different business functions within the buying firm. Thus, although a buying center or group of these industrial customers may be interested in the same product, their behavioral responses to its benefits may differ substantially (Kotler, 2000). Shapiro and Bonoma described the first steps of market segmentation in industrial markets. Their seminal paper, published in 1984 by Harvard Business Review, first presented the concept of the ‘nested’ approach to industrial market segmentation in response to these buying situations. Indeed, the buying behavior of industrial customers is the underlying rationale for the nested approach. In principle, five main variables or nests become the source of information for clustering customers in different market segments (see Figure 13.2). Market segmentation in industrial markets is normally done from the supplier’s view, as most of these nested variables—company demographics, situational factors, and personal characteristics of the purchase—are not difficult to obtain. Information about industry, company size, and customer location can be determined without even visiting the customer. On the other hand, situational factors can greatly affect purchasing approaches and require more detailed knowledge of the customer, but this information is more likely to be gathered Demographics Operating variables Purchasing approach Situational factors Personal characteristics

Figure 13.2 Nested approach first presented by B. P. Shapiro and T. V. Bonoma, in ‘How to Segment Industrial Markets,’ Harvard Business Review, 62 (May 1984), pp. 104–110. C hapter

1 3 

161

L uis P rat o

by sales teams. These include the urgency of order fulfillment, product application, and order size. Finally, personal characteristics of the purchase refer to stakeholders in organizations making purchase decisions, although the organizational framework in which they work constrains their final decisions. The information about typical roles, buyer motivation and individual perceptions, and risk-management strategies can trigger more comprehensive groups of buyers (individuals) depending on their similarities. This information can also be obtained, with little effort, from external sources. Nonetheless, the operating variables and purchasing approach of the buying firm require somewhat more intimate information. Variables such as the type of technology resources invested and deployed by the buying firm, as well as their level of capabilities to employ such technologies, require knowledge of internal processes that are not easily obtained from outside sources. Moreover, the purchasing approach of the buying firm is often neglected as the most important means of segmentation. OEMs seldom assess the formal organization of the purchasing function, the power structures, the nature of buyer–seller relationships, the general purchasing policies, or the purchasing criteria (Shapiro & Bonoma, 1984). Nowadays, digital technologies make possible the seamless integration of the physical (e.g., product-related human services) and virtual (e.g., digital and analog information) worlds in industrial markets. This connection offers new challenges as these markets migrate from product-based models to more service-based businesses. Markets can leverage these digital capabilities to allow industrial smart services to pave the way to performance-based business models (Selviaridis & Wynstra, 2015). However, most industrial markets are accustomed to purchasing products, not services. Market segmentation has been commonly done based on customer market potential, customer service cost reduction, and customer retention and service-level improvements when positioning services (Gilmour et al., 1994). Service organizations of OEMs have focused on demographic and geographic data of the buying firm to investigate these variables instead of using the nested approach proposed in industrial markets (Shapiro & Bonoma, 1984). On the other hand, when performance management and measurement enters the picture, buyers approach multiple dimensions or performance indicators to assess their business relationships with OEMs. Thus, the performance orientation of industrial smart services is seldom mentioned. In particular, buyers tend to use the total cost of ownership (TCO) of products as a critical differentiator between high and low performers (OEMs). Buyers focus on the fact that supplier performance management and measurement can support different collaborative, commodity, and strategic supplier segments. OEMs, by contrast, face the challenge of the value intangibility of services, for which it remains difficult to demonstrate tangible TCOs because of the absence of similar approaches. Also, when introducing industrial smart services with a performance orientation (e.g., achievement of outputs or outcomes), OEMs still tend to adopt an unsophisticated approach to segmentation. In conclusion, market segmentation in modern marketing is noticeably lightweight not only in the service literature but also in new performance-based contexts. Alternative framework: Defining the key triggers from a buyer’s perspective

This known, complex, and multidimensional nature of market segmentation in industrial digital markets may thus require additional considerations. OEMs can unlock competitive advantages in growing niche markets where these novel industrial smart 16 2 

C hapter

1 3

S mart S ervice a n d D ata Value

Technological resources Technological capabilities Service use Purchasing policies Purchasing functions, Purchasing organization, Power structures, Risk management Buyer–supplier relationships Purchasing criteria Outsourcing strategy

Value of digital capabilities

Selected NEST Variables

Data Analytics

Data Transformation

Data Capturing & Connectivity

C

F

I

B

E

H

A

D

G

Outputs

Outcomes

Inputs

Resulting Segments Segment A Segment B Segment C Segment D Segment E Segment F Segment G Segment H Segment I

Performance orientation

Figure 13.3  Framework representation.

services act as a main source of growth. Figure 13.3 presents an alternative framework in light of digital transformation—one based on the nature of employment of digital capabilities (value of data) and a performance orientation embedded in industrial smart services. This framework taps previous studies (Shapiro & Bonoma, 1984), but from a buyer’s view. Also, the framework reflects my special interest in understanding more deeply both operational and purchasing variables presented in the nested approach but often neglected. The framework is rather simplified for ease of interpretation, thus expanding the service and the marketing literature. Further, practitioners may use this study to refine their internal policies for sales (i.e., OEMs) and purchasing (i.e., buyers) of industrial smart services. In this framework, the vertical axis summarizes the value of digital capabilities embedded in industrial smart services to serve untapped markets. Firms engaged in such services should exploit all their resulting digital capabilities from technology assets (i.e., sensor, monitoring, control, and analytics) to create an augmented value in industrial operations. Previous literature has distinguished four distinct digital capabilities. First, data-capturing capability (Porter & Heppelmann, 2014, 2015) enables firms to measure field data at an unimaginably lower cost. For this capability to be effective, technology assets should be aligned to sense, monitor, and collect field operational data on product usage and performance from connected products with the lowest human involvement. Second, the ability of firms to transmit data remotely from connected products has a remarkable impact on the maintainability, availability, and reliability of the installed base. This data-connectivity capability (Porter & Heppelmann, 2014, 2015) empowers the use of cloud-based communication to optimize the information exchange of functionally connected products and their performance in field operations. Third, data transformation (Lee et al., 2015; Porter & Heppelmann, 2014, 2015) refers to the data-processing ability in which available data are transformed into actionable and valuable information. The main rationale for developing these capabilities is to allow firms to act preventively (Rymaszewska et al., 2017) rather than reactively thanks to the presence of new digitally enabled competitive advantages (Porter & Heppelmann, 2014, 2015). Finally, the capability to interpret (Daft & Weick, 1984) data is commonly C hapter

1 3 

16 3

L uis P rat o

embedded in the well-known data analytics capability (Porter & Heppelmann, 2014, 2015). However, both are interrelated to enable the technical ability of the organizations to analyze processed field data (i.e., the product and service). While the former concentrates on the conversion of data into analytical insights, the latter aims to use this interpretive capability to add value by developing actionable measures for the optimization of operations and processes (Ulaga & Reinartz, 2011). The key to the latter is the combination of field operations insights and product and domain understanding. This dimension was built upon the inputs provided by OEMs about their lack of information regarding the operational variables of the nested approach to market segmentation in industrial markets coined by Shapiro and Bonoma. For that reason the main objective of this dimension is to understand the digital maturity of organizations, for instance, type of technology resources (digital technologies) invested and deployed. Also, how they recognize the true value of digital capabilities developed for levering the benefits embedded in the service offering. This information about buyers’ internal resources and processes is not easily obtained when selling industrial smart services. The horizontal axis, on the other hand, describes the shift to performance (outputs and outcomes) specifications embedded in industrial smart services instead of the required inputs, activities, or processes to be achieved by OEMs (Martin, 1997, 1999). However, for buyers this concept is limited or misleading in some instances. When discussing the performance-based orientation of any service, firms should address two main concepts: outputs and outcomes. Outputs are defined as the direct results of the provision of any service activity; for instance, outputs of a maintenance activity may influence a machine’s uptime. Outcomes, on the other hand, are defined as the value derived from a given service (Bonnemeier et al., 2010; Ng et al., 2009); the actual production of that machine (miles flown and tons excavated) or outcomes may even be expressed in monetary terms (sales income generated by that production). In some cases, outcomes need not always be quantified monetarily but may well include elements that are hard to monetize immediately—such as learning outcomes from an outsourced training program. Nevertheless, buying firms generally neglect these concepts when buying these services, firms often specify the inputs (e.g., time, personnel) required instead. This dimension simply encapsulates most of the purchasing variables according to the nested approach to market segmentation but considering performance orientation of industrial smart services. Thus, OEMs seldom assess the general purchasing policies, formal organization of the purchasing function, the power structures, the nature of buyer–seller relationships, or the purchasing criteria (Shapiro & Bonoma, 1984). For that reason, this second dimension intends to capture this buyer’s view in order to maximize resources and profit impact (Beane & Ennis, 1987) for OEMs. The initial considerations are coded into five categories by having a buyer’s view. For the vertical axis, (1) digital maturity and (2) recognition of the true value of digital capabilities inform about how buying firms may value digital transformation or the value of data embedded in these services. While the horizontal axis informs about the (3) service procurement and contracting strategy, (4) outsourcing appetite, and (5) purchasing behavior decisions of buying firms during the acquisition of industrial smart services. This framework tries to reflect the alignment of goals between buyer and suppliers, the type of purchasing behavior based on performance orientation of industrial smart services, and, more importantly, the value provided by newly developed digital capabilities to boost the buyer’s experience. The matrix (see Table 13.1) is simple and 16 4 

C hapter

1 3

S mart S ervice a n d D ata Value Table 13.1  Description

of typical archetypes

Digital Data analytics capabilities (value of data)

Segment C

Segment F

Segment I

1. Moderate level 2. Interested in getting processed data from machines, packaged into customized APIs and data analytics. 3. Based on preventive activities to guarantee my cost predictability 4. I do it myself because I don’t want any risk related to planning and operations 5. OEMs to assist only with repair activities of my machines (i.e., only spare parts and some labor)

1. High level 2. Interested in getting processed data from machines, packaged into datadriven services (e.g., machine-availability services) 3. Based on preventive activities to guarantee my machine availability 4. I let others do it. Willing to outsource SOME KEY activities supported by new technology 5. OEMs to optimize machine maintenance, connectivity, and logistics (i.e., SLA, availability contracts)

1. Very high level 2. Interested in getting processed data from machines, packaged into data-driven services (e.g., business model innovation and orchestration of my supply chains) 3. Based on predictive activities for my additional productivity and economical gains 4. I do nothing, and rely on my suppliers/ OEMs. Willing to outsource ALL KEY activities supported by new technology 5. OEMs to optimize business models by outsourcing all competencies into long-term partnerships (i.e., outcome contracts, risk-sharing agreements) (Continued )

C hapter

1 3 

16 5

L uis P rat o

Table 13.1 Continued

Digital Data analytics capabilities (value of data)

Segment C

Segment F

Segment I

Data transformation

Segment B

Segment E

Segment H

1. Low level 2. Interested in getting processed data from machines— (e.g., customized APIs and databases) 3. Based on TCO 4. I do it myself because I don’t want any risk related to planning and operations 5. OEMs to assist only with repair activities of my machines (i.e., only spare parts and some labor)

1. Moderate level 2. Interested in getting processed data from machines, packaged into datadriven services (e.g., machine-availability services) 3. Based on preventive activities to guarantee my machine availability 4. I partially let others do it. Willing to outsource some activities 5. OEMs to assist with machine maintenance and connectivity to reduce TCO (i.e., SLA, maintenance contracts)

1. High level 2. Interested in getting processed data from machines, packaged into data-driven services (e.g., business processes and business model optimization) 3. Based on predictive activities to guarantee my machine OEE 4. I let others do it. Willing to outsource SOME KEY activities supported by new technology 5. OEMs to optimize machine, operations, and business processes (i.e., outcomebased contracts)

Segment D

Segment G

1. Low level 2. Interested in getting processed data from machines—(e.g., customized APIs and databases)

1. Moderate level 2. Interested in getting processed data from machines—(e.g., customized APIs and databases)

Data capturing Segment A and connectivity 1. None 2. Interested in getting only raw data from machines— (e.g., standard factory APIs)

(Continued )

16 6 

C hapter

1 3

S mart S ervice a n d D ata Value

Table 13.1 Continued

Digital Data analytics capabilities (value of data)

Segment C

Segment F

Segment I

3. Based on the price 4. I do it myself when it breaks 5. OEMs to assist only with repair activities of my machines (i.e., only spare parts and some labor)

3. Based on TCO 4. I let partially others do it. Willing to outsource some activities 5. OEMs to assist with machine maintenance and connectivity to reduce TCO (i.e., SLA, maintenance contracts)

3. I let others do it. Willing to outsource SOME KEY activities supported by new technology 4. OEMs to optimize machine, operations, and business processes (i.e., outcomebased contracts)

Inputs

Outputs

Outcomes

Nature of performance orientation

largely self-explanatory when analyzing the resulting archetypes (groups of similar buyers).

The process: Results of pilot project

A pilot project was conducted with firms active in construction and metals production environments. The project unveiled the importance of having the buyer’s perspective when selling smart services in industrial markets. Field interviews shed light on the possible benefits of this holistic understanding of both operation and purchasing variables of the nested approach (Shapiro & Bonoma, 1984), as the main strength of this framework. The process to unpack the benefits of this framework has three steps. 1. Determine the target audience. The first step was to identify the target audience to be investigated. For this study, the sample consisted of seven buying companies (BCs) selected by an OEM active in industrial smart services (Table 13.2). All companies are present in construction and metals production environments on a global scale. The OEM was for many years unsuccessful in demonstrating the value delivered by industrial smart services. Mostly, the buying firms claimed data ownership of most data sets supporting the development and delivery of smart services. Thus, these buying firms considered paying for such novel services to be not worth the money. 2. Research the buyer’s view of the target audience. Second, data were collected through pilot interviews with 13 informants (i.e., service, key account management, and C hapter

1 3 

16 7

L uis P rat o Table 13.2  Sample

job titles and companies (buyers and OEM)

Job categoriesN = 29

OEM

Product engineering 2 management R&D/digital 3 transformation Business development 5 Contract management Operations 3 Procurement and corporate development Total 13

BC1

1

1

BC2

BC3

BC4

1

1

2

BC5

BC6

BC7

Total 6

1

5

1

1

1 2 2

3

2

7

1

1

1

1

1

1

5 1 9 3

29

regional operations) from the OEM and 16 informants from all BCs. This step provided the foundation for identifying major trends in the active industry. The interviews followed a prewritten interview structure containing seven questions for operating variables and four questions for purchasing variables in order to generate information about the ‘value of data’ and the ‘performance orientation’ of industrial smart services. The respondents described their core business and market environment. Later they explained the role and scope of digital technologies available and how organizations have evolved on the digital transformation roadmap. The employment of qualitative research explores the phenomena from a company’s perspective, yielding broader insights than quantitative research (Suter, 2014) to understand how the managers judged performance (outputs and outcomes) beyond revenues and margins obtained through the provision of service activities (inputs). Thus, a much broader understanding of the seven companies’ core operating and purchasing variables and their efforts to purchase industrial smart services was gathered. The aim was to generate examples of successes and failures when buying industrial smart services and probe the differences across various types of service offerings. The validation of these insights was done via quantitative research. Respondents from 124 companies (a 25% response rate) were asked to complete an electronic-based survey containing 32 items derived from the qualitative questions. The scale ranged from 1 (strongly disagree) to 5 (strongly agree). 3. Group potential buyers into market segments. Initially we found that buyers’ interests are fundamentally reflected in purchasing decisions based on their loyalty to and trust in OEMs. Second, the service-procurement strategy of the buying firm plays an important role in deciding how certain responsibilities of field- and data-enabled services may be outsourced to suppliers (i.e., do it myself vs. let others do it partially or fully vs. do it together). However, buying companies still lack a clear understanding of the organization’s performance orientation (i.e., cost, TCO, outputs, outcomes). In general, purchasing policies, internal processes, and habits in purchasing products remain a barrier to engaging these organizations in buying industrial smart services. Further, the value of data will depend on organizations’ digital maturity. Buyers may have acquired new tangible assets (i.e., sensors, IoT, cloud computing, platforms, 16 8 

C hapter

1 3

S mart S ervice a n d D ata Value

analytics); however, their intangible assets (i.e., data delivery, data connectivity, data aggregation, data analytics) may remain underdeveloped. Therefore, the latter becomes central to evaluating the value of data because organizations require the employment of proper digital capabilities to extract such value. The results of our pilot interviews and our electronic survey present the following distribution from the total sample. Segments A (8%), B (21%), and C (6%) represent a limited market in terms of attractiveness for industrial smart services. Although buyers expect foundational capturing and connectivity capabilities from OEMs, buyers themselves do not own such in-house capabilities. Commonly, these market segments tend to focus on raw data and data sets rather than on the overall service provided by an OEM. These buying firms are mostly very immature in terms of their digital transformation journey; therefore, they do not understand the value either of developing such capabilities or of outsourcing them. Because they try to be self-sufficient, their outsourcing strategies remain underdeveloped. Such companies do not think about the availability or outcome of machines; they care more about the inputs from suppliers like field technicians onsite and streaming data on time. For a proper service-to-market fit, OEMs should consider establishing a multisided business-model platform to make available data at different levels of aggregation (raw data, processing, and analytics). The underlying logic is to consider the use of OEM-developed digital capabilities to stream data to such markets rather than providing a full smart service solution. Therefore, the aggregation of several layers of digital capabilities enables delivering such value, even though the buying firm still does not consider the achievement of outputs and outcomes to be core to their service procurement and contracting strategy. Segments D (6%), E (28%), and F (21%), by contrast, have already migrated their outsourcing to either outputs or outcomes. Further, these segments are willing to partially let OEMs exercise responsibility for industrial processes and operations through the implementation of smart services. This market encapsulated 50% of potential buyers of such services. Even if OEMs and buyers have upgraded tangible assets (i.e., sensors, IoT, cloud computing, platforms, analytics), their intangible assets or digital capabilities (i.e., data capturing, data connectivity, data transformation, data analytics) may yet be underdeveloped. Thus, the latter piece becomes central to evaluating the value of data, as organizations require the employment of proper capabilities to extract such value. Therefore, the buying firms seek more collaborative relations in order to employ the analytical or interpretative capabilities to optimize business processes and performance from the use of information (field data intelligence). Surprisingly, none of the 52 respondents considered segments G, H, or I appealing to their organizations. This sample of firms informs us that industrial smart services that use outcomes as a main performance orientation are less likely to be purchased by companies that are active in construction and metals production environments, whereas outputs as embedded performance orientation are more likely to be considered as purchasing criteria. Conclusions

The operational dimension of the buyers is closely related to the service use and applicability of the technology through developed capabilities. This study shows that the value of data will depend not only on the digital maturity of organizations but also on the degree of buyer–supplier interdependencies and flexibility to make sense of data C hapter

1 3 

169

L uis P rat o

that has already been collected and processed. Also, performance-based partnerships fail to materialize unless OEMs do not exercise more comparative or complementary capabilities (i.e., product or domain knowledge) instead of competing for the same responsibilities as their counterparts. These findings also confirm that in terms of performance orientation (i.e., inputs, outputs, outcomes), buying firms rely on OEMs to deliver on procurement/contracting strategy (i.e., cost reduction, TCO). This reflects how certain responsibilities (i.e., do it myself vs. let others do it partially or fully vs. do it together) for these services may be outsourced to OEMs. Therefore, formal organization of the buying function, the nature of buyer–supplier relationships, and the general purchasing policies of the buying firm are key pieces of the puzzle for their adoption as well.

Bio

Luis Prato has more than 15 years of experience in turning technology into commercially viable services for critical infrastructure (e.g., mining, oil and gas, water supply, and marine) and manufacturing sectors. He has held different leadership roles at Schlumberger, Parker Hannifin, Aker Solutions, and GKN. Luis is an affiliated PhD candidate at the Rotterdam School of Management—Erasmus University and a member of the Purchasing and Supply Management (PSM) center at the Department of Technology and Operations Management. He can be reached at lprato​@rsm​.​nl.

Key objectives 1. Identify the most common challenges industrial natives face during the segmentation of markets when selling novel industrial smart services. 2. Explain the market segmentation framework most utilized in industrial markets. 3. Be exposed to an alternative market segmentation framework that includes the relationship between Value of data and Performance orientation of novel industrial smart services. 4. Distinguish the key criteria for buyers when purchasing novel industrial smart services 5. Describe how to employ the proposed alternative market segmentation framework.

Key summary points 1. Previous segmentation practices revolve around the use of secondary data (e.g., demographics, country, revenue, service application), although behavioral segmentation is acknowledged in B2B markets. Nonetheless, this paper presents an alternative in light of digital transformation in performance-based relationships. 2. The findings acknowledge that market segmentation is a complex and multidimensional process when selling and buying novel industrial smart services.

17 0 

C hapter

1 3

S mart S ervice a n d D ata Value

3. The buying process of industrial smart services offerings needs a holistic treatment, as most buying firms fail to comprehend the difference between buying products and buying services. This gap becomes a constraint on organizations’ appetite for the outsourcing of services. 4. OEMs rarely assess the purchasing dimensions (i.e., procurement/contracting strategy, outsourcing appetite, purchasing behavior decisions) of buyers during the process of market segmentation because this information is difficult to obtain.

Key questions 1. How prepared are organizations in regard to market segmentation in light of digital transformation on industrial markets to determine the potential acquisition of novel industrial smart services? 2. In what ways do firms recognize the value of data embedded in industrial smart services for their market segmentation? What do organizations understand about the performance orientation of novel industrial smart services, is there a clear consensus on what performance-based orientation means?

References Allmendinger, G., & Lombreglia, R. (2005). Four strategies for the age of smart services. Harvard Business Review, 83(10), 131. Beane, T. P., & Ennis, D. M. (1987). Market segmentation: A review. European Journal of Marketing, 21(5), 20–42. https://doi​.org​/10​.1108​/ EUM0000000004695. Bonnemeier, S., Burianek, F., & Reichwald, R. (2010). Revenue models for integrated customer solutions: Concept and organizational implementation. Journal of Revenue and Pricing Management, 9(3), 228–238. https://doi​.org​/10​.1057​/rpm​. 2010​.7. Cenamor, J., Rönnberg Sjödin, D., & Parida, V. (2017). Adopting a platform approach in servitization: Leveraging the value of digitalization. International Journal of Production Economics, 192(October), 54–65. https://doi​.org ​/10​.1016​/j​.ijpe​. 2016​.12 ​.033. Daft, R. L., & Weick, K. E. (1984). Toward a model of organizations as interpretation systems. Academy of Management Review, 9(2), 284–295. https://doi​.org ​/10​. 2307​/258441. Gilmour, P., Borg, G., Duffy, P. A., Johnston, N. D., Limbek, B. E., & Shaw, M. R. (1994). Customer service: Differentiating by market segment. International Journal of Physical Distribution & Logistics Management, 24(4), 18–23. Kotler, P. (2000). Marketing management: The millennium edition. Upper Saddle River, NJ: Prentice Hall. Lee, J., Ardakani, H. D., Yang, S., & Bagheri, B. (2015). Industrial big data analytics and cyber-physical systems for future maintenance & service innovation. Procedia CIRP, 38, 3–7. https://doi​.org ​/10​.1016​/j​.procir​. 2015​.08​.026. Lightfoot, H., Baines, T., & Smart, P. (2013). The servitization of manufacturing: A systematic literature review of interdependent trends. International Journal of Operations and

C hapter

1 3 

17 1

L uis P rat o Production Management, 33(11), 1408–1434. https://doi​.org​/10​.1108​/ IJOPM​- 07​-2010​ -0196. Luz Martín-Peña, M., Díaz-Garrido, E., & Sánchez-López, J. M. (2018). The digitalization and servitization of manufacturing: A review on digital business models. Strategic Change, 27(2), 91–99. https://doi​.org ​/10​.1002 ​/jsc​. 2184. Martin, H. H. (1997). Contracting out maintenance and a plan for future research. Journal of Quality in Maintenance Engineering, 3(2), 81–90. https://doi​.org​/10​.1108​ /13552519710167700. Martin, L. L. (1999). Performance contracting: Extending performance measurement to another level. Public Administration Times, 22(1), 1–8. Ng, I. C. L., Maull, R., & Yip, N. (2009). Outcome-based contracts as a driver for systems thinking and service-dominant logic in service science: Evidence from the defence industry. European Management Journal, 27(6), 377–387. https://doi​.org ​/10​.1016​/j​.emj​. 2009​.05​.002. Opresnik, D., & Taisch, M. (2015). The value of big data in servitization. International Journal of Production Economics, 165, 174–184. https://doi​.org ​/10​.1016​/j​.ijpe​. 2014​.12 ​.036. Porter, M. E., & Heppelmann, J. E. (2014). How smart, connected products are transforming competition. Harvard Business Review, November. https://hbr​.org ​/2014​/11​/ how​-smart​ -connected​-products​-are​-transforming​- competition. Porter, M. E., & Heppelmann, J. E. (2015). How smart, connected products are transforming companies. Harvard Business Review, October, pp. 1–39. Rymaszewska, A., Helo, P., & Gunasekaran, A. (2017). IoT powered servitization of manufacturing—An exploratory case study. International Journal of Production Economics, 192(October), 92–105. https://doi​.org​/10​.1016​/j​.ijpe​. 2017​.02​.016. Selviaridis, K., & Wynstra, F. (2015). Performance-based contracting: A literature review and future research directions. International Journal of Production Research, 53(12), 3505– 3540. https://doi​.org ​/10​.1080​/00207543​. 2014​.978031. Shapiro, B. P., & Bonoma, T. V. (1984). How to segment industrial markets. Harvard Business Review, 62(May), pp. 104–110. Stoll, O., West, S., Gaiardelli, P., Harrison, D., & Corcoran, F. (2020). The successful commercialization of a digital twin in an industrial product service system. In B. Lalic, V. Majstorovic, U. Marjanovic, G. von Cieminski, & D. Romero (Eds.), Advances in production management systems: Towards smart and digital manufacturing (pp. 275–282). Cham, Switzerland: Springer International. https://doi​.org​/10​.1007​/978​-3​- 030​-57997​-5​_ 32. Stoll, O., West, S., & Hennecke, L. (2021) Procurement of advanced services within the domain of servitization: Preliminary results of a systematic literature review. In S. West, J. Meierhofer, & C. Ganz (Eds.), Smart services summit: Digital as an enabler for smart service business development (pp. 73–81). Cham, Switzerland: Springer Nature. https://doi​.org​/10​.1007​/978​ -3​- 030 ​-72090 ​-2 ​_7. Suter, W. N. (2014). Qualitative data, analysis, and design. In W. N. Suter (Ed.), Introduction to educational research: A critical thinking approach (pp. 342–386). London: SAGE. https:// doi​.org​/10​.4135​/9781483384443​.n12. Ulaga, W., & Reinartz, W. J. (2011). Hybrid offerings: How manufacturing firms combine goods and services successfully. Journal of Marketing, 75(6), 5–23. https://doi​.org​/10​.1509​/ jm​.09​.0395. Van der Valk, W., & Rozemeijer, F. (2009). Buying business services: Towards a structured service purchasing process. Journal of Services Marketing, 23(1), 3–10. https://doi​.org​/10​ .1108​/08876040910933048. West, S., Wenting, Z., & Stoll, O. (2021). The contracting of advanced services based on digitallyenabled product-service systems. In A. Dolgui, A. Bernard, D. Lemoine, G. von Cieminski, & D. Romero (Eds.), Advances in production management systems: Artificial intelligence for sustainable and resilient production systems (pp. 228–237). Cham, Switzerland: Springer International.

17 2 

C hapter

1 3

C hapt e r 1 4

Three Considerations for Data Monetization and Value Creation in the Digital Age Bill Schmarzo

Introduction

It is all about data. Search Google for the term data and you will submerge with content. Data is now front and center in the world of digital transformations. That makes sense. Companies have focused their digital efforts on collecting as much data as possible through the billions of sensors that are now in operation across all applications in the consumer and business world. The focus on data collection is now old news. The new challenge ahead of digital professionals is to clean, connect, and mine the data to extract valuable insights to improve performance or generate innovations. Well, we are not there yet. Having more data does not make anyone digitally savvy. On the contrary! More data means more cleaning and more complications. In this chapter, I propose three considerations for data practitioners to start thinking more about data monetization. I am the Dean of Big Data, and I have helped dozens of organizations apply these useful principles to make sure data brings value and ultimately profit. Let us get started!

First consideration: The four stages of data monetization

Of all the weird things to happen in 2020, who would have ever contemplated a day when the price of a barrel of oil would go negative? But on April 20, 2020, and for the first time ever, the price of the main US oil benchmark fell to about −$37.63. Yes, the price of a barrel of oil fell so far that traders were paying buyers to take oil off their hands. You see, there are substantial costs—and risks—associated with just storing oil. And this analogy holds true for data as well. There is no value in just storing data. There are storage, management, security, backup, and governance costs as well as potential compliance and regulatory liabilities associated with just storing data. So, there are similarities between oil and data, as illustrated in Figure 14.1. Data, like oil, has latent value. That is, data and oil have potential value that is not yet realized. Possession of data, like possession of oil, in and of itself is not sufficient in DOI: 10.4324/9781003226192-18

17 3

B ill S chmar z o

Economic Similarities of Oil and Data If “Data is the new Oil”

...then “Analytics is the new Gas”

Oil is raw and is of little direct use

Data is raw and is of little direct use

However, Oil has Potential Energy

However, Datahas Potential Value

Gas (refined Oil) has 5x to 10x more Potential Energy than Oil

Analytics (refined Data) has 5x to 10x more Potential Value than Data

Burning Gas to create motion converts Potential Energy to Kinetic Energy

Applying data science to optimize decisions converts Potential Value to Kinetic Value

Kinetic Energy is the energy of motion; the movement of objects. Objects that are not in motion possess Potential Energy, which is converted to kinetic energy when some force (catalyst) acts upon the object to set it in motion.

Figure 14.1 Economic similarities of oil and data.

providing economic value. Data, like oil, must be used to convert that latent (potential) value into kinetic (realized) value. The value of data is only realized when you apply the customer, product, and operational insights buried in the data to optimize the organization’s key business initiatives and supporting use cases. To get value from data, one needs an economics-based monetization strategy where value is created in the ‘use,’ not the ‘possession,’ of the data. It is how you put that data to work that converts latent value into realized value, and that is the power of a data science value engineering methodology that focuses on identifying, validating, valuing, and prioritizing where and how the data can derive and drive new sources of customer, product, and operational value. It requires a management mindset focused on data monetization—the process of generating quantifiable economic benefits from the data.

Data: The source of economic value creation

The article ‘$21 Trillion in Intangible Assets Is 84% of S&P 500 Value’ (Berman, 2019) highlights the changing nature of ‘value’ as reflected in the market cap (market value) of today’s digitally proficient companies. The intangible assets of the five largest market cap companies in 2018 were five times ($25 trillion vs. $4 trillion) more valuable than their tangible (nonphysical) assets. The reason why Apple, Alphabet (Google), Microsoft, Amazon, and Facebook have an outrageously high percentage of their market cap reflected in intangible assets is not because they have so much data. Heck, companies like Yahoo, AOL, K-mart, and Sears had tons of data. The reason why the intangible assets of these companies represent such a staggering percentage of their market cap is that these companies are masters of extracting value out of these intangible assets, which include data.1 Traditional accounting GAAP rules struggle to articulate the value of intangible assets in determining the value of the modern digital company. Accounting uses a ‘value in exchange’ valuation methodology, which 174 

C hapter

1 4

T hree Co n siderati o n s f o r D ata M o n eti z ati o n

determines the value of an asset based on how much was paid for that asset. But that valuation methodology does not work well for intangible assets. Economics, on the other hand, uses a ‘value in use’ asset valuation methodology that determines the value of an asset based on how much revenue or ‘value’ the use of that asset can generate. That is why companies trying to excel in the age of digital transformation would be better served if they embraced an economics, instead of an accounting, value generation mentality.

Mastering the ‘four stages of data monetization’

Companies need a roadmap to help them exploit the unique economic behaviors of this intangible asset called data—an asset that never depletes, never wears out, and can be used across an unlimited number of use cases at near-zero marginal cost. Companies must transition their executive mindset from ‘data as a cost to be minimized’ to ‘data as an asset that will fuel the economic growth of the 21st century.’ Organizations need the ‘4 Stages of Data Monetization’ (see Figure 14.2). The ‘4 Stages of Data Monetization’ provides both a benchmark against which organizations can measure their data monetization effectiveness and a guide for helping organizations master data as the asset that will fuel the economic growth of the 21st century. Here is a description of each of the four stages of data monetization: ■

Stage 1: Data is a cost. This is the stage where data is a cost to be minimized. This stage reflects the increasing costs associated with the storage, management, and governance of the data, as well as potential regulatory and compliance risks/costs associated with not effectively managing or protecting your data. Bottom line: the ever-accelerating volume and variety of data are growing faster than the declining costs of data storage.

Stages of Data Monetization There is no value in possessing data. In fact, there are costs and potential liabilities associated with the storage of data. The value of data is only realized when you leverage the customer, product, and operational insights buried in the data to optimize the organization’s top-priority use cases 500

400

300

200

Stage 1: Data as a Cost Increasing costs and liabilities associated with possessing data

Stage 2: Data Monetization Exploration Initial use case POV’s build organizational support for data monetization

Stage 3: Data Monetization Value Realization

Stage 4: Data Monetization Value Acceleration

Re-use of data shrinks use case time-to-value and derisks projects

Data and Analytic enhances ripple through previous use cases

100

-

2

1

3

4

(100)

(200)

Bill Schmarzo “Big Data MBA” Curriculam

Version 1.0

Figure 14.2 Four stages of data monetization. C hapter

1 4 

175

B ill S chmar z o ■

■

■

Stage 2: Data monetization exploration. In stage 2, initial data monetization pilots start to take root driven by close, grassroots collaboration between business and IT leadership. This is the proof of values (POV) stage, where pilots around wellvetted use cases build organizational awareness and hands-on experience around the potential of an organization-wide data monetization effort. Stage 3: Data monetization value realization. The hiring of a chief data monetization officer lays the foundation for the rapid operationalization and subsequent governance of the organization’s data monetization efforts by driving data and analytics reuse and refinement. The reuse and refinement of the organization’s data and analytics accelerate use-case implementation time-to-value while derisking the projects. Stage 4: Data monetization value acceleration. Through the proactive management and governance of companies’ data and analytic assets, data and analytic enhancements ripple through previous use cases, causing a rapid acceleration in value realization. This is the stage where Apple, Alphabet (Google), Microsoft, Amazon, and Facebook live. This is also the foundation for the ‘Economic Digital Asset Valuation Theorem’ presented in Figure 14.3.

Second consideration: Creating a data strategy that delivers value

The internet and globalization have mitigated the economic, operational, and cultural impediments traditionally associated with time and distance. We are an intertwined global economy, and now we realize (the hard way) that when someone sneezes in some part of the world, everyone everywhere gets sick. We are constantly getting punched in the mouth, and while we may not be sure whence that punch might come next (pandemic, economic crisis, financial meltdown, climate change, catastrophic storms), trust

Schmarzo Economic Digital Asset Valuaon Theorem The more the data and analycs get used, the more accurate, more complete, more robust, more predicve and consequently more valuable they become Hi Effect #3: Economic Value Accelerates

Cumulave Value ($$$)

• Refining Analyc Module effecveness ripples thru previous use cases that use same Analyc Module – The Google Effect

Effect #2: Economic Value Grows • Data and analyc module re-use shrinks me-tovalue and de-risks use cases

Effect #1: Marginal Costs Flaen • Reusing “curated” data and analyc modules reduces marginal costs for new use case (no data silos or orphaned analycs)

Lo Number of Use Cases

Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.3 The Schmarzo economic digital asset valuation theorem.

176 

C hapter

1 4

T hree Co n siderati o n s f o r D ata M o n eti z ati o n

me when I say that in a continuously transforming and evolving world, there are more punches coming our way. So, how does one develop and adapt data and AI strategies in a world of continuous change and transformation? It is not that strategy is dead, but it is that strategy—like every other part of the organization and the world—needs to operate in an environment of continuous change and transformation.

Data strategy: Tragic mismatch in data acquisition versus monetization strategies

Organizations spend hundreds of millions of dollars in acquiring data as they deploy operational systems such as ERP, CRM, SCM, SFA, BFA, e-commerce, social media, mobile, and now IoT. Then they spend even more outrageous sums of money to maintain all the data whose most immediate benefit is regulatory, compliance, and management reporting. No wonder CIOs have an almost singular mandate to reduce those data management costs. Data is a cost to be minimized when the only ‘value’ one gets from that data is regulatory, compliance, and management risk reduction. Organizations have a tragic mismatch in their investments in acquiring and managing data versus their investments in monetizing data (see Figure 14.4). I have always emphasized the unique economic characteristics of data—data never wears out, never depletes, and can be used across an unlimited number of use cases at near-zero marginal cost. But if data possesses this outsized economic potential, why is there a tragic mismatch in organizations’ investments in acquiring and managing data versus their investments in monetizing data? As in most organizational transformations, step 1 starts by reframing the conversation.

Mismatched Data Acquisition vs Monetization Strategy Organizations have a tragic mismatch in the investments that they make in acquiring and managing data versus the investments that they make in monetizing data. DATA INVESTMENT STRATEGY $1’s of millions invested in advanced AI / ML analytics

$10’s of millions invested in reporting and dashboard apps $100’s of millions invested in operational, web, mobile and IOT data acquisition

DATA VALUE MONETIZATION REALITY

$

$$$$$

$$$

$$$

$$$$$

$

Data yields customer, product, operational insights to optimize use cases and create revenue opportunities

Data used for regulatory compliance and management reporting

Data is a cost to be minimized

Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.4 Mismatched data acquisition versus data monetization strategies. C hapter

1 4 

17 7

B ill S chmar z o Reframing the data strategy conversation

Many moons ago, I stated, ‘Organizations do not need a Big Data strategy; they need a business strategy that incorporates Big Data.’ To quote the Harvard Business Review: ‘The problem is that, in many cases, big data is not used well. Companies are better at collecting data—about their customers, about their products, about competitors—than analyzing that data and designing strategy around it’ (Gerdeman, 2017). Too many organizations are making big data, and now IoT, an IT project. Instead, think of the mastery of big data and IoT as a strategic business capability that enables organizations to exploit the power of data with advanced analytics to uncover new sources of customer, product, and operational value that can power the organization’s business and operational models (see Figure 14.5). To exploit the unique economic value of data, organizations need a business strategy that uses advanced analytics to interrogate/torture the data to uncover detailed customer, product, service, and operational insights that can be used to optimize key operational processes, mitigate compliance and cyber-security risks, uncover new revenue opportunities, and create a more compelling, more differentiated customer experience. But exactly how does one accomplish this? By focusing on becoming value-driven, not data-driven.

Adopting a value engineering mentality

The value of data is not having it (data-driven). The value of data is using it to derive and drive new sources of ‘wealth’ (value-driven). To exploit the economic potential of data, executives must transition from a data-driven mindset to a value-driven one that focuses on exploiting data to uncover new sources of customer, product, and operational value (see Figure 14.6).

Big Data Business Model Maturity Index How Effecve is Your Organizaon at Leveraging Data and Analycs to Power your Business Models? CULTURAL TRANSFORMATION

Data Products

Key Business Processes Big Data Economics

Connuous Learning & Adapng INSIGHTS MONETIZATION

BUSINESS OPTIMIZATION

BUSINESS MONITORING

BUSINESS INSIGHTS

Prescripve Recommendaons

Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.5 Five steps to building a big data business strategy.

17 8 

C hapter

1 4

T hree Co n siderati o n s f o r D ata M o n eti z ati o n

Value Engineering Framework Business Initiative Stakeholders Decisions Analytics Data Architecture & Technology Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.6 Value engineering framework.

The heart of the value engineering framework in Figure 14.6 is the collaboration between the different stakeholders to identify, validate, value, and prioritize the decisions (use cases) that support the organization’s key business initiatives. It is these use cases that help you determine which data is most relevant (because not all data is of equal value)—to differentiate the signal from the noise buried in the data. Data may be the new oil and one of the most valuable resources in the world, but it is the analytic insights buried in the data that will determine the winners and losers in the 21st century. Data lake as collaborative value creation platform

If you do not care whether your data lake turns into a data swamp, then just go ahead, and toss your data into your unmanageable gaggle of data repositories. Mission accomplished! But if you seek to exploit the unique characteristics of data—assets that never deplete, never wear out, and can be used across an infinite number of use cases at zero marginal cost—then transform your data lake into a ‘collaborative value creation’ platform that supports the capture, refinement, and reuse of your data assets across the organization (see Figure 14.7). There are two important lessons I learned about data lakes: ■

■

Lesson 1. Use the organization’s key use cases to drive organizational alignment on identifying, capturing, and operationalizing new sources of customer, product, and operational value buried in the data. Lesson 2. Do not implement a rigid technology architecture that interferes with Lesson 1.

Modern digital companies realize that they are in the data monetization business, and the data lake is the platform for driving that data monetization. C hapter

1 4 

17 9

B ill S chmar z o

Data Lake 3.0: Collaborave Value Creaon Pla orm Improve Campaign Effecveness Increase Customer Loyalty

Opmize Store Remodeling

Improve Manager Retenon

Increase Customer Store Visits

DATA LAKE Reduce Customer Arion

Improve Hiring Effecveness

Increase Customer Crosssell

Improve New Product Introducons

Increase Customer Advocacy

Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.7 ‘Unlearn to unleash your data lake.’

Third consideration: Value engineering—The secret sauce for data science success

The business, economic, and social good that can be delivered courtesy of data science is almost unbounded; it has the potential to improve health care, public safety, transportation, education, environment, manufacturing, communities, and the overall quality of life. If what your organization seeks is to exploit the potential of data science to power your business models, then your next question is ‘How do I achieve that?’ And that is the topic of this third consideration.

The how: Data science value engineering

Let me introduce you to the data science value engineering process (see Figure 14.8). Let us drill into each of the steps of the data science value engineering framework— the ‘how to do it’ framework.

Step 1: Identify a key business initiative

As Stephen Covey discussed in his famous book The Seven Habits of Highly Effective People, ‘Begin with an end in mind.’ We have found that the how conversation must begin with a focus on the organization’s key business initiatives: that is, what is important to the business over the next 12 to 18 months. Your organization may have business initiatives such as these: ■ ■ ■ ■ ■

18 0 

reduce inventory costs improve supply chain reliability reduce unplanned operational downtime improve customer retention improve yield C hapter

1 4

T hree Co n siderati o n s f o r D ata M o n eti z ati o n

Data Science Value Engineering Framework Business Iniave Stakeholders Decisions Predicons Data & Instrumentaon Architecture & Technology Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.8 Data science value engineering framework. ■ ■

improve ‘first time fix’ improve supply chain reliability and quality

These are all most excellent business initiatives. You just need to invest the time to understand and research them thoroughly including the business, customer, environmental, and operational benefits, and the metrics and key performance indicators against which progress, and success, will be measured.

Step 2: Identify key business stakeholders

Once you identify the targeted business initiative, we want to identify the business stakeholders who either impact or are impacted by the targeted business initiative. This should be at least four to five different organizations because you want diverse perspectives on how these organizations plan to address or support the targeted business initiative. We use Personas (a design thinking tool) to help us understand the stakeholders with respect to their work objectives, work environment, key decisions, questions, and impediments.

Step 3: Identify, validate, value, and prioritize the decisions

Now the money step! Yep, once we know the targeted business initiative and the key stakeholders, then we want to drive facilitated collaboration across the different stakeholders to identify, validate, value, and prioritize the decisions these stakeholders need to make in support of the targeted business initiative. We use the prioritization matrix to drive consensus across the different stakeholders on the top-priority decisions (see Figure 14.9). C hapter

1 4 

18 1

B ill S chmar z o

Prioritization Matrix

• Focusing on a targeted Business Initiative is critical as it defines the framework around which the relative value and implementation feasibility placement discussion can occur

Hi

A

Targeted Business Initiative F D

Business Value

• Prioritization Matrix drives collaboration between Business and IT stakeholders around the relative placement of each use case based upon its business value and implementation feasibility

Lo

B

E

C

Implementation Feasibility

Hi

Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.9  Prioritization matrix.

The prioritization matrix is the most powerful business alignment tool I have ever used. It works every time … if you do the proper preparation work and are willing to put yourself in harm’s way. After completing step 3, everything else is easy.

Step 4: Identify supporting predictions

For each of the top-priority decisions, you want to next identify the predictions that each stakeholder needs to make in support of those decisions. Sometimes it is easier, when working with the business stakeholders, to ask them what questions they need to answer to support their key decisions. Then it is a simple process of converting those questions into predictions. To help organizations frame or understand the potential of data science, I start my customer conversations with a simple question: How effective is your organization at leveraging data and analytics to power your business models? Figure 14.10 shows some questions and supporting predictions from an agriculture example. For example: ■ ■

‘What were revenues and profits last year?’ (the question) converts to ‘What will revenues and profits likely be next year?’ (the prediction). ‘How much fertilizer did I use last planting season?’ (the question) converts to ‘How much fertilizer will I likely need next planting season?’ (the prediction).

See, a pretty simple process. Step 5: Identify potential data sources and instrumentation strategy

The next step is to work with the business stakeholders to identify what data might you need to make those predictions. The trick for fueling data brainstorming builds on the 18 2 

C hapter

1 4

T hree Co n siderati o n s f o r D ata M o n eti z ati o n

From Questions to Predictions Descriptive Questions

Predictive Analytics

(What happened?)

Prescriptive Actions

(What is likely to happen?)

(What should we do?)

What were revenues and profits last year?

What will revenues & profits likely be next year…?

Plant X and Y crops across N acres

How much fertilizer did I use last planting season?

How much fertilizer will I likely need next planting season…?

Pre-order X amount of fertilizer at 5% discount

When will my equipment likely need maintenance next month…?

Service your harvester and tractor #2 in January

How much downtime did I have last month due to unplanned equipment maintenance?

How many workers will I likely need next month and when will I need them…?

How many workers did I use last month?

Hire X number of workers for Y days

Bill Schmarzo “Big Data MBA” Curriculam

Figure 14.10 Transitioning questions into predictions.

‘predictions’ identified in step 4. We simply add the phrase ‘and what data might you need to make that prediction?’ to the prediction statement. For example: ■

What will revenues and profits likely be next year … and what data might you need to make that prediction? The data source suggestions might include commodity price history, economic conditions, trade tariffs, fertilizer and pesticide prices, weather conditions, fuel prices, and more.

Data Value Assessment Assesses relative business value of each of the data sources vis-à-vis the specific business and operational Use Cases Increase Store Traffic Local Events

Increase Store Traffic Loyalty

Point of Sales

3

0

0

4

4

3

3

Market Baskets

2

2

3

1

3

3

3

Store Demographics

3

0

0

3

3

3

2

Local Competition

1

3

3

0

1

2

2

Store Manager Demo

1

1

2

0

2

2

1

Consumer Comments

4

0

0

3

4

4

1

Social Media

2

0

1

2

3

2

0

Weather

1

0

0

2

0

1

1

Local Events

2

2

1

1

2

1

1

Traffic

1

3

2

1

3

2

Use Cases

Increase Shopping Bag Revenue

Increase Corporate Catering

Bill Schmarzo “Big Data MBA” Curriculam

Increase Improve New Non-corporate Product Catering Introductions

Low

High

Improve Promotional Effectiveness

0 0

4

Low

High

Figure 14.11 Data value assessment matrix example. C hapter

1 4 

18 3

B ill S chmar z o

How much fertilizer will I likely need next planting season … and what data might you need to make that prediction? The data source suggestions might include pesticide and herbicide usage history, weather conditions, crops to be planted, pest forecasts, soil conditions, and more.

■

In the end, we will get a matrix of data sources mapped to each key decision (use case) that we can use to prioritize our data and instrumentation (IoT sensor) strategy (see Figure 14.11).

Step 6: Identify supporting architecture and technologies

Finally, we will need big data and IoT architecture and technologies on which we can build the solution that delivers business value. For example, in an IoT architecture, one will need to consider the architecture and technology choices at the edge, platform (sometimes also referred to as ‘fog’), and enterprise (or cloud) levels. While the architecture and technology choices and integration are never easy, at least you’ll understand what technologies you will need and what technologies you won’t need.

Bio

Bill is Chief Data Monetization Officer | Recognized innovator, educator, practitioner in Data Science, Design Thinking | Creator Big Data MBA | Author of four books including ‘Economics of Data, Analytics and Digital Transformation.’

Key objectives 1. Understand the concept and stages of data monetization. 2. Connect the topics of business strategy, value creation strategy, and data strategy. 3. Emphasize the need for a value creation mindset as a strong connector to a data mindset.

Key summary points 1. The economic digital asset valuation theorem provides a framework for how digitally literate organizations can exploit the unique economic characteristics of data and analytics—assets that not only never deplete but increase in value the more they are used—to reduce marginal costs, increase marginal revenues and profits, and create continuously learning and adapting assets that accelerates the value of these digital assets. 2. While most organizations have a data strategy, there is a tragic misalignment between the investments made in data acquisition and storage versus the investments made in data monetization.

18 4 

C hapter

1 4

T hree Co n siderati o n s f o r D ata M o n eti z ati o n

3. Data may be the new oil and one of the most valuable resources in the world, but it is the analytic insights buried in the data that will determine the winners and losers in the 21st century. 4. If what your organization seeks is to exploit the potential of data science to power your business models; then the data science value engineering framework provides how the organization can do it. 5. The value engineering framework starts with the identification of a key business initiative that not only determines the sources of value but also provides the framework for a laser focus on delivering business value and relevance. 6. The heart of the data science value engineering framework is the collaboration with the different stakeholders to identify, validate, value, and prioritize the key decisions (use cases) that they need to make in support of the targeted business initiative.

Key questions 1. How do you get started with the value creation discussion with those in charge of data management? 2. What are the key activities or actions to move along the monetization stages and accelerate maturity development? 3. Who oversees developing the value engineering mindset when most of the work is very technical in nature?

Note 1 Not all intangible assets are composed of data. Other items included in intangible assets are patents, intellectual property, royalty agreements, brand equity, social media influence, goodwill, licensing, public access and use rights, customer lists, supplier agreements, franchising agreements, customer relationships, software licenses, and, of course, data.

References Berman, B. (2019). $21 trillion in intangible assets is 84% of S&P 500 value. IP Closeup, June 9. https://ipcloseup​.com​/ 2019​/ 06​/ 04​/ 21​-trillion​-in​-u​- s​-intangible​-asset​-value​-is​- 84​- of​- sp​-500​ -value​-ip​-rights​-and​-reputation​-included/. Gerdeman, D. (2017). Companies love big data but lack the strategy to use it effectively. Harvard Business School, Working Knowledge, August 21. https://hbswk​.hbs​.edu ​/item ​/companies​ -love​-big​-data​-but​-lack​-strategy​-to​-use​-it​- effectively. Schmarzo, B. (2019). Why tomorrow’s leaders MUST embrace the economics of digital transformation. LinkedIn, March 4. https://www​.linkedin​.com ​/pulse ​/why​-tomorrows​ -leaders​- embrace​- economics​-digital​-bill​-schmarzo/.

C hapter

1 4 

18 5

C hapt e r 1 5

The Economics of AI How to Shift Data Projects from Cost to Revenue Center Alex Fournier and Claire Gubian

Introduction

In the coming years, the ability of organizations to pivot their activities around enterprise AI will fundamentally determine their fate. Those able to efficiently leverage data science and machine learning techniques to improve business operations and processes (as well as to find new business opportunities) will get ahead of the competition, while those unable to shift will fall behind, swept away with the tide of rising costs and diminishing revenue. Three out of four C-suite executives believe that if they do not scale artificial intelligence (AI) in the next five years, they risk going out of business entirely. (Accenture, 2019)

Of course, the key word here is efficiently; it is not enough for organizations to simply leverage enterprise AI techniques at any price. Eventually, for enterprise AI strategy to be truly sustainable, one must consider the economics: not just gains, but cost.

The economics of AI

When organizations take their first steps into the enterprise AI world, the most common technique is to begin with a finite list of select use cases, ideally optimized for a balance between difficulty in execution and potential impact. The initial entry point might be costly in terms of technologies and change management, but assuming that the use cases are operationalized, the economic value is positive. In fact, Accenture’s ‘AI: Built to Scale’ uncovered that companies in the study that use this multiuse-case approach to get started report nearly three times the return from AI investments compared with companies pursuing siloed proofs of concept. DOI: 10.4324/9781003226192-19

18 7

A lex F o ur n ier a n d C laire G ubia n

Real-world example GE Aviation focused specifically on its self-service data program as a first step in enterprise AI with a goal of scaling and streamlining (i.e., making processes smoother and business operations more efficient overall). With these first efforts, they have been able to quantify their efficiencies and savings via the self-service data program to the tune of millions of dollars (Saxena & Tudor, 2021).

But what happens after companies find success with this first list of use cases? Well, they tend to repeat the process, adding increased use cases. Getting to the 10th or 20th AI project or use case usually still has a positive impact on the balance sheet, but eventually, the marginal value of the next use case is lower than the marginal costs. There is a point in time when the economic value of enterprise AI decreases because ■ the marginal cost of the supplemental use cases is not decreasing ■ the marginal value of the supplemental use cases is decreasing (i.e., the first use case has more value than the Nth use case) ■ the marginal profit of supplemental use cases quickly becomes negative

One might see this analysis and conclude that the most profitable way to approach enterprise AI, then, is to only address the top five to 10 most valuable use cases and stop. But this does not consider the continued cost of maintenance of this core of AI projects. Adding marginal cost to the maintenance costs will generate negative value and negative numbers on the balance sheet. It is therefore economically impossible to scale use cases, and it is a big mistake to think that the business will be able to easily generalize enterprise AI everywhere by simply taking on increasingly more AI projects throughout the company. To continue seeing returns on investment (ROI) in AI projects at scale, taking on exponentially more use cases, companies must find ways to decrease both the marginal costs and incremental maintenance costs of enterprise AI.

The cost of enterprise AI

Before looking at how organizations can reduce costs, it is important to understand what some of those costs of enterprise AI are. Of course, there are obvious, tangible costs (like that of tools and technology), which should certainly be managed to successfully scale.

Cloud costs and ROI As organizations’ data teams grow and as more staff outside of data teams begin working with data, having a modern approach to architecture that allows for scaling up and down of resources is critical. Indeed, elasticity will be the name of the game in 2022.

18 8 

C hapter

1 5

T he E co n o mics o f A I

Even though the cloud is growing in popularity, most companies will take a hybrid approach, investing in AI platforms that sit on top of the underlying architecture to provide a consistent user experience for working with data no matter where it is stored. Yet in early 2019, The Information reported that more and more companies find themselves surprised by their rising cloud costs (Efrati & McLaughlin, 2019). Companies that do not actively develop a larger cloud strategy and manage cloud costs in 2022 will face an uphill battle to prove positive ROI with AI projects, racking up a bill that is not offset by the financial gains or savings from the projects themselves.

But it is the following fewer tangible costs that tend to bog down organizations’ efforts by adding up over time, hampering their ability to scale and profit from Enterprise AI

Data cleaning and preparation

By now, most have heard the adage that data scientists spend about 80% of their time finding, cleaning, and preparing data. Indeed, 43% of overall respondents to an AI Maturity survey by Dataiku in 2019 said that data cleaning and wrangling is ‘the most difficult or time-consuming part of data processes at my organization,’ including 63% of the respondents in the C-suite (Heidmann, 2019). What is most challenging about data cleaning and preparation is that it is a huge task—and therefore a huge cost in terms of employee time—that needs to be done for every individual use case or AI project. Data cleaning and preparation are critical parts of an AI project and, if not done well, can translate into inferior quality models. So, reducing this cost is not necessarily about simply discouraging time spent or outsourcing the work. Rather, it is about ensuring efficiency, putting systems in place that allow data to be found, cleaned, and prepared once, and then used a maximum number of times in different use cases. Instead, what often happens for organizations today, as Data Science Senior Director Chris Kakkanatt explains (n.d.), is costly repeated work across people, teams, and the wider organization. Operationalizing and pushing to production

In the process of operationalization, there are multiple workflows: some internal flows correspond to production, while some external or referential flows relate to specific environments. Moreover, data science projects are composed of not only code but also data (including code for data transformation, configuration and schema for data, public referential data, and internal referential data). That is why, to support the reliable transport of code and data from one environment to the next, they need to be packaged together. Consistent packaging, release, and operationalization are complex, and without any way to do it consistently, it can be extremely time-consuming. Dataiku surveyed more than 200 IT professionals, asking, ‘On average, how long does it take to release the first version of a machine learning model in production?’ More than half said ‘between three and six months.’ C hapter

1 5 

18 9

A lex F o ur n ier a n d C laire G ubia n

That is a massive cost not only in person hours but also in lost revenue for the time the machine learning model is not in production and able to benefit the business. Multiply this not by one model but by hundreds, and the cost is debilitating.

Data scientist hiring and retention

Data scientists by nature are curious and entered the field to make a difference and to have an impact. They are also driven by efficiency, which means they do not like to do things twice if they do not need to. If all data scientists at the organization are doing is spending time playing with data in a sandbox (never seeing their projects in production having an impact on real data), spending time on data cleaning and prep instead of problem-solving or cutting-edge technologies, or doing repetitive work, they are not going to be very happy in their jobs, and, in turn, the company will spend a lot of money dealing with constant turnover. Reducing this cost is a matter of proper tooling: providing the resources for staff to capitalize on past projects and reuse work.

Model maintenance

Machine learning models are not like software code where they can be put in production once and work, untouched, until something about the system fundamentally changes. Data is constantly changing by nature, which causes models to drift over time. That means that continual AI project maintenance cannot be ignored (or at least not without an effect on profit). Depending on the use case, the model can become less and less effective in a best-case scenario; in the worst case, it can become harmful to the business. Maintenance becomes even more challenging the more use cases the company takes on, which drives up the costs even further. MLOps has emerged as a way of controlling the cost of maintenance, shifting from a one-off task managed by a different person— usually the original data scientist who worked on the project—for each model into a systematized, centralized task. One additional subdimension under model maintenance is the cost of maintaining infrastructure. That is, the maintenance of models necessarily requires maintaining the infrastructure on which they run. Given the speed at which the average organization’s infrastructure stack is evolving, this can become costly (in terms of both time and money) quickly.

Complex technological stacks

It is not just infrastructure that needs to be maintained; all technologies in the AI space are moving at the speed of light. That means switching from one to another happens often, and when it does, it can be costly. Also, in large organizations, different teams or different geographies might be using completely different technologies altogether. Without the ability to stitch together the larger technology picture—and allow reuse and sharing of knowledge across these teams or geographically—things can get even more expensive at scale. 190

C hapter

1 5

T he E co n o mics o f A I Capitalization and reuse

Common sense and economics tell us not to start from nothing every time, and that is exactly the principle behind reducing costs associated with data cleaning, preparation, operationalizing, model maintenance, and even hiring woes. Reuse is the simple concept of avoiding rework in AI projects, from minute details (like code snippets that can be shared to speed up data preparation) to the macro level (like ensuring that two data scientists from distinct parts of the company are not working on the same project). Capitalization in enterprise AI takes reuse to another level—it is about sharing the costs incurred from an initial AI project (most commonly the cost of finding, cleaning, and preparing data) across other projects, resulting in many use cases for the price of one. But how exactly do reuse and capitalization ensure scale? By increasing the number of use cases addressed with AI projects while reducing the impact of the costs outlined above. For example, say the business has a list of four uses cases in mind to start their enterprise AI efforts. In addition to these four, of course, there are lots of other potential use cases across the business. Capitalization means that while tackling these larger, high-priority use cases, the organization can also take on lots of other smaller use cases by reusing bits and pieces, eliminating the need to reinvent the wheel with data cleaning and prep, operationalization, and monitoring, and—in doing all of that—ensuring that data scientists are happy, spending their time on high-value tasks. It can also spur the discovery of hidden use cases. By capitalizing on the work of existing projects to spin up new ones, teams might find previously untapped use cases that bring a lot more value than expected, opening businesses to new possibilities (and sources of profit or savings). In addition, the surfacing of these hidden use cases via reuse often comes from the work of analysts or business users; in other words, it is one of the keys to unlocking data democratization, where it is not just data scientists that are bringing value from data. This idea often goes together with self-service data initiatives. Capitalization and reuse sound easy in principle, but in practice, they require strong, enterprise-wide, centralized processes where ■ ■ ■ ■ ■ ■ ■

people can easily access information, including who is working on what projects people can transparently consume things done by others (including seeing data transformation, models, etc.) people can take, reuse, and adapt work done by others data experts can capitalize (and monitor) a portfolio of data treatments to be used across the organization data experts can easily build and share projects to be used by others people—whether coders or not—can work efficiently in their preferred way data leaders can ensure the quality of AI projects, ensuring that capitalization and reuse are being used properly

To many organizations, this is a scary list—it is a level of transparency that many find uncomfortable. Some industries are hampered by regulations that make transparency more difficult, but certainly not impossible (see Dataiku, 2021). But it is the level of C hapter

1 5 

191

A lex F o ur n ier a n d C laire G ubia n

transparency one needs to execute on capitalization and reuse and profit from enterprise AI efforts. It is also worth noting that to build a solid and scalable enterprise AI foundation, these principles must be considered and built into the strategy from the start—not after the nth use case when costs are already impacting profits.

Capitalization and reuse with an AI platform

Data science, machine learning, and AI platforms are tools to enable enterprise AI by allowing people within the organization to scale, providing transparency and reproducibility throughout—and across—teams and all the dimensions previously touched on in the costs section of this chapter (Table 15.1).​ A tool that can enable capitalization and reuse should provide, at a minimum: ■ ■ ■ ■ ■ ■

Robust documentation so that contributors can explain what has been done in a specific project via wikis, to-do lists, versioning, activity logs, and so forth. A built-in, centralized, and structured catalog of data treatments (from data sources to data preparation, algorithms, and more) for easy consumption. The possibility of grabbing parts of data projects and inputting them into new projects or mixing components of two different projects together. The possibility for advanced users to package data products as plug-ins to be used by others without the need to understand all the underlying complexities. An advanced console to monitor usage, versions, and quality to ensure easy and efficient operationalization. The possibility for automating scenarios using complex triggers as well as automating test production and deployment.

Table 15.1  AI

mitigation by cost category

Cost

Mitigated with an AI platform via …

Data cleaning and preparation

Reuse of already cleaned and prepared data across projects as well as between personas (i.e., data scientists can use data prepared by analysts). Reuse from design to production (i.e., without the need to recode models and pipelines from scratch to operationalize). Reuse of project elements across users, allowing data scientists to spend more time on higher-value tasks (which also happen to be more interesting for them, which means lower risk of brain drain) as well as the ability to code in preferred languages and leverage open-source tools. Automated scenarios, monitoring, and reuse of infrastructure across technology stacks. Freedom to reuse and adapt even across changes in technology with Dataiku as an abstraction layer, freeing people from the underlying technology.

Operationalizing and pushing to production Data scientist hiring and retention

Model maintenance Complex technological stacks

192

C hapter

1 5

T he E co n o mics o f A I

An AI platform provides all these things while adding one extra layer: a user interface accessible to anyone on a data team, from data scientist to beginner analyst. True inclusivity and democratization of data efforts bring capitalization and reuse to another level, as it is no longer just a question of data scientists, but of reuse by everyone across the organization.

Conclusion

Indeed, as previously mentioned, and worth reiterating, capitalization and reuse sound easy in principle but are difficult in practice. Besides the previously discussed transparency strategies for building a solid foundation of reuse, what are the next steps? What does it take to really execute? It comes down not only to having the right technology in place but also to the right foundation of people and processes. Organizations that see real success from reuse and capitalization begin with a strong data science center of excellence, which can help enormously at the start by establishing best practices and ensuring that people across the business—whether data scientists, analysts, or business users— are following them.

Bios

Alexis Fournier is Regional VP, EMEA, AI Strategy at Dataiki. He began his career as a data scientist in the telecommunications industry before joining an international organization, where he applied this expertise to economic research. Following this international experience, Alexis worked at SAP to support many customers in their innovation journey around data science. Alongside Dataiku’s customer teams, Alexis supports organizations in understanding the value of AI in the enterprise and its processes, as well as on the different paths to everyday AI. Claire Gubian leads the Business Transformation practice at Dataiku, which helps customers accelerate their transformation thanks to AI. Claire is a seasoned leader who spent most of her career in management consulting, advising large organizations on their digital transformation, and at PayPal, where she was leading the peer-to-peer payments product line, notably during the mobile revolution. Claire is enthusiastic about how large organizations, but also everyone, embrace change and transform their way of working and making decisions thanks to technology and data. Claire has traveled the world, and what she loves the most about her job is connecting the dots and sharing best practices across multiple geographies and industries.

Key objectives 1. Dive into the economics of AI: why scalability is not simply a matter of more use cases. 2. Identify ongoing costs of enterprise AI: what they are and why they exist. 3. Focus on capitalization and reuse: how these techniques can be used to reduce costs and scale enterprise AI efforts.

C hapter

1 5 

193

A lex F o ur n ier a n d C laire G ubia n

Key summary points 1. Capitalization and reuse are essential drivers of strong AI economics. 2. To continue seeing ROI in AI projects at scale, taking on exponentially more use cases, companies must find ways to decrease both the marginal costs and incremental maintenance costs of enterprise AI. 3. The right AI platform can help capitalize on data and maximize data reuse while mitigating the overall cost of programs. 4. An AI platform provides all these things while adding one extra layer: a user interface accessible to anyone on a data team, from data scientist to beginner analyst. True inclusivity and democratization of data efforts bring capitalization and reuse to another level, as it is no longer just a question of data scientists, but of reuse by everyone across the organization.

Key questions 1. 2. 3. 4.

Should AI be considered a source of costs or revenues? What are the potential costs associated with enterprise AI programs? What are the potential challenges of deploying enterprise AI programs? How do you convince teams and leadership to think in terms of revenue growth and not pure costs?

References Accenture. (2019, November 14). AI: Built to scale. https://www​.accenture​.com​/ro​- en​/insights​/ artificial​-intelligence​/ai​-investments. Dataiku. (2021). Executing data: Privacy-compliant data projects. A guide for data teams. https://content​.dataiku​.com​/data​-privacy. Efrati, A., & McLaughlin, K. (2019, February 25). As AWS use soars, companies surprised by cloud bills. The Information. https://www​.theinformation​.com ​/articles​/as​-aws​-use​-soars​ -companies​-surprised​-by​- cloud​-bills. Heidmann, L. (2019, August 28). AI maturity survey: Where are we in the path to enterprise AI? Dataiku. https://blog​.dataiku​.com​/ai​-maturity​-survey. Kakkanatt, C. (n.d.). Scaling data science for a global analytics team [Video]. Dataiku. https:// videos​.dataiku​.com​/watch​/rdD​H B1D​H kQY​1kMT​x FXSHRx. Saxena, S., & Tudor, J. (2021). GE Aviation: From data silos to self-service. https://content​ .dataiku​.com ​/ge​-aviation​-ssa.

194

C hapter

1 5

C hapt e r 1 6

The Pricing of Data An Interview with Jian Pei, Simon Fraser University

Stephan Liozu: Thank you for taking the time to give us an interview. I’d like to start this with maybe introducing yourself, your area of research, and your affiliation for the readers. Jian Pei: I’m a professor at Simon Fraser University, Canada. My research area is data science, data mining, applied AI, database systems, and information retrieval. In the last 20 years, I have worked on many applications, facilitating data meeting the applications, and empowering data-intensive applications. Some of my algorithms and methods have been adopted or implemented by industry production systems and open-source packages. For example, the Spark MLlib machine learning packages. Stephan: How did you then move into the pricing area from your scientific area of interest? Jian: Among many applications, one important thing I observed is that it is hard for data to be moved from one owner to another user because people always have all kinds of concerns like privacy and security. We need a dynamic ecosystem. If you want to build a dynamic ecosystem that can cover many different parties, data owners, model builders, applications, service providers, and users, you need some mechanism that is simple and trustworthy. Data marketplaces are one such mechanism. Once we talk about marketplaces, then pricing becomes an essential element. However, when you talk about the marketplaces, it doesn’t have to be public marketplaces. Even within a large enterprise, we may still need to have some internal mechanism for data exchange or sharing. In such an internal system, many people may think ‘we don’t need pricing.’ But pricing is still the best mechanism to fairly evaluate the value of data and the utility of or the gain by using data. Therefore, all the efforts of data-intensive applications come to a point. We need some fairpricing mechanisms.

DOI: 10.4324/9781003226192-20

195

T he P rici n g o f D ata

Stephan: That’s interesting. I’ve done my own research on the pricing of data and the value of data. We always see a lot of papers and publications around new service models, new business models. But we don’t see a lot of published content about the actual monetizing of the value of the data itself. And it seems to me like there’s less interest in the business model of selling data—in the topic of pricing data sets and selling them. I’m not understanding why. Why is it that companies or data owners are not interested in selling data? Jian: There are several reasons that selling data directly or without any constraint becomes very risky and people don’t want to do it. First of all, privacy and security are a big concern. If you look at what privacy really is, it’s about the right to be forgotten. If you just sell the data, then you immediately lose the right to be forgotten because you don’t know who the new owner is and whether the new owner will propagate the data to some downstream transactions. This really constrains the possibility of selling data directly. In the future, if the government does not set a clear policy and regulations on data second usage, selling data directly will still be very challenging. In many situations, data pricing may be through data-based services. For example, if two companies using their own data collaborate with each other to provide a machine learning model, it is just some applications of the data, some summarization of data. Selling these kinds of models would be more flexible. Another example is customer targeting, like marketing campaigns. These kinds of services are heavily based on data. I would categorize them as some ways of selling data. Stephan: But what’s interesting when you read about all the consultants that write about data, they all say we have petabytes and petabytes of data, and that data is the new oil. It’s like when I have data, I’m rich! But if you look at how many companies are actually deriving value from the data, it’s a whole different ballgame. I see that contradiction where we’re still very focused on collecting data and not enough on extracting the value. How do you react to this? Jian: I think with data, there are several important things we need to get clear. Data itself, where the data is coming from, for example, how you collect the data, the context of data, the second use of data, and the capability of managing, processing, and making it for the next downstream applications—these several resources and capabilities are typically owned by different people. For example, when you collect data, you collect data from the original field or from many users. The people collecting the data may not have the context of the data. You contribute your transaction, but you may not know the big picture behind your transaction. You may not have the corresponding data on how to do it, understanding and extracting the context. You may not have the second use of the data. For example, you may not have the cross-market capability to do that. From the very high level, if you put all those things together, then definitely data is the new oil. The more data going through this whole pipeline, the more benefits you get. But the challenge and the opportunity are how you bridge the gaps among different parties and then the new oil is really going down from oil to gasoline, to different kinds of products. That’s exactly why we need the data pricing and marketplaces.

196

C hapter

1 6

T he P rici n g o f D ata

Stephan: That makes a lot of sense. Why do you think that area is so underresearched? You go to HBR and you read about the data-quality problem, you read about the processing of data, but that’s it. I mean, at the end of the day, we’re still focused so much on the collection of data. What’s going on that the industry or the data world is not talking more about this, you think? Jian: There are at least three reasons. The first is that data pricing and data marketplaces are an area beyond many traditional areas. It’s an interdisciplinary area. You need to understand competitions, you need to understand economics, you need to understand marketing, and how enterprises are running. It’s always difficult at the very beginning because you don’t know where to start. Second, the feasibility to run experiments in a practical and applied way is not as easy as collecting data. To clean up the data, you need to have a clear measure. But for data pricing, data marketplaces, sometimes measuring is very difficult, just like there are many ideas in economics and also finance, but very few of them can go for experiments because it is very costly to do an experiment about economics. You cannot just shut down one thing in the marketplace and see what’s the effect, right? The same happens to data pricing. To make it even more challenging, many enterprises are not ready for data-intensive applications yet. For example, the traditional manufacturing industry is not quite ready yet. The same is true for agriculture and even education. If you look at the students in the classroom, we don’t have enough mechanisms and channels to collect data about the progress of the students’ learning. The third issue is people’s mindset. In management, many people still don’t have the right mindset. They say data is a strategic resource, as important as your cash flow. However, in many places, the CIO is reporting to the CTO. Information looks like a technology to many people. Indeed, information is not technology. Technology helps to process information, but information itself should be a resource. This kind of misplacement makes it pretty challenging even nowadays. Stephan: Right now, you look at the mainstream consulting reports, they enjoy publishing the number of assets that are connected, the number of data points, and so forth. When are we going to see greater interest in the monetization and the pricing of data? How many years do you think we’re away from it? Is that 10 years, 20 years? Jian: I don’t see anything happening in the next few days, but we really will see substantial changes that may happen in the next five years. The reason is that a lot of substantial, dramatic changes are coming in the next few years. For example, the post-pandemic makes everything as a virtual business so popular and so important. Once you get virtual, then naturally a lot of data becomes available. I see in the next five to 10 years, there is a good chance that we will see a boost. Stephan: Do you think that the speed of the boost is impacted by legal, ethical, or technological issues? Is it an issue of mindset? Or is it all of the above before we see really a big boost? What are the biggest barriers, in your opinion? Jian: I think that the legal, ethical, and the enterprise mindset, the business part, are important. I can see that this is catching up quickly. For example, we had no privacy regulations in the past and it affected data adoption and acceptance. Many companies realized there was a red line not to cross by testing and learning. Now

C hapter

1 6 

197

T he P rici n g o f D ata

we can see more and more legal efforts that have been put forward to define such boundaries. This is very important. The lack of a clear and practical legal framework is the first barrier that needed to be somehow removed. Then there are technological barriers or enablers that are needed to be successful in pricing data. Machine learning is important. It is a key technique to transform data from different sources into different kinds of models for different applications. The second use of data becomes the business opportunities for enterprises and startups. This is very, very important, and we saw a huge trend before the pandemic. If you think about 5 years ago, there were tons of new companies running on second use of data. And through the pandemic, these have been slowed down. Everything was slowed down, but I expect that after the pandemic, this will come back even stronger because people will have more opportunities to switch gears, to change their paths that will create new opportunities. Then the next big thing is the information infrastructure and that will also help to facilitate the data collection and data transformation. Stephan: When people hear the term data pricing, they may ask themselves, ‘How do you define data pricing?’ What does that mean for the world of business? And then the second question they may have is, ‘How can we apply some of the scientific pricing concepts like value-based pricing, dynamic pricing, to the concept of data pricing?’ Jian: Data pricing should not be anything new from an economic point of view. We need to be scientific and follow principles of economics. Pricing measures the value and the utility of products to users. In that context, data pricing is very simple. It is just how to set the price appropriately so that suppliers and consumers can come together. But the real challenge in data pricing research is that data is different from traditional products, mainly in the much lower costs. For example, data has a very low, almost zero cost in reproduction. You can reproduce a data set, almost at no cost. This will create a huge challenge for economic theory. If you want to price a cup, it’s very easy. What’s the cost of making another one? And then you just assume you can make 1,000 such cups in one design and then you have the design costs divided by 1,000 plus the production cost, just to keep our example simple. But data sets are completely different because they have some special properties. I can reproduce a data set at almost no cost. When I pass the data to the next step, I may not know exactly the value I can gain from the second use. It is important for regulation, for policy, for legal efforts, to define and understand ‘second use.’ When we sell the data, the seller should exactly know what the data would be used for, what its second use is, and then anything beyond the second use should be prohibited or should be renegotiated because that is the boundary of privacy. Stephan: So then, if you know the context of the data and if you know the objective of the second use, is it easier, then, to do a little bit more of dynamic value-based pricing based on how the data is, where it comes from, what’s in it, and what it’s going to be used for? Jian: The challenge is the scale. If every factor is set and if the problem is stationary, it’s just back to whatever the regular pricing practice is. You can use game theory or any well-established principles. You can always assume there is some stable state.

198

C hapter

1 6

T he P rici n g o f D ata

However, the challenge is the scale. First, there are different second uses. A user may not feel comfortable or a patient going through thousands of possible second uses of her or his data. Then we need to define some better ways of pricing, which leads to new challenges. For enterprise, when you want to get the data, get the second use, how can you estimate the audience for the second use and the scale of second use? Like, Uber never knew how many users it would have when it was created. Stephan: But you could argue, then, if you go to a plant, a production plant, right, and you have data already in your plant, but you’re missing data and you’re going to call the data warehouse or go to a marketplace and say, ‘I’d like to get vibration data.’ Now I’m sure they’re going to qualify you. They’re going to ask you why, to try to evaluate the value of completing that data set. I’m sure you could develop an algorithm, a machine learning algorithm that could predict that value. Jian: The overall scenario you described is valid. Things may be even a little bit more sophisticated when we investigate the details. When you call it a data warehouse and say, ‘Oh, I want some data.’ For example, ‘I want some image data to enhance my current model, specifically this step of the process.’ Then, not all the image data will be useful to you because your model already has this capability. Your model already can tell the cat from the dogs, but you just want to distinguish between the British short-hair and the American short-hair. You want to look at those details. Once you want to look at the details, then you look at the microdata, you may step into the privacy zone. If I have 1,000 pictures about cats as a data owner, but 400 of them are about those short-hairs, do you want my data? And how much is my data worth? It becomes subtle. Stephan: Obviously, it’s not as easy as I described, but I’m sure this is where we’re going. I’m sure that the marketplaces and the warehouses, eventually, they want to optimize their pricing, right? But truly improving algorithms, right, for someone who wants to use data sets in their applications and the predictive power of algorithms. So how do we get there, then? My question is, how do we get to a better predictability by acquiring data and through the marketplace, for example? How do we know? Jian: My vision is that there won’t be a unifying, completely transparent, one-marketfits-all solution. Instead, you won’t call the data warehouse and say, ‘I want this data.’ Indeed, you call a few vendors. I need a better model to make the prediction so that when I see a cat, I immediately can tell what cat is. And those models will try to provide you the algorithms and the tools. Those models will also try to explore the data from the upstream data sources. This is what the product supply chain we have nowadays is. You don’t have a uniform global market where everything is exchanged. Instead, you have a very sophisticated supply chain network. I will expect data pricing, data marketplaces to be shaped in a similar style. Stephan: Okay. So, the buyers on the marketplaces will not be end users; they will be model builders. Jian: Yes, and it will be many different parties. A very important part is of course the data owners. The data owners can be customers, manufacturers, and all kinds of different places, as long as you can collect data. Then you have data collectors. Their task is just to collect the data. And then we get the data sets cleaned and

C hapter

1 6 

199

T he P rici n g o f D ata

labeled, some aggregated privacy protected as the very raw product of data. Then we pass to the next level and maybe some simple modeling vendors. They build the models. Models may be integrated together into some longer production lines, and they will be passed to enterprises and put into production. Stephan: So it sounds complex and long. Jian: Yeah. So that’s exactly what I mean. The scale is a real challenge for data pricing. If we have all the pieces already there and we just want to solve this pricing problem, it is not much different from the traditional product pricing. Stephan: Do you think that data warehouses are the future of data pricing and data monetization? At some point there’s going to be a few experts, big data house with marketplaces, or it’ll be more specialized by industry, by vertical, or by function. How do you see the organizing as an industry? Jian: I think it’ll be more specific to industry, to scenario, to optimization. I mean, the data industry will be more vertical, because vertical has a few benefits. Most importantly, it’s easier for marketplaces to be constructed, to be formed. It’s a dynamic ecosystem. And then it may be even associated with the current ecosystem of industry verticals. Stephan: If I’m a chief data officer today in a big company and I want to start really paying attention to my data and potentially, leverage, monetize my data, from your standpoint, what are three things I need to pay attention to? Jian: I think the first and most important thing is that you need to work with your CEO, your president, to make sure they treat data as a resource instead of just a subcategory of technology. You need to make sure that they don’t think, ‘oh, if you develop software, you also will get the data.’ They should see data directly, instead of seeing software or technology. Second, you should start to assess the value of data, its transformation, and the flow of value in your enterprise. You need to make sure that when data comes into your company from different sources, value is being considered and being put to work. When we say data value, we should measure that in terms of business impact. For example, in a search engine, we should clearly see, what’s the amount of time users spend on which part of the page? And which link the user clicks? And why we get this link? This link is built on the recommendation using this data. This is very important. If you don’t have this fine, data-driven supply chain there, or value chain there, people cannot understand how data is transformed from one place to the other and how it brings in value. The third thing is probably around value optimization. You improve your value from there. For example, a retail company can get the customer data and ship it to some vendor and then, of course, the vendor would be pleased to put some ads to you, create some business gain. But you need to think about whether this is the best way to run your business. Maybe you can just set up one more layer: that is, I can take the vendor’s requests and build a recommendation system, I can try to do cross-marketing. That will bring you more value. Stephan: We have the field of data and the field of pricing. Do you think eventually in the data space, all these data functions, we’re going to have data pricing managers? Some function which is fully responsible at some point, with the pricing and

200

C hapter

1 6

T he P rici n g o f D ata

monetization of data? Because right now they’re truly separate. No one is really talking about, in the pricing space, the pricing of data. Jian: I think that this is happening, but it doesn’t carry the title data pricing officer. They are marketing and product owners that monetize simple data sets in advertising or digital marketing. They have the data, and they know the users. They know the value of clicks and can put a price on that! So, it is already happening. But I agree with you that the next step probably will be going further to make this business happen everywhere in the whole enterprise and to formalize functions and titles. Stephan: I thank you for sharing your knowledge and for diving into the topic of data pricing. Where can readers find more information about your work? Jian: Thank you for having me in this inspiring discussion. Interested readers can find my publications on my webpage (www​.cs​.sfu​.ca/​~jpei). Recently, I wrote a few pieces on data pricing, including ‘A Survey on Data Pricing: from Economics to Data Science.’

Bio

Jian Pei is a Professor in the School of Computing Science at Simon Fraser University and an associate member of the Department of Statistics and Actuarial Science. He is a well-known leading researcher in the general areas of data science, big data, data mining, and database systems. His expertise is in developing effective and efficient data analysis techniques for novel data-intensive applications and transferring them to products and business practice. He is recognized as a Fellow of the Royal Society of Canada (Canada’s national academy), the Canadian Academy of Engineering, the Association of Computing Machinery (ACM), and the Institute of Electrical and Electronics Engineers (IEEE). Jian Pei is one of the most cited authors in data mining, database systems, and information retrieval. Since 2000 he has published one textbook, two monographs, and over 300 research papers in refereed journals and conferences, which have been cited extensively in the literature. His research has generated a remarkable impact substantially beyond academia. For example, his algorithms have been adopted by industry in production and popular open-source software suites. Jian Pei’s professional leadership is also demonstrated by his leadership in many academic organizations and activities. He was the Editor-in-Chief of the IEEE Transactions of Knowledge and Data Engineering (TKDE) in 2013–2016, the Chair of the Special Interest Group on Knowledge Discovery in Data (SIGKDD) of the Association for Computing Machinery (ACM), and a general co-chair or program committee co-chair of many premier conferences. He maintains a wide spectrum of industry relationships with global and local industry partners. He is an active consultant and coach for industry on enterprise data strategies, health care informatics, network security intelligence, computational finance, and smart retail. His industry partners and customers include Fortune Global 500 companies and unicorn startups. During his last sabbatical and extended study leave, he held business executive and technical leadership positions for two Fortune Global 500 companies.

C hapter

1 6 

201

T he P rici n g o f D ata

He received many prestigious awards, including the 2017 ACM SIGKDD Innovation Award (the highest award for technical excellence in data science), the 2015 ACM SIGKDD Service Award, the 2014 IEEE ICDM Research Contributions Award, the British Columbia Innovation Council 2005 Young Innovator Award, an IBM Faculty Award (2006), a KDD Best Application Paper Award (2008), an ICDE Influential Paper Award (2018), a PAKDD Best Paper Award (2014), a PAKDD Most Influential Paper Award (2009), and an IEEE Outstanding Paper Award (2007).

202

C hapter

1 6

SECTION 4

The Pricing of Platforms and Marketplaces



C hapt e r 1 7

Marketplace Monetization Methods Jacek Łubin´ski

Our investment strategy at Market One Capital (https://moc​.vc/) concentrates on early stage marketplaces and network effects platforms across Europe. As we’re reviewing hundreds of marketplaces projects every quarter, below I share my findings and thoughts on monetization methods used by marketplaces as well as the current trends we’re observing in that respect.

Monetization methods

Marketplaces have evolved in the last 25 years from light horizontal platforms to smart comprehensive products integrating additional value including payments, reviews, curation, smart matching, communication, complete value chain coverage, and so much more. The plurality of marketplace types gave rise to the proliferation of various monetization methods. Let’s begin with a summary of all the methods we’re observing.

Commission on all transactions

This is the most popular method—by providing value as the middleman, the marketplace gets a commission. The commission as percentage value of gross merchandise value (GMV) is usually called take rate or rake. The range for take rate is enormous—it can vary from less than 1% to more than 50%—and it depends on numerous factors: ■

What value is delivered by the marketplace. The higher and broader the value provided, the higher the justifiable rake. For managed marketplaces such as Packhelp* (https://packhelp​.com/),1 a custom packaging platform, or Welcome Pickups* (www​.welcomepickups​.com/), an in-destination travel marketplace, which takes responsibility for the whole process, product, and quality, usually the rake is 20% to 50%. For low-touch, lightly managed marketplaces, it is usually below 10%.

DOI: 10.4324/9781003226192-22

205

J A C E K Ł U B I N´ S K I ■

■

■

■

■

How much money suppliers make. It is rare that a marketplace makes more money on a transaction than a supplier. If suppliers make a single-digit or low double-digit return on a transaction, which is, for example, the case for many fintech marketplaces like EstateGuru (https://estateguru​.co/), the marketplace cut is usually in the low single digits. Cost of providing the product/service by the supplier. Shutterstock (www​.shutterstock​.com/), Fotolia (fotolia​.c​om), or other marketplaces of digital products usually have a very high rake, as there are practically no variable additional costs of selling a photo or a font to an additional buyer, so any additional revenue will be welcomed by a supplier. Set of services provided. For example, Lunching* (www​.lunching​.pl​/pl/), a B2B food-ordering marketplace, has one rake set for orders delivered by food providers and another, higher one for orders delivered by Lunching. Different rakes for different products. CGTrader (www​.cgtrader​.com/), a 3D model marketplace, takes a different cut for readily available stock models (zero marginal cost) than for custom models requested by the demand side (additional value of matching sides). Competition. As everywhere, competition puts pressure on the margins—a lower take rate means that a marketplace is more attractive for those paying a commission. That was one of the several reasons why Booking​.c​om famously grew so significantly at the expense of Expedia (12–15% vs. 25–30% commission)— it attracted more accommodation providers, which attracted more demand side (a crucial factor for marketplace success), created more liquidity, and ultimately increased the value of their marketplace.

Usually the commission is paid by a supplier, who gets additional revenue from a marketplace, but this is not always the case. For example, Collection Hub (https://collectionhub​.com/), a debt collection marketplace, is taking a cut from both debt collection agencies (a bigger part) and clients (a smaller part). The reasoning is that without Collection Hub, it would be much more difficult for the vast majority of businesses to find a debt collection agency in a remote country they don’t deal with on a regular basis (e.g., a Czech company trying to collect receivables from a Mexican counterparty). Also, Agriniser (www​.agriniser​.com ​/en), a grain marketplace, is introducing a commission from both sides of the transaction, as they replace ‘offline’ middlemen who do the same and take a bigger cut. We see cases where a commission takes the form of a predefined transaction fixed fee. This is usually the case where the transaction GMV does not vary much and the logic is to simplify the settlements. Baculum (www​.baculum​.es/), a marketplace for caregivers and nursing homes, gets a fixed fee for bringing in a patient. Viantro (www​ .doctari​.de​/viantro), a recruitment marketplace for doctors, had a fixed fee for placement but later changed it to a percentage of annual salary. Agriniser’s pricing is based on a fixed fee per ton of grain. There are also marketplaces where a commission has a floor in the form of a fixed minimum fee. This is usually used in microtransactions to preserve the unit economics of a transaction for a marketplace. For example, Symmetrical​.​ai* (https://symmetrical​ .ai/), a fintech marketplace and salary finance provider, charges a small commission on 206

C hapter

1 7

M arketplace M o n eti z ati o n M eth o ds

employee cash advances but with a minimum fee per transaction because of the associated costs. Some full-stack marketplaces like Sonar Home* (https://sonarhome​.pl/), dealing with residential real estate, or Spot a Wheel (www​.spotawheel​.com/), dealing with used cars, complete the transaction in two separate steps—first they buy the asset from the previous owner, and later they sell it to a new one. Therefore, the take rate is not known until an asset finds a new buyer—the marketplace takes full risk and full reward on making the arbitrage. By providing value for both previous sellers and new buyers, these marketplaces are able to monetize it and make attractive annual percentage rates (APRs) on capital. Generally, commission is the most popular and attractive monetization method in terms of absolute dollars that a marketplace can get from a transaction.

Commission on transactions generated through the marketplace

There are cases where a marketplace is unable to generate a commission on all the transactions. Imagine that you’re a hairdresser and you invite all your existing clients to book appointments through a new app—why would you let a marketplace get a commission from transactions with your clients? On the other hand, if a marketplace brings you a new client, who discovered your salon through the platform, a cut is understandable—this is simply your marketing cost. This is the case for the likes of Eversports* (www​.eversports​.at/), a sports and fitness booking platform, Freidesk (https://freidesk​.com/), a freight marketplace, and Versum (www​.versum​.pl/), a beauty marketplace. This method is very often combined with the monetization of software tools in SaaS-enabled marketplaces, which I describe below.

Recurring fee for tools provided (SaaS-enabled marketplace)

Monetization of tools provided is also among the most popular monetization methods for marketplaces overall. Usually these are essential workflow management and CRM tools that make the lives of suppliers a whole lot easier. With the companies mentioned above—Eversports is using this method with sports venues, and Versum does so with beauty salons. All relevant SaaS pricing considerations, which I won’t explore in detail in this chapter, apply here to optimize revenue per account and LTV. Cohosting​.​io (https://cohosting​.io/) offers a cross-selling travel and tourism tool for hospitality professionals like B&B hosts or hotels for a monthly fee. The professionals can then increase their cross-selling revenue by using the tool to sell tours and experiences, car rentals, luggage transfers, or any of their own services to their guests. ClickClickDrive (www​.clickclickdrive​.de/), a marketplace for driving lessons, provides a tool for managing driving schools, including CRM, calendars, and so on.

Payment fee

If a marketplace is not taking a commission on a transaction for some reason, it can still provide the possibility to pay for a product or service and charge a fee for it. For example, Eversports charges a payment fee if someone books a sports venue or a spot C hapter

1 7 

207

J A C E K Ł U B I N´ S K I

in a class and pays directly through the platform. It’s a convenient method of payment for some people, which streamlines providers’ operations as well; thus, getting a fee for this (which also covers the costs of a payment provider) makes sense.

Subscription fee

A subscription fee in a marketplace means that one of the sides pays a fixed monthly fee, sometimes simply to be on the platform and sometimes for the ‘full experience.’ In some cases, a subscription fee is used if there is a disintermediation risk and it’s difficult for a marketplace to take a commission out of a transaction. This is the case with Niania (www​.niania​.pl/), a marketplace for childcare, which charges users seeking a babysitter a fee for a premium account that gives you unlimited access, contact details of all the babysitters, and so on. It’s a similar story with Dobry Mechanik* (https://dobrymechanik​.pl/), a car repair and maintenance marketplace. Although they take a commission on part of the transactions, they also offer premium cards to car mechanics paid as a subscription fee. With a premium card, mechanics get a neat, organized website with contact details, photos, an appointment application form, premium visibility on the platform, and so forth. Dobry Mechanik also introduced a very interesting tiered pricing mechanism. Based on the number of leads they send to mechanics via telephone, the estimated conversion rate from phone call to visit at the mechanic based on their experience and the average value of the ‘order’ (visit) in the marketplace, they estimate the GMV generated for a mechanic and apply one of the tiered pricing packages based on the value of the estimated GMV. Subscription fees are also used by some of the peer-to-peer marketplaces. On Brainly (https://brainly​.com/), an edtech P2P marketplace, heavy users can get a premium account where they get unlimited access to the platform, including expert-verified answers, faster answers to their questions, and no ad interruptions. Some marketplaces, especially ‘bits’ marketplaces (those focused on digital products), use a subscription fee as a retention driver. This is the case with the previously mentioned Shutterstock and Crella (https://crella​.net/). Above a certain usage frequency, it is cheaper for demand-side users to pay a subscription fee than to purchase many single items. Since the marginal cost of selling an additional unit of a font or a business card template is nearly zero, this is still a very profitable model for these marketplaces. Meanwhile, Fachmistrz (www​.fachmistrz​.pl/), a handyman services marketplace, sells a monthly subscription fee to businesses that includes a package of services to be used within a month. Some companies on the demand side with a certain frequency of demand simply prefer this option to a typical pay-per-use model. Wagestream (https://wagestream​.com ​/us/), a salary finance provider, is charging companies a monthly recurring fee based on the number of employees enrolled in the system.

Fee for additional services provided by a marketplace or by third parties

Many marketplaces make money by offering additional complementary paid products or services provided by the platform itself or by third-party providers. Marketplaces gather a lot of data about their users and can offer additional tailored products or services addressing their specific needs. 208

C hapter

1 7

M arketplace M o n eti z ati o n M eth o ds

Boxmotions (www​.boxmotions​.com/), a managed marketplace for storage, and Liki24 (https://liki24​.com ​/uk/), a marketplace of medicines and other pharmacy products, offer delivery services for a fee. Boxmotions also sells packing materials and other add-on services such as insurance and moving. Spotawheel provides bank financing and insurance. Clients of Shypple (www​.shypple​.com/), a digital freight forwarder, can buy insurance for a container that is provided by an insurance company. Dot Residential (https://dotresidential​.com/), a managed residential real estate investment platform, gets, inter alia, a fee for arranging financing.

Listing fee

Fee for a listing is usually used by low-touch, old-generation marketplaces and classifieds. If you want to sell your car on OtoMoto (www​.otomoto​.pl/) or your real estate on Gratka (https://gratka​.pl​/nieruchomosci), you have to pay for your listing to be visible on the platform. But other marketplaces, such as Allegro (https://allegro​.pl/), are also using this method as an additional revenue source along with others.

Advertising revenue

Advertising revenue is another method used by some marketplaces as complementary, for example, Brainly. Also, driving schools buy ads on Super Prawo Jazdy (www​.superprawojazdy​.pl/). Pharma companies, which have difficulties advertising on Google and Facebook, are happy to buy ads on Liki24.

Pay per lead

Pay per lead is a method popular among marketplaces where the disintermediation risk is high and usage is infrequent. This is the case, for example, with Fixly (https:// fixly​.pl/), a B2C-focused handyman services marketplace, where supply side providers pay a fee for a lead. Once you establish a connection with a plumber you are happy with, you don’t need to come back to a marketplace—that is why lead-based models usually work for new connections only. Rynek Pierwotny (https://rynekpierwotny​.pl/), a residential real estate aggregator, also charges a fee per each qualified lead passed to real estate developers. We have seen this monetization method also used by some recruitment marketplaces, including GastroJob (https://gastrojob​.pl/), which gets paid for an organized meeting with a candidate. This might make sense, especially for temporary jobs with no significant qualifications required, when there is a high volume of candidates and conversion from meeting to recruitment is known and pretty stable. For more qualified jobs, getting a commission as a percentage of annual salary might be more attractive and also better aligns the interests of the marketplace and the companies recruiting people.

Promotion fee per premium listing

A premium listing is sometimes offered within a premium account, but it can also be a separate monetization method, among other ones. The likes of OLX (www​.olx​.com/) C hapter

1 7 

209

J A C E K Ł U B I N´ S K I

or Allegro rely heavily on monetizing this option, as suppliers want to stand out from the fierce competition and are often eager to pay for better visibility.

Data monetization fee

Although vast amounts of data collected by some marketplaces should be valuable to some businesses, we don’t see many projects using data monetization. In one example of a B2B marketplace working with CPG wholesalers and retailers, the platform sells reports to CPG brands about the market shares and sales potential of certain products based on aggregated and anonymized transactional data.

Asset management fee

For fintech marketplaces that support the management of funds or financial assets, a typical management fee as a low percentage of assets under management is usually applied. This is, for example, the case with Stableton Financial (www​.stableton​.com/), a marketplace for alternative investments, which provides advisors and their high-networth investors with curated alternative investment products.

Summary of monetization methods

These are all the methods I came across when preparing this chapter. Now let’s summarize and compare them briefly. Figure 17.1 includes all the monetization methods mentioned, subjectively ranked on two dimensions—attractiveness and popularity. We are most thrilled by marketplaces using monetization methods from the right side of the chart. The general idea is that the higher the attractiveness measured as a percentage of GMV of monetization methods used by a marketplace, the lower the GMV required to achieve a company of a certain size and valuation (measured by net revenues generated by a business). Second, it is possible to combine many monetization methods in one marketplace. Obviously, it’s less so for marketplaces with high double-digit rakes, but especially if your total combined rake is below 10%, look for opportunities to combine less attractive methods. Now let’s move on to the second part of the chapter, where I share some of the recent trends we are currently observing with respect to monetization methods.

Current trends in marketplace monetization Adding fintech capabilities

We see more and more signals that every software company becomes fintech-enabled. A basic financial product offering can be built in-house using integrations with financial services providers. Moreover, the dynamic proliferation and significant progress of fintech companies in the last five years, also at the infrastructure layer, will make it easier to integrate financial services and products into a tech company’s offering. What is more important is that these solutions should also be available for small and medium startups. Many marketplaces know a lot about their users—they gather thousands of data points about their situation in real time. Especially in cases of frequent and 210 

C hapter

1 7

C hapter

1 7 

Asset management fee

Payment fee

Figure 17.1 Marketplace monetization matrix.

Low attractiveness, low popularity

Advertising revenue

Promotion fee per premium listing

Lead fee

Listing fee

Low attractiveness, high popularity

Data monetization fee

Fee for additional services

Subscription fee

Recurring fee for tools provided

Commission on all transactions

Commission on transactions brought by marketplace

Popularity

High attractiveness, low popularity

Attractiveness (measured as % of GMV)

High attractiveness, high popularity

M arketplace M o n eti z ati o n M eth o ds

211

J A C E K Ł U B I N´ S K I

heavy usage, they can create a very detailed profile of users and their needs. Sharing this knowledge with fintech tubes in exchange for tailored financial products creates a win-win-win situation. Payments were the first step. Loans, factoring, bank accounts, and others will come next, as it will be easier to do tight integrations. As of now, we see fully managed marketplaces like the previously mentioned Spotawheel, Dot Residential, and others integrating financial services offerings with their products, but these are not fintech-enabled integrations yet.

Low or no initial monetization for B2B marketplaces in enormous old-school industries

B2B marketplaces are arguably one of the next waves. There are currently numerous B2B marketplaces changing the way communication and ordering are conducted between businesses and their suppliers (day one value). This includes the likes of Rekki (https://rekki​.com ​/en​-us), Choco (https://choco​.com ​/us/), and Katoo (www​ .katooapp​.com/) in the restaurant industry or Simplo (www​.simplo​.com​.tw​/index​ .php​?lang​= en), bex (https://bexapp​.de/), and Schüttflix (https://schuettflix​.com ​/de​/ de/) in the construction industry. I don’t know about all of them, but at least some of these are initially focusing on changing super-old habits of how ordering happens and creating new ones—and the less friction, the higher the chances of success with this. This usually means no/small monetization in the first place, and definitely not on the demand side. But the end game is that once you have the harder side of the marketplace onboarded (the demand side) with good stickiness and retention, then you will find ways to monetize the relationships, usually on the supply side. This strategy requires a lot of capital, but the processes in these industries are so oldschool, and total addressable markets are so enormous, that there are many great investors ready to deploy capital in many of these startups and take the risk, at least for now ;-).

Dynamic pricing

What was famously introduced by Uber in ride-hailing some time ago has become more popular among fully managed marketplaces. As a marketplace gathers a lot of data and is confident to increase the price for a service by even a small amount, this usually translates into pure margin, as suppliers’ costs are kept the same. This is not possible in all industries, but it’s worth considering, especially in industries where prices fluctuate anyway, such as travel and hospitality.

Bio

Jacek Łubiński is a software engineer turned finance professional turned VC. He has nine years of experience in VC, during which he analyzed thousands of startups, invested in more than 40 of them, and became a trusted supporter and sparring partner for many portfolio companies. Apart from digital platforms and software, he is interested in crypto. He holds three master’s degrees from Austrian and Polish universities (two in computer science and one in finance) and enjoys reading books and drinking yerba mate.

21 2 

C hapter

1 7

M arketplace M o n eti z ati o n M eth o ds

Key objectives 1. Understand the concept of monetization. 2. Dive into the concept of platforms and marketplaces in the context of growing digital transformation projects. 3. Discover the monetization tools for capturing value in digital platforms.

Key summary points 1. Pricing and monetizing in digital platforms and marketplaces are not business as usual. There are tools and techniques that need to be used to extract value. 2. Although digital platforms and marketplaces are booming, we do not see advanced monetization strategies. Most business professionals are not fully aware of all the monetization tools at their disposal to capture value through platforms and marketplaces. 3. Digital trends reveal that platforms and marketplaces are booming and are touching sectors that were considered traditional and ‘protected’ from disruption.

Key questions 1. What are the trends in digital transformations? 2. Are the pricing and monetization of platforms and marketplaces the same concepts? 3. What is important in platforms and marketplaces to capturing value?

Note 1 Companies marked with an asterisk are part of the Market One Capital portfolio.

C hapter

1 7 

21 3

C hapt e r 1 8

The Monetization of Marketplaces and Platforms in the Context of Web 3.0 Murali Saravu

Introduction

We all basically understand standard subscription billing. The way you pay for your Microsoft Office software, your monthly Netflix subscription, your Peloton account, and your coffee club membership essentially is the same—you pay a fixed fee, billed on a predetermined basis (usually monthly), to a company offering the goods or services. The roles are constant—the seller is a company; the buyer is a consumer. The sale is one-to-one. There are no middlemen, and there is no role-switching. This has been the de facto standard for billing over the past several years. In these common scenarios, each department within the finance function at a company has its own defined tasks to complete. Sales negotiate; sales and legal work out the contracts; billing sends out invoices; accounts receivable collects payments, and so forth. Again, roles are defined and static. But the world is changing, and things aren’t looking so simple in tomorrow’s landscape. The internet is on the cusp of a major structural shift toward Web 3.0. This transformation is being driven, in part, by a rebellion against the concentration of power in the hands of a few industry behemoths (Facebook, Amazon, Apple, Google) as well as by the emergence of blockchain and AI. What does the future of monetization have to do with Web 3.0? Everything. The very nature of how the web is structured is shifting. The rise of marketplaces and platforms will accelerate, bringing new opportunities for creators and for monetization. But wait. First, how did we get here?

Web 1.0: The read-only web

In the early days of the internet, the web mainly consisted of static pages of information you could visit. These pages lived on ISP-hosted web servers or on free web-hosting services. Visitors went to each page to consume content, and the creators were web DOI: 10.4324/9781003226192-23

215

M urali S aravu

developers who built websites to serve up content, which was mainly displayed in text or images. This lasted from approximately 1991 to 2004. During Web 1.0, monetization was straightforward. You paid for web hosting, banner ads, or storage. If you could download the application, you couldn’t see how it was constructed, or created, and you couldn’t change it in any way. You would download the application, but the code and back end were proprietary to the company. This marked a time when billing was simple to administer, simple to calculate, and simple to collect. For instance, in these early days of the internet, you might pay for a handful of servers to be placed in your offices where they physically housed your network and storage. You would pay for the hardware and the software one time. Just as Web 1.0 was static, so was monetization. One time, perpetual, and unchanging.

Web 2.0: The interactive and social web

Then came the internet most of us have lived with for nearly 20 years, Web 2.0. This is the interactive and social web, which brought the rise of Facebook (B2C) and SaaS companies (primarily B2B). As Web 2.0 evolved over the past 20 years, the way we paid for goods and services, and the way companies monetized those goods and services, also changed. On the B2C side, consumer data became the currency. While the apps themselves remained free, consumer data has been broadly monetized and sold to the highest ad bidder. On the B2B side, the popular idea of a subscription licensing model—as opposed to a single-use, perpetual license—gained momentum. This was radical when it began. Cloud computing suddenly meant that not everything needed to be hosted locally, and software-as-a-service became possible. Today, IT departments pay for everything from JIRA to Asana to Salesforce​.c​om via subscription. SaaS companies have even sprouted up for specific lines of business, such as sales/marketing (Hubspot), legal teams (LinkSquares), or finance (Chive). Monthly, straightforward subscription billing, consistent roles, and so on. While subscription billing was a leap forward, there was still centralized power, centralized billing, and centralized monetization with static roles. One-to-many, where the company is the operator, and the subscribers are the buyers.

In the latter part of Web 2.0, the sharing economy took hold

The sharing economy, popularized by Airbnb, Uber, Upwork, and others, offered a glimpse into a new way for companies and individuals to buy and sell goods and services. The sharing economy depends on a new economic model linking excess capacity to real-time demand. By connecting consumers to providers via online ecosystems or platforms, the sharing economy changes business-to-consumer models into consumerto-consumer models. By bringing decentralization to the exchange of goods and services, the sharing economy introduced a new way for businesses and consumers to realize value: via multisided marketplaces or ecosystems. Airbnb and the others created marketplaces or ecosystems that then, in turn, enabled individuals to exchange everyday goods and services. Even though they don’t directly offer goods or services to consumers, the marketplaces connect individuals who want to do business with each other. Billing,

216 

C hapter

1 8

M o n eti z ati o n i n the Co n text o f W eb 3 . 0

naturally, becomes significantly more complex here. How? Well, for one, the roles shifted. Additionally, changes are based on actual usage or consumption, not on a standard, one-size-fits-all monthly fee. In the case of Airbnb, there is the operator (Airbnb), the seller (the property owner), and the buyer (the person taking a great vacation). And the seller can be a buyer (property owners can go on vacation too!), and the buyer can be a seller (vacationers might have their own home to put on the marketplace), depending on the situation. The sharing economy introduced an ecosystem-driven, multiparty world where ■ ■ ■ ■ ■ ■ ■

the buyer can be the seller peer-to-peer transactions are supported distributed apps in the network are enabled pricing can be highly dynamic (seasonal prices on Airbnb, surge pricing on Uber, etc.) revenue sharing between seller, app provider, and network operator is possible the app provider can set the price the peer network is critical for transaction fulfillment

Invoicing in this world requires that the operator generates charge invoices and payout invoices. Or, with the right system, these would be combined to include fees owed, payments due, and surcharges to the marketplace operator. It isn’t just possible that these invoices would change over time; it is expected. This scenario requires robust back-end office technology that offers flexibility and efficiency, and that importantly offers highend automation for core activities, such as price setting, billing, rating, and invoicing. However, the power is still in the hands of the operator or business. For instance, consider the Salesforce marketplace, known as AppExchange. Party 1 can be both an app developer and an app consumer, and Party 2 can also have her own app—and make her own purchases. Both are buyers and sellers. But Salesforce gets income from both of them because it is still the ‘hub’ of the transaction. The traditional marketplace or ecosystem structure is complex, but it is essentially a hub-and-spoke model. As traffic, money, and products flow through the spokes, the hubs essentially hold money in escrow to make sure everyone is treated appropriately. Companies in the sharing economy take on the risk of operating in a more complex monetization world. ■ ■ ■ ■ ■ ■

How much usage was there? How and when do I collect the payment? How and when do I pay the seller? What if there was a discount to factor in? What if I typically pay on day 30 but I still haven’t been paid? What if the customer comes back and disputes the charge?

For taking on the management of these factors, as well as the ‘risk,’ the operator gets a slice of the payment from seller to buyer, and they also often charge the seller a participation fee.

C hapter

1 8 

217

M urali S aravu

According to PwC (2015), the sharing economy is set to reach $335 billion by 2025, and companies working in the sharing economies will grow by 2,133% in 12 years. Big business indeed. But wait. The journey isn’t over.

Web 3.0 is about to take hold. here is where things get really interesting

Web 3.0 comes from the rise of AI, decentralized data architecture, and edge computing. It is about taking the power out of the hands of the few and putting it in the hands of the many via blockchain. Remember the sharing economy example above? Web 3.0 is ushering in a world where all those variables are present—plus more. Web 3.0 is ■ ■ ■ ■ ■ ■ ■

verifiable trustless self-governing permissionless distributed and robust stateful native built-in payments

In the Web 3.0 scenario, there might not be an Airbnb or a Salesforce performing the role of operator. All buyers can be sellers, all sellers can be buyers, and there is no intermediary required. The marketplaces are not hub-and-spoke but distributed and cooperative. In the future of Web 3.0 monetization, anyone can be a buyer or a seller, or a bit of both, because the decentralized marketplace is built on blockchain. Moreover, in some instances, the ‘user’ won’t even be a human being but an IoT sensor, using services or perhaps consuming or creating content. These will be deployed by the tens of thousands and will be billed with limited (if any) human interaction. In these marketplaces of the future, (1) ledgers are shared; (2) transactions are immutable, which means they don’t change over time; and the network is both (3) trustless, meaning without a trusted intermediary, and (4) distributed. Here there are fewer delays, fewer errors, and more complexity is allowed for. All intermediate results are eliminated with two-sided peer-to-peer networks, where you don’t need a middleman company to hold your money in escrow. Companies will need to support multiple monetization models to grow their businesses, not just one-to-many or one-to-one, but also buying and reselling, and importantly for the future, multiparty. This model requires the ability to make peer-to-peer transactions on emerging blockchain-based networks that include revenue sharing, consolidated billing, and payments and payouts. To monetize, you need to set prices, going into revenue-sharing structures within contracts, which are implemented via technology and your infrastructure, and set up all the roles and sharing mechanisms. It’s a far cry from our one-to-one or one-to-many models of the past. Because this is all happening in real time, needs must be defined in advance via AI-based forecasting. There are essentially hundreds of millions of transactions taking

218 

C hapter

1 8

M o n eti z ati o n i n the Co n text o f W eb 3 . 0

place with no human interfacing happening—it is all planned interfacing. Companies are getting to the point where they are playing together and sharing revenue and doing it in a way that’s automated, fully monetized, and transparent and secure. An obvious benefit is that the right people are involved—and they’re being paid the right amount. But less obvious benefits are the speed of transaction time and the speed of optimization. Consider this example. You’ve been in the hospital, and it is time to go home. But your doctor wants to make sure you stay healthy, so she wants to monitor your progress at home for a bit. She suggests that you take home a remote patient monitoring device so the medical team can keep an eye on you. The device captures a start-stop point and calculates usage, and then the complex process begins. You can imagine the following players—the insurance company that needs to approve use of the device; the hospital which is responsible for its distribution, management, and billing; the device manufacturer; the software developer; and there is you, the patient, who only is paying for the times you are using the device. The hospital coordinates on what the insurance will pay, and then on the other side is billing you, the patient. And we all have different rates and fees depending on our individual insurance, so what might be true for you is not necessarily true even for your neighbor. The future is where all of these parties are enabled by a monetization partner, who provides a platform that comprises user interface and application logic. The partner handles the complexity inherent in these transactions such as service dimensions and price models and pricing rules, and the other parties are free to focus on their individual roles, and you’re able to focus on getting better. Here’s why invoicing/billing is challenging in these environments. For one thing, taxes. Consider this scenario. Company A buys a service from the marketplace to be used across various locations in the US. The service is provided by several service providers in the network, and the marketplace charges Company A for the service. However, tax calculations would necessarily factor in the tax jurisdiction of the individual transactions. And two-sided transactions will have sales tax levied on charges to Company A. For all payouts, the service providers who fulfilled the transactions would have income tax to address. Let me explain further. In marketplaces that offer two-sided transactions, a user of the service can also be a provider of a service. A simple analogy of this would be a household using solar panels. The house can consume electricity from the grid but also can provide generated electricity back to the grid. Utilities have custom solutions to support billing and invoicing of such use cases. With Web 3.0, these use cases are going to be part of many of the marketplaces. Unlike one-sided transactions where the user of the service pays for the service, which is an account receivable function from a finance perspective, two-sided transactions span multiple finance functions: accounts receivable for using the service and accounts payable for providing the service. Billing and invoicing systems have two types of customer accounts in a hierarchy to support these use cases and have necessary charge and payout business flows. Added to this, applicable sales tax would need to be handled in the transaction’s jurisdiction and where the user of the service is as well as the service provider’s location, both of which are not fixed locations, unlike in a one-sided transaction.

C hapter

1 8 

21 9

M urali S aravu The future benefits

To clearly spell out some of the benefits, we have the following: 1. Trust. People can have multiple roles and relationships—but there is immutable trust in transactions even at the most complex levels. You can trust your partners and customers, just as they can trust you. 2. Improved monetization. You can’t bill what you can’t measure, and too many companies see the future of marketplaces but don’t know how to do the monetization part. We’re on the edge of a major shift, and they’ll need to figure this out to stay competitive. 3. Streamlined revenue accounting. Revenue share between multiple parties involved in the fulfillment of transactions is accounted accurately and in compliance with the financial accounting guidelines So, how do you get started? Many of the steps in today’s concept-to-cash paradigm will need to be reimagined to support Web 3.0 services and two-sided transactions, and most of today’s product offerings are simply not designed to handle the complexity that comes with these future applications. Most of today’s monetization solutions would require incredible amounts of custom coding, and thus it is important to invest in monetization platforms that are already future-proofed. The steps are as follows: ■

■

■

■

■

220

Idea to offer. Product and pricing teams can model and visualize various pricing dimensions and come up with a pricing strategy and rate cards that can be offered to customers. For two-sided transactions, teams also have to factor in payout dimensions and the number of parties involved in fulfilling the service and have optimized payout models to ensure profitability. Quote to order. Depending on the focus (B2B or B2C), onboarding customers into the marketplace would require them to agree to a commercial contract that covers various services, usage commitments, and mutually agreed rate cards. Order to invoice. Customer usage is tracked and brought to a monetizable format. This is then rated based on the agreed contract terms and an invoice is generated. In the case of two-sided transactions, there will be invoices for charges and invoices for payouts for the customers. Invoice to cash. This step supports the payment and payout processes to ensure that customers are paying for the usage and getting paid for fulfilling the service. Also, this step would cover any adjustments and exception-handling processes to address customers’ inquiries. Billing analytics. It is very important to ensure that no revenue is left on the table. When it comes to usage and marketplaces, ensuring that all transactions are tracked and accounted correctly is critical to marketplace success. Verification with such complexity isn’t simple. Moreover, unlike recurring charges, when usage-based monetization is involved, forecasting the transaction volume and fees will become more critical and requires employing data science and AI/ML models and technologies. Robust billing analytics and forecasting engines should be part of the platform from the get-go.

C hapter

1 8

M o n eti z ati o n i n the Co n text o f W eb 3 . 0 Conclusion

To succeed in monetizing in Web 3.0, you need technology that can do three things: 1. An agile monetization platform—the ability to model and create rules for your services with various dimensions such as these: ■ features to monetize ■ charge ■ cost ■ payout ■ value ■ usage transaction definition 2. Go beyond FIAT currencies to support tokens and virtual currencies. 3. Invest in instrumentation and flexible service usage tracking and processing engines that can bring all usage data into charging/billing.

Bio

Murali Saravu is the founder and CTO of Monetize360, a technology company that delivers dynamic monetization solutions to sales and finance departments. The company simplifies and accelerates the implementation of pricing and billing automation for any scenario, regardless of complexity. Monetize360 was inspired by Murali’s 20+ years architecting robust billing platforms at enterprise organizations as wide-ranging as Cable &Wireless, Cisco, and Intuit. In addition to leading the technical team at Monetize360, Murali educates the market on the value of turnkey monetization solutions.

Key objectives 1. Understand how the evolution of the web has impacted monetization strategies over time. 2. Recognize how Web 2.0’s standard subscription billing has given way to more complex, and interesting, ecosystem or marketplace models brought about through the sharing economy. 3. Take a glimpse into the future of Web 3.0 and come to understand the myriad monetization benefits that will accompany the shift.

Key summary points 1. In a Web 3.0 world, marketplaces and multisided transactions will be more prevalent. 2. There will be a fundamental shift in monetization. This requires reimagining the current concept of cash systems and processes. 3. Flexible and agile monetization flows that can also be deployed as smart contracts are critical to success.

C hapter

1 8 

2 21

M urali S aravu

Key questions 1. Are businesses thinking about their services, about how they will play in the Web 3.0 world? For example, Facebook is transitioning to Meta. 2. Blockchain as a technology platform is becoming mainstream. Are you thinking about non-fiat currencies, figuring out how customers can buy services using tokens? For example, buying Tesla using Bitcoin and merchandise using Dogecoin. 3. How are you modeling your monetization, which is shifting from users to usage, in this new world?

Reference PWC. (2015). Sharing or paring? Growth of the sharing economy. https://www​.pwc​.com​/ hu​/en​ /kiadvanyok​/assets​/pdf​/sharing​- economy​- en​.pdf.

222

C hapter

1 8

C hapt e r 1 9

Pricing in Platforms and Marketplaces A Primer in Understanding All the Dimensions of the Pricing Problem and Opportunity in Marketplace Platforms Simone Cicero

Platforms offer complex pricing opportunities

One of the highlights of a recent conversation I had with James Currier was that, increasingly, as we move into more vertical and managed marketplace opportunities, where a mix of platform-provided services often complement the more ‘transactional’ nature of the marketplace, pricing becomes a complex strategic matter (Boundaryless, 2021). Indeed, when thinking about pricing, in a platform-marketplace setting, one needs to understand that there are at least three contexts where pricing is a fundamental question to be addressed, as visualized in Figure 19.1: ■ ■

■

The marketplace where transactions are facilitated and intermediated (what we call the transactions engine). The product side, a mix of services that support continuous learning and improvement (the learning engine, a concept we introduced in 2016) and the services that support the execution of the workflow (a sort of workflow engine), normally more focused on the side of the ‘providers’ of the platform. The extensions, the place where third parties, often developers, create extensions (such as plug-ins, apps, templates, libraries) to the workflow engine one provides to the entities in the ecosystem

Pricing in the marketplace

Pricing and monetizing the marketplace side of the platform value proposition have two main aspects. The first aspect regards the mechanisms used to price the services or products being traded in the marketplaces, while the second, the so-called take rate, is about the percentage of the trade value that the owner of the marketplace-platform will take, as a counterpart of its facilitation, lead generation, and more.

DOI: 10.4324/9781003226192-24

223

S im o n e C icer o

Figure 19.1 The three pricing contexts. Sizing the take rate

Take rates tend to vary quite a lot in the landscape of marketplaces: from as little as 5% where the owner mainly takes care of the intermediation, up to 80% that stock photo marketplaces take—a rather specific case, as most of the other merchandise categories see an upper take rate of around 30%. There is one key aspect to consider when setting a take rate: the actual contribution of the marketplace owners, and how that is perceived by the entities in the transactions. Several things need to be executed to ensure a good outcome of a transaction on the marketplace, and the platform owner normally takes over on some of them in cooperation with the producers. There are several approaches to frame the ‘stack’ of elements that make the transaction possible, and I came out with this stack: ■ ■

■

■ ■ ■

224

Customer acquisition and attraction refers to the processes needed to attract customers. Discovery/matchmaking refers to facilitating the identification of the niche product being sold (by the independent producer) and the connection with the ‘other side of the apple’ (the perfect buyer looking for such a niche experience). Trust building and risk reduction refer to mechanisms for facilitating the transaction achieved by reducing risk perception in the parties and increasing trustworthiness, such as insurance. Customer services and refunds refer to managing potential issues post-transaction due to bad quality of the experience. Ancillary services to production refers to all services (e.g., logistics) that support producers to execute the niche value proposition they’re delivering. Production refers to the actual execution of the core of the niche value proposition (VP) (e.g., renting the room, delivering the freelancing gig outputs). C hapter

1 9

P rici n g i n P latf o rms a n d M arketplaces ■

Niche VP innovation refers to the processes needed to innovate the value proposition that independent providers are providing to customers.

Clearly, the more of this stack of enabling elements that is taken over by the platform owners instead of the providers, the more the owner of the platform is entitled to take a bigger rate of the transaction’s value. Another key aspect of sizing a good take rate comes from understanding the so-called BATNA (best alternative to a negotiated agreement). Both sides of the transaction normally consider alternatives to transacting on the platform. From the perspective of providers, the alternative could be ■ ■

self-managing demand creation and transactions management using an alternative marketplace (potentially multi-tenanting, using both at the same time when possible)

If the market where you’re playing (or planning to play) is highly competitive, there may be an ongoing dynamic pushing take rates toward converging into a common value. If the market is also subject to a winner-take-all form of competition, the pricing wars that may derive from it—either because of an initial land-grab or a long-term consequence of the focus on retention—will likely end up driving take rates to the bottom. If the market is not crowded yet, the BATNA that the users will consider will need to be off-platform, making it slightly easier to sustain a higher take rate. Another aspect defining the take rate in a marketplace will regard choosing the breakdown of who to take it from. As an example, Airbnb has an articulated fee structure on the take rate: for nonprofessional hosts, the take rate is around 3% on the host’s side (or a little more, up to 6% in some cases) and circa 14% on the guest’s side, while for professional hosts (like hotels and some other hosts), the host doesn’t get any visible ‘service fee’ and the host is charged around 14% to 16%. The articulated way that Airbnb uses to define its take rates is a good example of how strategic one can be about it. Later in this chapter I focus on strategically using pricing in platforms, but one thing we can anticipate is that for marketplaces, you want to normally lower the friction (and a high take rate is an expression of it) toward the side of the marketplace you’re seeking the most to attract. As legendary investor Bill Gurley (2013) explains well in his seminal essay ‘A Rake Too Far,’ you normally want to create the least amount of friction in a platform to attract users: from the strategic perspective, lowering the friction toward the side you are constrained with may be an extremely good idea, especially as you seek liquidity. Clearly, take rates impact consumer pricing in your marketplace; therefore, a couple of essential elements need to be added to the picture. Another key element making up the final price for the consumer is—of course—the price providers put on their products or services. Depending on the nature of the supply in the market, marketplaces may or may not be able to apply certain pricing strategies such as imposing the cost of services or bundling transactions in one-off subscriptions. There could be essentially two macro cases: ■ ■

the marketplace sets the price of the goods to be sold (and thus needs to attract sellers to sell at that price) the marketplace lets providers set their price independently

C hapter

1 9 

225

S im o n e C icer o

The relationship between the marketplace and the suppliers clearly impacts this choice, just as the nature of services does. In the rare cases that the service sold is a frequently purchased commodity such as with rides, price is normally set by the platform; in most other cases, price is set under the laws of competition and specialization by the providers on the platform independently. In some cases when supply is not commoditized but can be substantially owned by the platform or when even demand can be fully owned (essentially meaning that multi-tenanting is not possible either because of exclusivity or other reasons), certainly the platform owner has dramatically more leverage in setting prices and could—eventually—even consider pricing bundles. Netflix, for example, provides its whole inventory at a fixed monthly subscription price. In a few words, as a quick recap: ■ ■ ■

When ubiquitous commodities are traded fixed price, price standardization works. When unique, niche services or products are traded, self-set pricing is the standard. When there’s a fully platform-owned inventory, this allows platform owners to play more with bundles, pricing strategies that go beyond monetizing a single transaction.

In any case, to properly juggle in the take-rate pricing spectrum, you’ll need to understand well the components of the provider’s cost of fulfilling the transaction and, most importantly, the value you can create for them by facilitating transactions and connecting them with consumers. This provider’s customers’ lifetime value—the LTV for a single customer that you connect to a provider for the first time—varies quite a bit across platforms, and it may also constitute an incentive to move off-platform (when this LTV is particularly high). As we’ll see later, this challenge is one to tackle strategically. In most cases, and especially with commodities, price transparency is an important strategy to pursue to reduce the buyer’s friction and has been approached in different ways. Uber, for example, has adopted predetermined tariffs, effectively having to enforce a mechanism to let drivers bid for a ride (you can’t technically oblige an independent driver to drive for a certain ride at a certain price); on the other hand, Amazon has enforced the use of the same ‘units’ (e.g., kg/oz) to present goods on the marketplace so that the customer is not confused by products adopting different packaging. Advertising is another major way to complement take rates as a monetization strategy for a marketplace/transactions engine. Advertising and positioning can be monetized in two major ways: as a separate value proposition or as an extension of the take rate, raking in bigger take rates for products and services traded because of being positioned through ads. According to Ben Evans, advertising essentially becomes a form of ‘taxation’ that the ecosystem owner can put in place as the ecosystem matures. Amazon Marketplace is a good example of a very mature marketplace where monetization undergoes a set of complex elements (Table 19.1). Pricing the ‘product side’ of your platform strategy

Besides the take rate, related to the transactional nature of the platform, marketplace platforms also offer a substantial ‘product’ side of the value proposition. The product side of the platform-marketplace strategy is key to complementing the value proposition 226

C hapter

1 9

P rici n g i n P latf o rms a n d M arketplaces Table 19.1  Example

of Amazon pricing

Take rate (referral fee) Fixed costs (selling plan) Additional services

Advertising

Variable, often between 8% and 15% $0.99 per unit (individual), $39.99 (professional) FBA fulfillment fees Storage fees Refund administration … Cost per click

of the marketplace. As anticipated above, the ‘product’ side can be described as a mix of a ‘learning’ engine and a ‘workflow’ engine, and the value proposition normally includes the following: ■

■

Some sort of software-based technological solution hovering around a SaaS that supports the execution of the core tasks the provider needs to fulfill their value proposition. A complementary set of enabling services such as a mix of marketing, sales, commoditized services, supply chain management, logistics, and more.

Certainly, this side of the value proposition offers a multiplicity of pricing strategies. Pricing a SaaS is a mature matter of investigation, and the reader can refer to many pieces of work that go into deep details. Above all we suggest that the reader catches up with the work of Patrick Campbell. While SaaS pricing is a complex art, we’ll try to distill the basic aspects here. The first thing one needs to understand about pricing this side of the business is the idea of the so-called value metric. According to Campbell: ‘A value metric is essentially what you charge for. For example, per seat, per 1,000 visits, per CPA, per GB used, per transaction, etc.’ Still according to Campbell, ‘if you get everything else wrong in pricing, but you get your value metric right, you’ll do ok’; that’s how key the value metric is. Understanding what you are charging for, what your user is aiming to achieve, is essential because it reduces churn and allows you to ensure that you’re not charging small and large customers (in terms of the success they derive from your product) in the same way. Understanding your value metric is also crucial to designing and adapting your organizational processes around delivering it to customers. Once you understand your value metric, you can pursue several approaches to building a contextual pricing strategy. You can go for ■ ■ ■ ■ ■

a freemium approach tiered subscription-based pricing (also on top of a Freemium model) usage-based pricing (much more apt to capture and put to value your understanding of the value metric, also on top of a Freemium model) paid premium services that are only sold to premium providers price by users

One thing that is essential to understand is that the more your value proposition is in the ‘transaction engine’ (meaning that you offer network-effect-based demand C hapter

1 9 

2 27

S im o n e C icer o

aggregation), the more you’ll need to operate on low margins, falling back on competitor-based pricing and cost-based pricing. Low margins—in fact—will generate low prices and thus attraction that, in turn, will provide further defensibilities. As the pendulum of your company value proposition moves more into the learning/ workflow engine zone, your focus needs to switch to value-based pricing because your value proposition is no longer (or at least not entirely) sensitive to network effects, thus to growth or size. Substack, for example, doesn’t aggregate demand directly and thus has a value proposition for writers that is largely independent of scale. As you move toward this side of your product, it thus makes more sense to try maximizing your pricing by improving how you understand customer value. Because of this, you’ll need to understand the value metric much more specifically, together with a few additional elements, namely: ■ ■

■

the customer’s willingness to pay the customer’s demographics (your so-called buyer personas that in this case are essentially the same as saying the producers you support with your marketplace-platform) the valued features for each customer (Figure 19.2)

Once you understand the demographics, the valued features—and therefore the value metric—you can evaluate the price elasticity of your product offering by plotting price perception.

Another pricing context: Your extensions platform

The third major context of expressing a platform’s pricing strategy that remains to be analyzed is the extension platform side. Extension platforms strategies are used by platform-marketplace owners to extend the value proposition of the product side of their strategy. As in the case of the marketplace, pricing here has two major ‘sides’: the developers (or more generally the third parties creating extensions) and the users of the ‘product side’ (normally, as said, the providers in the marketplace). Monetizing the extension providers (developers) normally happens through two major approaches: an

Value Proposition Focused on P2P Transactions (Marketplace)

Value Proposition Focused on the “product side” (for example SaaS)

It’s key to understand: - The supply side BATNA - Cost of fulfilling a transaction for the suppliers - LTV of the suppliers’ customers

It’s key to understand: - Willingness-to-pay - Value metric - Buyer personas

Figure 19.2  The pendulum.

228

C hapter

1 9

P rici n g i n P latf o rms a n d M arketplaces

integration fee, either one-time or recurring, needed to create or publish the extension, and a take rate on the purchase of the extension (if paid) or of any other digital element purchased (such as credit, or additional items). An example of the former is Apple’s $99 yearly individual developer account fee (yearly, mainly related to distribution and security checks), while the latter is well represented by the (fairly complex) Salesforce AppExchange Marginal Royalty Model where, depending on the annual order value (AOV), one can end up with independent software vendor (ISV) Force partners (those distributing the apps via AppExchange marketplace) paying a take rate ranging from 10% to 15%, with discounting growing as the business grows, as a way to encourage more active partners. Note that Salesforce AppExchange also has a one-time integration fee ($2,550 for the security review) and a yearly listing fee of $150. In more complex ecosystems, transactions can also happen inside the extension (often as micropayments), and brands have been often battling around enforcing certain payment channels to their extensions or app builders. The obvious reference here is Apple’s battle with Epic Games that led the producers of Fortnite to be deplatformed as an answer to their refusal to use Apple Pay as the only payment option available (and thus pay a 30% take-rate fee to Apple for each micropayment; ‘Epic Games v. Apple,’ 2022). For as much this complex picture of pricing in platform-marketplaces may seem complicated to handle, we still see some additional cases that do not even fit in this scaffolding. One interesting example is that of Airbnb’s co-host program. According to Airbnb’s website: Co-hosts help listing owners take care of their home and guests. A co-host is someone the listing owner already knows. They are usually a family member, neighbour, trusted friend, or someone the host has hired to help with the listing. Is a co-host to be seen as an ‘extension’ provider? In some ways it could; in some other ways, it may be seen as a player in the marketplace (or as a proxy for one) The interesting thing is that price setting here is left totally to the relationship between the host and its co-host and Airbnb doesn’t impose any take rate on it. Pricing and unit economics

As you focus on the key customer—normally the supplier as the recipient of your product value proposition, and possibly also the constrained side of your marketplace strategy—you will need to understand how to fully close the pricing circle by understanding how these numbers fit with unit economics across the spectrum. According to Profitwell, founders building products (or promoting the product side of a platform strategy) should investigate ‘quantified buyer personas’ (QBPs). The essential idea of a QBP is to capture the relationship between ■ ■

customer acquisition cost (CAC) customer lifetime value (LTV)

and their ratio, done for all personas and their relative customer acquisition channels. One particularly important observation that emerges is indeed that of product–channel C hapter

1 9 

229

S im o n e C icer o

fit. In his seminal post ‘Product Channel Fit Will Make or Break Your Growth Strategy,’ Brian Balfour (2017) explains two key things: ■ ■

You can mold the product but not the channel; therefore, understanding how the channel works for your customer is key to tuning the value proposition accordingly. Channels follow a Pareto distribution; therefore, identifying the key channel is key to the success of your strategy.

In a few words, when thinking about pricing and its impact on CAC and LTV, the platform designer will need to truly factor in the different channels connected with the different personas. How to use pricing strategically in platforms and marketplaces

After having explored how pricing can work, and the dimensions of the problem, we must investigate how pricing can be leveraged strategically for platform-marketplaces, in relation to the peculiar nature of such a business, and the phases that platformmarketplaces go through. The discussion on this strategic aspect of platform pricing is both old and new: Dan Hockenmaier has an excellent post on his blog where he wrangles with two key perspectives. In ‘The Marc Andreessen and Bill Gurley Schools of Pricing’ (2019), Hockenmaier presents two essentially different approaches to pricing to drive growth: Andreessen seems to point out that a higher price can serve as a way to feed the sales and marketing growth loop, while Gurley points out something that we already mentioned in the post, the idea of reducing friction as a way to drive growth and therefore network effect. In ‘Pricing for Platform Powered Businesses’ (Coutts, n.d.) and the related posts from the series, the team at LaunchWorks offers a good way to look into using pricing strategically in the platform-marketplace setting that is worth integrating and building on top. Based on such a series of posts and other reflections, we can try to list key strategic drivers that a platform-marketplace wants to achieve and evaluate the related pricing tactics (Table 19.2).

230

C hapter

1 9

C hapter

1 9 

Increase the AOV (size of transactions)

Grow the number of transactions per user

Increase the number of transactions

Attracting high spenders/ growing single transaction spending

Facilitate completion of transactions

Grow top of the funnel Reduce onboarding bouncing

Reduce onboarding friction

Growth

Through

Main strategic aim

pricing by strategic driver

Focus on

Table 19.2  Platform

adopt freemium pricing ■ no upfront/one-time fee ■ adopt usage-based pricing NA ■

■

■

■

■

■

■

■

■

adopt take rates that decline according to annual order value/number of transactions (e.g., Salesforce AppExchange) adopt usage-based subscriptions up to a certain number of transactions create additional paid services such as embedded finance (e.g., instant loans) to reduce transaction frictions (e.g., Shopify capital used to invest in marketing, branding—from Shopify experts) provide options with free transactions cancellation policy (higher price, e.g., Booking​ .c​om) adopt ‘price fencing’ (upper limits) for the customer reducing the perception of risk for variable pricing (e.g., Uber, Lyft) provide a discount of take rates on high-value transactions cap transaction fees (e.g., eBay) (Continued)

adopt lower than average (competitors) take rates

Pricing tactics (product Pricing tactics (transactions engine) side/learning engine and workflow engine)

P rici n g i n P latf o rms a n d M arketplaces

2 31

232

Improve LTV

Retention and diversification

Provide better transaction experiences

Main strategic aim

Focus on

Table 19.2 Continued

■

NA

Improve inventory depth

■

■

■

■

■

■

■

NA

Evaluate BUYER personas deeply and provide a pricing option for each persona NA ■

provide discounted take rates on high numbers of transactions (e.g., Salesforce AppExchange) apply the price to lead generation and not transaction fulfillment (e.g., Thumbtack) adopt fixed subscription types with unlimited transactions (e.g., Bumble—pro/boost—on top of a free basic profile) strategically chose what side to address the take rate to adopt dynamic pricing (‘surge pricing,’ e.g., Uber) take rate discounts for high-performing suppliers (e.g., Salesforce AppExchange) adopt category-based take rates (lower for certain categories you want to grow)—e.g., Amazon

Provide discounted take rates on high numbers of transactions (e.g., Salesforce AppExchange)

Pricing tactics (product Pricing tactics (transactions engine) side/learning engine and workflow engine)

Improve inventory quality NA

Improve P/C ratio (liquidity)

Keep transactions on the platform (or monetize transactions that happen off-platform)

Reduce churn

Through

S im o n e C icer o

C hapter

1 9

P rici n g i n P latf o rms a n d M arketplaces Conclusions

As the reader will have understood at this point, pricing is an essential element of a marketplace-platform strategy and has a tremendous impact on the platform’s growth potential. Overall, there are so many levers you can play, but the best way to understand if you are building a sustainable relationship (in terms of the balance between creating and extracting value) with your ecosystem may be the so-called Bill Gates line: ‘A platform is when the economic value of everybody that uses it, exceeds the value of the company that creates it. Then it’s a platform’ (Thompson, 2018).

Bio

Simone is an entrepreneur, facilitator, thinker, writer, and host, with a focus on open business models, ecosystemic organizations, and platforms: he has worked on these topics in edge contexts such as emergent, self-managed organizations and startups, but also with large institutions, ranging from Fortune 500s to the UN, to explore and experiment a new theory of organizing for the 21st century. In 2013, because of his experience with open and systemic business models, Simone created the first fully open-source methodology for platform and ecosystems thinking—the Platform Design Toolkit—an approach that effectively contributed to kickstart a new design domain, and now accounts for more than 70,000 practitioners all over the world. This and other works allowed him to be featured in the Thinkers50 Radar class of 2020.

Key objectives 1. Explain how pricing plays a role in reducing frictions in platforms and marketplaces. 2. Show the implications of platforms and marketplaces for all players in the monetization game. Deep dive into the unit economy of all methods used to price in platforms and marketplaces.

Key summary points 1. There are three critical components in a platform or marketplace: the marketplace itself, the product or services, and the extensions to these products and services. All three need to be monetized and priced. 2. The pricing tools and methods used for each of the three components must be applied using the best-in-class practices in SaaS but also in product business models. This summarizes the complexity of design and launch marketplaces and platforms. Excellence in pricing across multiple sectors is required in order to scale. 3. There are different pricing mechanisms to grow a platform or to retain existing members. It is essential to define the proper key performance indicators for each strategic objective.

C hapter

1 9 

233

S im o n e C icer o

Key questions 1. Why is pricing so important in digital platforms and marketplaces? 2. How can pricing be a driver of growth in digital platforms and marketplaces? 3. How do you use pricing strategically to promote user adoption or retain current users?

References Balfour, B. (2017, July 12). Product channel fit will make or break your growth strategy. https:// brianbalfour​.com ​/essays​/product​- channel​-fit​-for​-growth. Boundaryless. (2021, March 3). The new growth landscape webinar #1 – James Currier, general partner at NFX [Video]. Youtube. https://www​.youtube​.com ​/watch​?v​=BIRzZuWZVWU. Coutts, J. (n.d.). Pricing for platform powered businesses. Launchworks. https://www​ .launchworks​.co​/insights​/pricing​-platform​-powered​-businesses/. Epic Games v. Apple. (2022, July 26). In wikipedia. https://en​.wikipedia​.org ​/wiki​/ Epic​_Games​ _v.​_ Apple. Gurley, B. (2013, April 18). A rake too far: Optimal platform pricing strategy. Above the Crowd. https://abovethecrowd​.com ​/2013​/04​/18​/a​-rake​-too​-far​-optimal​-platformpricing​-strategy/. Hockenmaier, D. (2019, February 6). The Marc Andreessen and Bill Gurley schools of pricing. https://www​.danhock​.com ​/posts​/andreessen​-gurley​-pricing. Thompson, B. (2018, May 23). The Bill Gates line. Stratechery. https://stratechery​.com ​/2018​/ the​-bill​-gates​-line/.

234

C hapter

1 9

C hapt e r 2 0

Online Pricing Experimentation Robert Phillips

Introduction

It is much easier and cheaper for an online seller or marketplace to change prices than it is for their offline, brick-and-mortar counterparts. For an online seller, changing a price usually requires only typing the new price into a computer system—or having an algorithm change it automatically—while for a physical retailer changing even a single price usually requires altering price tags and displays as well as reprogramming the transaction register. In addition, the online seller potentially has continual access to streams of data regarding competitive prices, costs, inventories, sales, and other information that can be used to power automated pricing systems. This provides the opportunity for online sellers to adjust their prices quickly and cheaply. It also means that online sales are an ideal environment for pricing experimentation. It is hardly a secret that price changes are far more frequent online than offline. Offline prices have been quite stable in the low-inflation environment of the last decades: not counting temporary promotions, prices for consumer items change about once a year on average (Nakamura & Steinsson, 2008). Online prices change much more frequently—often many times a day. Figure 20.1 shows the end-of-day retail prices displayed by Amazon for the ‘Rio Red Swingline Stapler’ over a multiyear period. The price varied frequently in a range from $9.49 to $36.49 (although it was usually between $10.00 and $20.00). Price variation was even greater than Figure 20.1 implies since the figure ignores intraday changes, which can be quite frequent online. A marketplace in which prices routinely change frequently is an ideal environment for online price experimentation. In this chapter, we provide a broad guide to the issues involved in planning, executing, and evaluating online pricing experiments. We begin with a discussion of the different reasons for performing experiments and their implications for experimental design. We then discuss the steps of designing an experiment, executing the experiment, and evaluating the results. Finally, we discuss how pricing experimentation should be organized. Throughout, we focus on the high-level issues involved specifically in DOI: 10.4324/9781003226192-25

235

R o bert P hillips

Figure 20.1 Amazon retail price for the Rio Red Swingline Stapler in the US from December 2008, through February 2021. The price varied from a low of $9.49 to a high of $36.49. Source: camelcamelcamel​.co​m.

pricing experimentation. For those interested in going deeper, general concepts of statistical inference can be found in standard textbooks such as Levine, Stephan, and Szabat (2017). Shadish, Cook, and Campbell (2002) is a good source for general issues involved in experimentation.

Why experiment

In a pricing experiment, one population of customers is temporarily exposed to a pricing policy to be tested—the treatment—while the other population is priced as usual. The first population is called the test group or the treatment group, while the second population is called the control group. After the experiment, the effectiveness of the treatment can be estimated by comparing the performance of the treatment group with that of the control group in terms of metrics such as sales, revenue, and profitability. A pricing policy is any approach to setting and updating prices over time: ‘We will set a cap of $50 on all rides less than 20 miles in length’ or ‘We will always be priced at least 1% lower than Walmart​.c​om’ are examples of possible pricing policies. The purpose of a pricing experiment is typically to determine how customers will respond to the treatment relative to ‘pricing as usual.’ This can be very valuable information, and well-designed experiments can be much more informative than alternatives such as offline simulations or analysis of natural price variation. However, pricing experimentation can be costly and disruptive, and a seller might want to consider alternative sources of data before deciding to run an experiment. One possible source is natural price variation. As shown in Figure 20.1, online prices can vary considerably over time in response to changes in inventory, market conditions, competitive actions, and so forth. It is tempting to treat this natural price variation as an experiment and use it to estimate customer response. For example, for the stapler in Figure 20.1, Amazon could run a regression of daily demand against price and other factors to generate a price-response curve showing how demand for the stapler would have varied as a function of price. Such a price-response curve could then be used to estimate the sales, 236

C hapter

2 0

O n li n e P rici n g E xperime n tati o n 180 150

Demand

120 90 60 30 0 $0

$100

$200

$300

$400

$500

Average Fare Figure 20.2 Average fares and demands for a commercial airline flight departure over a year.

revenue, and profit that would be expected from different prices or different pricing policies. While using natural price variation to estimate price response is tempting, it can introduce bias because historic changes in prices are usually not random but based on factors such as market conditions and competitive actions that might also independently influence demand. To the extent that these factors can be observed and incorporated in the model as explanatory variables, this bias can be minimized. However, factors that influence both price and demand that are not recorded in the data may still lead to bias. For example, assume that a seller raises his price in anticipation of high demand. In this case, high prices would be associated with high sales—but this does not mean that high prices are driving high sales. Figure 20.2 shows the relationship between the seats sold and the average realized fare on a particular passenger airline flight for different departure dates. Naïve regression would lead to the conclusion that higher fares drove higher demands. In reality, the revenue management system for the airline was monitoring bookings and, when demand was likely to exceed capacity, raising the average fare. Not accounting for this effect would lead to severe underestimation of the magnitude of customer response to price—perhaps even with the wrong sign! Another potential source of price variation are natural experiments, in which an unanticipated event such as a regulatory change or a system error results in some customers being exposed to an alternative price. For example, Uber, like other ride-sharing companies, adjusts its ride pricing in response to the local demand/supply situation—if demand in a region sufficiently exceeds the supply of drivers, the price will be increased in order to decrease demand and attract more drivers (Phillips, 2020, chap. 7). Hall, Kendrick, and Nosko (2016) describe a ‘natural experiment’ in which a computer error caused Uber’s dynamic pricing adjustment system to fail in New York City during the high-demand period just after midnight on New Year’s Eve, 2014. As a result of the error, Uber prices were ‘stuck’ at their base levels when the algorithm would normally have raised them much higher to balance supply with demand. Comparison of results during the system failure with the results just prior and just after enabled Uber to estimate the effect of their dynamic pricing on metrics such as rider waiting time and ride completion percentage. C hapter

2 0 

2 37

R o bert P hillips

To summarize, while analyses of natural price variations and natural experiments can provide important insights, they are often unsuited to provide the guidance needed to support important business decisions. Natural price variation is usually not random and is therefore subject to biases. Natural experiments are unpredictable and, by their nature, cannot be targeted and can be very difficult to generalize. For this reason, randomized experiments are usually the preferred way to understand empirically the likely effects of a proposed pricing policy—in fact, randomized experiments are typically referred to as the gold standard for statistical inference.

Steps in a pricing experiment

There are three broad steps in performing a pricing experiment—planning the experiment, running the experiment, and evaluating the results. We discuss each of these in turn.

Planning the experiment

A critical step in designing an experiment is to specify—ideally in writing—the purpose of the experiment. Broadly speaking, experiments are performed for one of the three reasons shown in Table 20.1. ■

■

A truth-seeking experiment seeks to verify or disprove a hypothesis. An example might be ‘presenting a 10% reduction as a sale will generate more than 3% additional revenue relative to presenting it as the new list price.’ An exploratory experiment seeks to determine whether a treatment or combination of treatments has different impacts on different segments of the population (e.g., US vs. Canada), different products (e.g., men’s shoes vs. women’s shoes), or under different conditions (e.g., Sunday vs, Monday). An example might be to determine which customer segments respond most strongly to a 5% price reduction.

Table 20.1  Types

Type of experiment

Purpose

Evaluation approach

Truth-seeking

Verify or disprove a hypothesis

Exploratory

Measure relative or absolute differences in response across customer segments, product types, and/or market conditions Determine whether the treatment policy should be adopted in place of the status quo

Apply measures of statistical significance such as p-value to the result Apply a measure of statistical significance to the differences in response

Decision support

238

of experiments

Calculate Bayesian statistics such as probability the effect is greater than 0 and conditional gain or loss

C hapter

2 0

O n li n e P rici n g E xperime n tati o n ■

Decision-support experiments seek to determine whether a particular policy should be implemented in place of the status quo or not. This is the most common—and most important—motivation for pricing experiments in business.

Classically, most scientific experiments were ‘truth-seeking,’ in which a hypothesis is posed prior to the experiment and then either accepted or rejected based on the results of the experiment. As a result, much of classical statistics focus on hypothesis testing and the use of statistics such as the p-value to determine whether a hypothesis should be accepted as ‘true’ or not. ‘Exploratory experiments’ can often be framed as a list of hypotheses: for example, ‘customers in the Northwest will be significantly more responsive than the average,’ ‘customers in California will be significantly more responsive than the average,’ and so on. While exploratory experimentation can often be useful, it can result in statistically meaningless relationships being considered important. We discuss this issue in the section ‘Evaluating the Results.’ Finally, decision-support experiments are designed to help determine whether the treatment policy should be adopted in place of the status quo or not. This is quite a different question than whether a hypothesis should be accepted as true or not. As a result, the results of a decision-support experiment should be evaluated differently than a truth-seeking experiment. Rather than a ‘p-value’ measuring the probability that the difference in metrics is significant, measures of conditional gain and loss might be more useful; for example, ‘If the treatment policy is adopted, we estimate that there is a 63% probability that revenue will be increased and the expected increase would be $21 million per year. On the other hand, there is a 37% probability that revenue will decrease and the expected decrease would be $8 million per year.’ Choosing the metrics. A key step in designing an experiment is specifying what effects we want to measure—that is, what metrics need to be calculated and reported. Experimental metrics can be grouped into three categories. ■

■

■

Headline metrics are a relatively small set—typically four to eight—that are calculated for every experiment and highlighted in the results. Typically, headline metrics include quantities such as total revenue, total contribution (or profit), and units sold. Experiment-specific metrics measure quantities that are important for the experiment being run but are not among the headline metrics. For example, for a particular experiment, we may be interested in the influence of the treatment on six-week engagement—the number of times customers log in to the website in the six weeks following exposure to the treatment. In this case, the additional metric, ‘six-week engagement,’ would be calculated for this specific experiment and reported in addition to the headline metrics but would not necessarily be calculated for other experiments. Secondary metrics are metrics that are calculated for all experiments but are not headline metrics. An example might be ‘returns’—the effect of the treatment on the fraction of purchased goods returned. We might expect most treatments to have a negligible effect on returns, but we may track them anyway in case a particular treatment has a substantial negative effect.

C hapter

2 0 

2 39

R o bert P hillips

It is important that all headline metrics be calculated for every experiment and given a prominent place in all reports of the results. In fact, it is best practice to mandate that any communication of the results of an experiment contain all of the headline metrics—which is why there should be a small number of them. This is to prevent stakeholders from—consciously or unconsciously—highlighting favorable metrics and downplaying unfavorable metrics (or vice versa): for example, prominently displaying that the treatment increased revenue by 2.3% while burying in the small print the fact that it reduced profit by 22.2%. Experimental design. Once a seller has decided on the policy to be tested and the purpose of the experiment, she needs to choose the test group and control group. There are three broad options available—customer randomization, product randomization, and switchbacks. Customer-randomized experiments. In a customer-randomized experiment, when a customer arrives at the website or opens the app, a (metaphorical) coin is flipped to determine whether she sees the control price or the test price. Customer randomization has several advantages. First, customer-randomized experiments typically have more power than product-randomized experiments—in most cases, marketplaces have more customers than products. Second, it is usually easier to generalize the behavior of a population of customers than it is to generalize from a collection of products—it is much easier to conceive of a representative customer, or set of customers, than a ‘representative product’ or set of products. However, there are cases in which customer randomization may not be possible. By definition, customer randomization involves showing different prices to different customers for the same product at (more or less) the same time. This may be a problem. While it is not necessarily illegal, charging different prices for the same product at the same time can be problematic if it results in, say, one gender being disproportionately exposed to a higher price than the other. More importantly, most customers hate ‘personalized pricing’ in which different customers are shown different prices. On the internet, any attempt at personalized pricing—whether or not it stems from experimentation—is likely to be detected quickly and communicated widely, generating negative publicity and potential backlash as well as contaminating the results. For these reasons, many companies have prohibited ‘personalized pricing’ even for experimentation. For a discussion of the pitfalls of personalized pricing, see Phillips (2020), chapter 12. While customer-randomized experimentation may not be feasible for experimenting on the level of pricing, it is often ideal for experiments focused on the presentation of prices. For example, a question such as whether conversion rates will be higher if prices are displayed on the website in black versus red is ideally suited for customer randomization. While the color of the price display will be different, as long as the price levels shown to different customers are the same, this is unlikely to generate any controversy or ill will. Product-randomized experiments. As noted in the previous section, customer randomization is often infeasible when an experiment involves showing different prices to test and control customers. A common alternative to customer-randomized experimentation is product-randomized experimentation, in which products are divided into a test group and a control group. The pricing policy to be tested is applied to the test group for some period, while the control group is priced according to ‘business as

240

C hapter

2 0

O n li n e P rici n g E xperime n tati o n

usual.’ At the end of the test, we estimate the difference in key metrics between the two groups. While product-randomized experiments avoid the ‘personalized pricing’ issues that can arise from customer-randomized experiments, they have two important drawbacks. First, the results of a product-related experiment can be difficult to generalize to the entire population—strong generalizability requires that the products in the test group and the control group are both ‘representative’ of the entire population. Perhaps, more importantly, product-randomized experiments are particularly susceptible to concurrent cannibalization in which increased demand for a product in the test group might partially be cannibalized by demand for products in the control group (and vice versa). We discuss this issue more fully in the section ‘Evaluating the Results.’ Switchbacks. We have seen that both customer-based and product-based experiments present challenges in designing and interpreting online pricing experiments. Switchback experiments are an attractive alternative. In a switchback experiment, test and control policies are applied in alternate time periods—say, alternating days or weeks. The difference between metrics in the treatment periods and test periods can then be evaluated using variations on standard A/B evaluation approaches (Bojinov, Simchi-Levi, & Zho, 2021). Consider the case in which we want to determine the additional lift that would be gained from reducing the price of televisions by 5%. For a switchback experiment, the price of all televisions would be kept at their usual price for some period (the control period) and then reduced by 5% for some period (the test period). There could be multiple test and control periods—for example, alternating weeks of test and control for two months. Switchback experiments can eliminate concurrent cannibalization. Since all relevant products are subject to the test policy at the same time—for example, all television sets have a 5% price reduction—there is no concurrent cannibalization. However, switchback experiments are potentially subject to intertemporal cannibalization. For example, a low test price on a replenishable item such as toilet paper might lead consumers to purchase more than usual during the test period and stockpile the excess. To the extent that these additional purchases reduce demand during a future control period, it constitutes cannibalization. These stockpiling effects have long been recognized in the salespromotion literature—see Lu (2017) for example—and the potential for intertemporal cannibalization needs to be considered in evaluating the results of a switchback test, just as concurrent cannibalization needs to be incorporated in the evaluation of the results of a product-randomized experiment. Bojinov, Simchi-Levi, and Zhao (2021) discuss some approaches to avoiding bias in switchback experiments. Once a general approach has been selected (customer-randomized, product-randomized, or switchback), the duration of the test and the test and control groups need to be selected. Typically, this involves random assignment—selecting a subset of the population and randomly assigning half the subset to the test group and half to the control group. Determining the size and duration of an experiment requires estimating factors such as the power of the test and the ‘minimum detectable effect.’ Larger and longer duration experiments are statistically more powerful but potentially more disruptive. Determining the size and duration of an experiment to reach a desired level of certainty is beyond the scope of this chapter but can be found in standard texts such as Levine, Stephan, and Szabat (2017).

C hapter

2 0 

2 41

R o bert P hillips Running the experiment

In general, once an experiment has been designed, running it requires monitoring to ensure that the treatment has been correctly implemented in the code and that the treatment is not leading to clearly catastrophic results. However, in the absence of errors or clearly catastrophic results, it is important to ‘stay the course’ and run the experiment to its planned end. In addition, it is best practice not to release ‘intermediate results’ while the experiment is in flight—say, two weeks into an experiment that is scheduled to last for six weeks. There are two reasons for this. First, because of random effects, key metrics are likely to fluctuate wildly during the early weeks of an experiment. Thus, a revenue gain (or loss) from treatment of 8% after week 2 may shrink to 0% by the end of a six-week experiment. It is even possible that the results after two weeks may be a revenue gain from treatment of 8% with 95% significance. In this case, it is likely that there will be pressure to terminate the experiment and institute the treatment policy immediately. The pressure might even be greater to terminate the experiment if the results after two weeks are a revenue loss of 8% with 95% significance. However, in both cases, it is possible that the results are transient: the experiment may well traverse several regions of ‘statistical significance’ before settling on the ultimate value. A study of more than 2,000 commercial experiments found that 73% of the experiments had been terminated when showing a positive effect at the 90% confidence level, resulting in significant bias (Berman et al., 2018). More discussion of the issues involved can be found in Johari et al. (2017).

Evaluating the results

Once an experiment is complete, the results must be analyzed to estimate the expected effect of the treatment on the key metrics relative to the control group. While there are different ways to do this, the most common in a business environment is difference-indifferences (DiD). To understand DiD, assume that we are interested in understanding how participation in club sports influences weight gain among university freshmen. Table 20.2 shows the average weight in pounds of college club sports participants and nonparticipants at the beginning and end of their freshman years. A naïve approach would be to note that club sports participants gained an average of five pounds during their freshman year and conclude that participating in club sports led on average to a five-pound increase in weight. An equally naïve approach would be to compare the end-of-year weight of participants (147) with nonparticipants (164) and conclude that

Table 20.2  Beginning

and end-of-year weights for college freshmen who are club sport participants and nonparticipants Weight (lbs.)

242

Group

Beginning of year

End of year

Difference (lbs.)

Club sports participant Club sports nonparticipant Difference-in-differences

142 153

147 164

5 11 −6

C hapter

2 0

O n li n e P rici n g E xperime n tati o n

participation in club sports leads to an average 17-pound weight reduction relative to nonparticipation. DiD assumes that if the participants had not participated in club sports, they would have gained the same amount of weight as nonparticipants—that is, their weight would have increased by 11 pounds. The estimated average effect of participating in club sports is thus the difference between the weight gain percentages for club sport participants and nonparticipants, or −6 pounds. While DiD is a standard approach for evaluating the results of experiments, it does not necessarily establish causality. In the club sports example, it could be that the students who chose to participate in club sports had better eating habits or other behaviors that could be responsible in part or entirely for the differential weight gain. This is an example of selection bias and illustrates why randomized assignment to treatment and control groups is so important. For online pricing experiments, there is the additional risk of cannibalization between the test and control groups. This is a particular danger for product-randomized experiments. For example, assume that a retailer wants to estimate the effects of reducing the prices of television sets by 5%. Under product-based randomization, the retailer would choose test and control groups of television sets and would lower prices for the test group by 5% while holding the prices of the control group at their base level. Consider a pricing experiment run on two television sets: a JVC television in the test group and a VIZIO television in the control group We wish to test the effect of a 5% price drop on the JVC television, whose price was $659 before the test. During the test period, we drop its price by 5% to $626 and then raise the price back to its original level of $659 at the end of the test. The control (VIZIO) television was originally priced at $649, and we hold its price at that level during and after the end of the test. The results of the test in terms of unit sales and revenues are shown in Table 20.3. Using difference-in-differences to evaluate the results would conclude that the treatment was resoundingly successful, increasing sales by 30.1% and increasing revenue by 23.7%. However, a little thought would suggest that these results might be misleading. The test television set (JVC) was priced higher than the control (VIZIO) prior to the test but was cheaper during the test. It is thus highly likely that some of the sales for the test set were cannibalized from the control. Assume that 50% of the sales increase—or 118 units—for the test was cannibalized from the control. Then, the actual revenue gain from the price reduction was 118´ $626 - 118´ ( $649 - $626 ) = $71,154 . This means that the ‘real’ (net) increase in sales was 118 units and the increase in revenue was Table 20.3

Results from a hypothetical price test

Test (JVC) Control (VIZIO) DiD

Pre-test period

Test period

Sales

Revenue

Sales

Revenue

Sales

Revenue

810 1,020

$533,790 $661,980

1,046 1,010

$654,170 $655,490

236 −10

$120,380 −$6,490

246

$126,870

C hapter

2 0 

Change

2 43

R o bert P hillips

$71,154. This indicates that the treatment was much less effective than unadjusted DiD would suggest. Cannibalization is an example of the broader phenomenon of interference, which occurs whenever the application of the treatment influences results from the control group—in the case above, sales in the control group are influenced by the discount offered to the treatment group. A more subtle example of interference might occur in a ride-sharing platform that divides a city into two regions and applies a test pricing approach in the north region while maintaining ‘pricing as usual’ in the south region. In this case, the movement of drivers through the city could be influenced by the test pricing in the north region, which would change the level of service and prices in the south region. Since the treatment ‘interfered’ with the control, unmodified DiD may not provide an unbiased estimate of the magnitude of the treatment effect. Designing experiments to minimize interference and adjusting DiD estimators to account for possible interference is an active area of research. One approach is clustering. In a clustered experiment, items that are likely to cannibalize each other are grouped together in a cluster—for example, all men’s jeans may be assigned to a single cluster. Then, clusters are assigned to test and control groups for the purpose of the experiment. The idea is that, by assigning each cluster entirely to either the test group or the control group, cannibalization between the two groups can be minimized. While clustering can reduce interference, it also reduces the statistical power of a test. It also makes it difficult to evaluate policies that would apply entirely or largely to items in the same cluster—for example, how to design an experiment to evaluate a price reduction in men’s jeans if all men’s jeans are in the same cluster. For this reason, designing and evaluating experiments in the presence of cannibalization is an active area of research—see Hayes and Moulton (2017) and Karrer et al. (2021). Finally, once an experiment has been completed, it is natural to dig into the results in search of deeper insights—for example, were there customer segments, products, or geographies where the treatment was particularly effective or ineffective? While postexperimental analysis can provide such insights, too much ‘slicing and dicing’ of the experimental data can lead to spurious conclusions. Consider an experiment that was run nationwide in the US and had no statistically significant effect at the national level. Yet, even if the treatment was completely ineffective relative to the control group as a whole, there is a high probability that at least one of the 50 states will show a statistically significant effect purely by chance. For example, ‘The proposed pricing strategy was not effective except in Massachusetts, where it led to a gain in revenue that was significant at the 0.05 level.’ Someone is then likely to suggest that ‘we need to dig deeper into the results to understand why the strategy worked in Massachusetts but nowhere else,’ sending the analysts on a wild goose chase to understand the apparently idiosyncratic response of Bay Staters to the proposed policy. Cannibalization and ‘slicing and dicing’ are two of the important sources of potential misinterpretation of experimental results—however, they are hardly the only ones. They do illustrate that correct interpretation of the results of pricing experiments can raise subtle issues of statistical inference. Furthermore, effective pricing experimentation requires discipline in designing, running, and interpreting the results of experiments, which has implications for how pricing experimentation should be organized and managed.

244

C hapter

2 0

O n li n e P rici n g E xperime n tati o n How should pricing experimentation be organized?

To ensure rigorous interpretation of experimental results and that pricing experiments are efficiently and effectively designed, best practice is to establish an independent group that is responsible for all aspects of pricing experimentation including vetting and prioritizing experiments, validating the experimental design, running the experiments, and analyzing and interpreting the results. For such a group to be successful, three aspects are key. 1. To avoid biases, the experimentation group should be functionally separate from the business stakeholders. Those who have a stake in the outcome of an experiment should not be the same people who are evaluating the outcomes. 2. The experimentation group should be ‘science driven.’ This means that it should include statisticians and econometricians who can deal with the—sometimes subtle—issues involved in experimental design and interpretation such as cannibalization. 3. The capability to perform pricing experiments is typically a scarce resource—the demand for experiments often outstrips the capacity to run them, since it may be difficult to run many different experiments at the same time without having them interfere with each other. For these reasons, the experimentation group should have the ability to say ‘no’ to experiments. The experimentation group should also play a consultative role. While it does not necessarily need to design every experiment, it should be able to help business stakeholders design pricing experiments in a way that can deliver insights most efficiently and effectively. Furthermore, the group should also be able to help decision-makers and senior managers understand the results of experiments and how they should be used in making policy decisions. One of the most important roles of a pricing experimentation group is to avoid success bias—interpreting (or presenting) the results of an experiment in a way that is overly favorable to the treatment. Success bias can arise when the person who has suggested the treatment policy has a stake in its success. If that person has a role in interpreting the results, they may—consciously or unconsciously—find ways of analyzing the results that are tilted in favor of the treatment. These can include hand-picking ‘outliers’ to eliminate, choosing the metrics to emphasize that favor the treatment and de-emphasize those that favor the control, and slicing the data to highlight segments where the treatment beat the control. A pricing experimentation group that is independent of the business stakeholders substantially reduces the risk of success bias. Finally, the pricing experimentation group should be able to help management put the results of a pricing experiment in context when making a decision. Typically, the results of an experiment are only one among a number of different factors that will inform the decision whether to adopt a treatment or not. For example, a short experiment cannot measure the effects of a pricing policy on the long-term behavior of customers. The pricing experimentation group should be able to help management understand the strength and reliability of experimental results in a way that ensures that they are used appropriately in decision-making.

C hapter

2 0 

2 45

R o bert P hillips Summary

The ability to run and evaluate experiments to determine how different pricing policies will affect the business before implementing a policy is a key advantage that online retailers and marketplaces enjoy relative to their brick-and-mortar counterparts. However, realizing this advantage requires a structured method for designing, running, and evaluating pricing experiments. It requires statistical sophistication to accurately evaluate results in the face of cannibalization. Furthermore, to avoid biases, the group designing, running, and interpreting experiments should be separate from those who have a stake in their outcomes. For these reasons, companies that plan on running multiple pricing experiments should develop an internal science-led group to manage the design and execution of the experiments as well as the validation of the experimental results.

Acknowledgments

I am grateful to Joe Cooprider and Tara Mardan of Amazon for their careful reading and comments that greatly improved the chapter.

Bio

Robert Phillips is former Director of Pricing Research at Amazon and Director of Marketplace Data Science at Uber. Prior to that, he was a Professor at Columbia University and Founder of Nomis Solutions. He is the author of Pricing and Revenue Optimization and Pricing Consumer Credit and a co-editor of the Oxford Handbook of Pricing Management. He can be contacted at robert​.phillips​@nomissolutions​.​com.

Key objectives 1. Explain when online pricing can be used, the reasons it is used, and how it can provide an advantage over brick-and-mortar competitors. 2. Describe different approaches for selecting test and control groups and their strengths and weaknesses. 3. Describe the three steps of designing, running, and evaluating the results of an experiment and some of the challenges associated with each step. 4. Advocate for an independent pricing experiment organization whenever the volume of online experiments is sufficient to justify it.

Key summary points 1. The ability to run and evaluate experiments to determine how different pricing policies will affect the business before implementing a policy is a key advantage that online retailers and marketplaces enjoy relative to their brick-and-mortar counterparts.

246

C hapter

2 0

O n li n e P rici n g E xperime n tati o n

2. Realizing this advantage requires a structured method for designing, running, and evaluating pricing experiments. It also requires statistical sophistication to accurately evaluate results in the face of cannibalization. 3. To avoid biases, the group designing, running, and interpreting experiments should be separate from those who have a stake in the outcome of the experiment. For this reason, companies that plan on running multiple pricing experiments should develop an internal science-led group to manage the design and execution of the experiments as well as the validation of the experimental results.

Key questions 1. 2. 3. 4.

Why should online marketplaces run pricing experiments? What are the different types of experiments that can be run? What are the steps in a pricing experiment? What are the challenges in each step? Who in an organization should be in charge of designing, running, and evaluating pricing experiments?

References Berman, R., Pekelis, L., Scott, A., & Van den Bulte, C. (2018). p-Hacking in A/B testing. https:// ssrn​.com​/abstract​=3204791. Bojinov, I., Simchi-Levi, D., & Zhao, J. (2021). Design and analysis of switchback experiments. https://papers​.ssrn​.com ​/sol3​/papers​.cfm​?abstract​_ id​=3684168. Hall, J., Kendrick, C., & Nosko, C. (2016). The effects of Uber’s surge pricing: A case study [Working Paper, Uber, San Francisco]. https://leeds​-faculty​.colorado​.edu​/ leachj​/ BCOR1015​ /Readings​%20not​%20linked​%20to​%20Library​%20Page​/ Effects ​_ of​_uber’s ​_ surge ​_ pricing​ %20CASE​​.pdf. Hayes, R. J., & Moulton, L. H. (2017). Cluster randomized trials. Boca Raton, FL: CRC Press. Johari, R., Koomen, P., Pekelis, L., & Walsh, D. (2017). Peeking at A/B tests: Why it matters and what to do about it. In KDD ’17: Proceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining (pp. 1517–1525). https://doi​.org​/10​ .1145​/3097983​. 3097992. Karrer, B., Shi, L., Bhole, M., Goldman, M., Palmer, T., Gelman, C., Konutgan, M., & Sun, F. (2021). Network experimentation at scale. In Proceedings of the 27th ACM SIGKDD conference on knowledge discovery and data mining (pp. 3106–3115). Singapore. Levine, D. M., Stephan, D. F., & Szabat, K. A. (2017). Statistics for managers using Microsoft ® Excel (8th ed.). Boston, MA: Pearson. Lu, A. (2017). Consumer stockpiling and sales promotions (DIW Berlin Discussion Paper No. 1680). https://ssrn​.com ​/abstract​=3034151. Nakamura, E., & Steinsson, J. (2008). Five facts about prices: A reevaluation of menu cost models. Quarterly Journal of Economics, 123(4), 1415–1464. Phillips, R. L. (2020). Pricing and revenue optimization (2nd ed.). Stanford, CA: Stanford University Press. Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Belmont, CA: Wadsworth Cengage Learning. C hapter

2 0 

2 47

SECTION 5

Pricing and Artificial Intelligence



C hapt e r 2 1

Artificial Intelligence and Its Impact on Pricing Technology Louis Columbus

Introduction

We are going through an AI pricing revolution. There is no denying it. The last five years have been nothing short of a major leap forward in science in the pricing space. Companies are beginning to notice the impact of these developments and are pouring major investments into new technologies. Look at these impressive statistics! ■ ■

■

AI-based pricing and promotion have the potential to deliver $259.1 billion to $500 billion in global market value, according to McKinsey Global Institute (2018). The global revenue management market is expected to grow from $14.5 billion in 2019 to $22.4 billion by 2024, attaining a compound annual growth rate (CAGR) of 9.6% (Global Newswire, 2020). BCG found that automating revenue management systems’ pricing rules with AI can increase revenues by up to 5% in less than nine months (Gartner, 2020).

Eighty-five percent of B2B management teams believe their pricing decisions need improvement, and just 15% have effective tools and dashboards to set and monitor prices, according to a recent Bain (2020) global survey of more than 1,700 business leaders. For the many companies that rely on pricing as a competitive advantage, they need to begin evaluating AI and machine learning on their IT platform roadmaps now. Staying at competitive parity and turning AI- and machine learning-based expertise into a pricing and revenue management strength needs to be a priority. Data is a proven panacea for fear, and given the new market dynamics many companies are facing, it is the most reliable way to make decisions. Fourteen ways AI impacts and improves pricing

The following are the ways AI is improving pricing and revenue management today. Using AI to identify and then eliminate the most unproductive customer discounts and segments, freeing up more financial resources and time for those that contribute DOI: 10.4324/9781003226192-27

2 51

L o uis Co lumbus

to profits. A recent Bain & Company research brief, ‘Bringing Order to Discounts Gone Haywire’ (Davenport et al., 2020) provides an excellent example of how AI can be used to determine the effectiveness of discounts by customer segment and discount type. The brief mentions how focused analysis of discounts can help stop revenue leakage due to suboptimal, expensive customer investments. Automating pricing rules with AI in revenue management systems increases total revenue by 5%. Boston Consulting Group (BCG) found that 95% of successful digital transformation initiatives used one or more revenue growth levers (Chugani et al., 2020). Seventy-seven percent of a given digital transformation’s financial impact was achieved through the combined use of six revenue growth levers. Improving pricing optimization with advanced techniques including AI has the potential to deliver a 5% increase in total revenue. BCG believes that automating pricing rules in revenue management systems and enforcing contractual pricing changes increase revenue. Capitalizing on the many insights that transactional data can provide by using AI and machine learning to look for patterns in pricing, volume, and mix analysis is delivering measurable results today. The patterns and trending insights in transaction data include new insights every business can use to become more competitive. Unlocking those insights takes an AI-based approach to interpreting the price, volume, and mix fluctuations often locked within the constraints of transactional data. Combining transactional data analysis and price, volume, and mix fluctuations have proved difficult and a challenge to combine in a unified, intuitive application. AI and machine learning are helping pricing managers capture more revenue and profits by finding what a given customer is willing to pay or optimizing prices across their customer and product mix. Identifying blind spots in pricing, discount, and deal size decisions is difficult for customers and products using spreadsheets alone. AI and machine learning help pricing managers analyze whether existing discounts make sense by correlating deal size to discounts made and identifying outliers where discounts have been granted due to the negotiating insight of the customer (Bain & Company, 2020). AI is making it possible to create propensity models by persona, and they are invaluable for predicting which customers will act on a bundling or pricing offer. By definition, propensity models (Barga et al., 2015) rely on predictive analytics, including machine learning, to predict the probability that a given customer will act on a bundling or pricing offer, e-mail campaign, or other call-to-action leading to a purchase, upsell, or cross-sell. Propensity models have proved to be amazingly effective at increasing customer retention and reducing churn. Every business excelling at omnichannel today relies on propensity models to better predict how customers’ preferences and past behavior will lead to future purchases. Price optimization and price elasticity are growing beyond industries with limited inventories, including airlines and hotels, proliferating into manufacturing and

252

C hapter

2 1

A I a n d I ts I mpact o n P rici n g T ech n o l o gy

services. All marketers are increasingly relying on machine learning to define more competitive, contextually relevant pricing. Machine learning apps are scaling price optimization beyond airlines, hotels, and events to encompass product and service pricing scenarios. Machine learning is being used today to determine pricing elasticity by each product, factoring in channel segment, customer segment, sales period, and the product’s position in an overall product line pricing strategy (Azure, 2017). Fine-tuning price segmentation strategies with insights gained from AI is helping to stabilize and increase margins and revenues today. Each customer segment has a different price they are willing to pay for a given product or service. By using AI and machine learning to know the price by segment that customers are most willing to pay for a given product, AI applications can suggest them to sales and revenue managers. Automating segment-specific pricing guidance using CRM and CPQ systems is pivotal to pricing segmentation strategies’ success. AI is providing sales and revenue managers with more accurate deal price guidance than was available in the past, leading to more effective use of pricing discounts. Facing greater pricing pressure in sales cycles, they want to close quickly, and sales reps are quick to provide deep discounts that sacrifice margin. This is especially true in enterprise software. McKinsey found that using dynamic deal scoring indexed to discounts provides the guidance that sales reps need in determining what level of discounting will win the deal and not sacrifice margin (Baker et al., 2018). Relying on AI to monitor risk-based metrics and KPIs to gain greater visibility into the root cause of potential risks to revenue. Lost sales, accounts, and customers often happen because sales and service teams do not know soon enough that there is an issue. AI-based alerting on key revenue, pricing, and quoting metrics can save a sale or a customer and help pinpoint a specific product issue as well. AI-based risk alerts are customizable for specific metrics and conditions and are sent to relevant team members, supporting customers. The most valuable aspect of these alerts is getting to the root cause of any issue. Eliminating pricing errors on orders that lose sales while providing a more precise approach to automating special pricing requests (SPRs) drives more sales. Price lists proliferate across the many selling channels any given manufacturer relies on, often leading to incorrect quotes and orders. When customers find out the pricing is wrong, they cancel orders and go to competitors. This happens because consumer-packaged goods (CPQ) systems are not integrated with pricing, ERP, and back-end systems. Combining historical selling, pricing, and buying data in a single machine learning model improves the accuracy and scale of sales forecasts. Factoring in differences inherent in every account given their previous history and product and service purchasing cycles is invaluable in accurately predicting their future buying levels. AI and machine learning algorithms integrated into CRM, sales management, and

C hapter

2 1 

2 53

L o uis Co lumbus

sales planning applications can explain variations in forecasts, provided they have the data available. Forecasting demand for new products and services is an area where AI and machine learning are reducing the risk of investing in entirely new selling strategies for new products. Knowing the propensity of a given customer to churn versus renew is invaluable in improving customer lifetime value. Analyzing a diverse series of factors to see which customers are going to churn or leave versus those that will renew is among the most valuable insights that AI and machine learning are delivering today. Being able to complete a customer lifetime value analysis for every customer that a company has provides a prioritized roadmap of the health of client relationships: where these are excellent and where these need attention. Many companies are using customer lifetime value analysis as a proxy for a customer health score that gets reviewed monthly. Improving demand forecasting, assortment efficiency, and pricing in retail marketing has the potential to deliver a 2% improvement in earnings before interest and taxes (EBIT), a 20% stock reduction, and 2 million fewer product returns a year. In CPQ and retail marketing organizations, there’s significant potential for AI and machine learning to improve the entire value chain’s performance. McKinsey found that using a concerted approach to applying AI and machine learning across a retailer’s value chains has the potential to deliver a 50% improvement in assortment efficiency and a 30% online sales increase using dynamic pricing (McKinsey Global Institute, 2017). McKinsey finds that AI is improving demand forecasting by reducing forecasting errors by 50% and lost sales by 65% with better product availability. Supply chains are the lifeblood of any manufacturing business. McKinsey’s initial use-case analysis is finding that AI can reduce costs related to transport and warehousing and supply chain administration by 5% to 10% and 25% to 40%, respectively. With AI, overall inventory reductions of 20% to 50% are possible (McKinsey & Company, 2017).

Bio

Louis Columbus is a Software Product Marketing and Product Management Leader with experience in marketing management, channel management, and direct sales. Previous positions include product management at Ingram Cloud, product marketing at iBASEt, Plex Systems, Senior Analyst at AMR Research (now Gartner), and marketing and business development at Cincom Systems, Ingram Micro, a SaaS startup, and at hardware companies. He is also a member of the Enterprise Irregulars. His background includes marketing, product management, sales, and industry analyst roles in the enterprise software and IT industries. His academic background includes an MBA from Pepperdine University and completion of the strategic marketing management and digital marketing programs at the Stanford University Graduate School of Business. He teaches MBA courses in international business, global competitive

254

C hapter

2 1

A I a n d I ts I mpact o n P rici n g T ech n o l o gy

strategies, international market research, and capstone courses in strategic planning and market research. He has taught at California State University, Fullerton; University of California, Irvine; Marymount University, and Webster University. You can reach him on Twitter at @LouisColumbus.

Key objectives 1. Understand how AI science might revolutionize pricing technology. 2. Evaluate the impact of pricing technology and AI on business outcomes. 3. Realize the penetration of advanced scientific methods in pricing and sales automation.

Key summary points 1. The impact of AI in pricing branches out to other functions such as supply chains, sales, and marketing. 2. Companies that are ahead of the curve with the adoption of pricing technology and pricing science (such as AI) are already benefiting from the efficiencies and margin gains. They are poised to accelerate as more data become available and are connected to other parts of the business. 3. Leveraging data and pricing science allows companies to recover faster from market and economic disruption, as they can quickly pivot business models, test new models with data, and automate their new go-to-market strategies.

Key questions 1. Will personalized pricing through AI be the new normal in decades? 2. What might be some of the limitations to AI pricing that may slow down adoption and implementation? 3. How do pricing, data analytics, and sales analytics get integrated to provide the right customer-centric AI platform?

References Ai4. (2020, March 10). Machine learning for pricing and inventory optimization @ Macy’s. https://ai4​.io​/ blog ​/ 2020​/ 03​/10​/machine​-learning​-for​-pricing​- and​-inventory​- optimization​ -macys/.

C hapter

2 1 

2 55

L o uis Co lumbus Azure. (2017, July 28). Cortana interactive pricing analytics pre-configured solution: User guide. https://github​.com ​/Azure ​/cortana​-intelligence​-price​-analytics​/ blob​/master​/ User​%20 Guide​/ UserGuide​.md. Bages-Amat, A., Baker, W., Magnette, N., & Winkler, G. (2018, October). What matters in B2B dynamic pricing. McKinsey & Company. https://www​.mckinsey​.com​/ business​-functions​/ growth​-marketing​-and​-sales​/our​-insights​/what​-really​-matters​-in​-b2b​-dynamic​-pricing. Bain & Company. (2020, February 24). Harnessing pricing power to create lasting value. https:// www​.bain​.com ​/globalassets​/noindex ​/2020​/ bain ​_ report ​_private ​_ equity​_ report ​_ 2020​.pdf. Baker, W., Kiermaier, M., Roche, P., & Vyushina, V. (2018, June 14). Advanced analytics in software pricing: Enabling sales to price with confidence. McKinsey & Company. https:// www​. mckinsey​.com ​/ industries ​/technology​- media​- and​- telecommunications ​/our​-insights ​/ advanced​-analytics​-in​-software​-pricing​- enabling​-sales​-to​-price​-with​- confidence. Barga, R., Fontama, V., & Tok, W.-H. (2015). Predictive analytics with Microsoft Azure machine learning (2nd ed.). Apress. Baumgartner, T., Hatami, H., & Valdivieso, M. (2016, June 10). Why salespeople need to develop ‘machine intelligence.’ Harvard Business Review. https://hbr​.org ​/2016​/06​/why​ -salespeople​-need​-to​-develop​-machine​-intelligence. Burns, D., Schottland, D., Murphy, J., & Whiteley, T. (2020, April 3). Revving up sales ROI for a downturn. Bain & Company. https://www​.bain​.com ​/insights​/revving​-up​-sales​-roi​-for​ -a​-downturn/. Chugani, S., Ghosh, S., Greenlee, N., Fæste, L., & Pototschnik, L. (2020, August 18). How to grow revenue quickly and sustainably in transformations. Boston Consulting Group. https:// web ​- assets​.bcg​. com ​/ 87​/ 01 ​/e5e​b10e​8 4d9​d a69​9182​b c3f76727​/ bcg​- how​- to ​- grow​- revenue​ -quickly​-and​-sustainably​-in​-transformations​-aug​-2020​-rev​.pdf. Davenport, C., Burns, D., & Narayan, A. (2020, May 26). Bringing order to discounts gone haywire. Bain & Company. https://www​.bain​.com ​/insights​/ bringing​-order​-to​-discounts​ -gone​-haywire/. Fontes Gerards, F., Goodin, C., Logan, B., & Schmidt, J. (2018, December). Powerful pricing: The next frontier in apparel and fashion advanced analytics. McKinsey & Company. https:// www​. mckinsey​. com ​/ industries ​/retail ​/our ​- insights ​/powerful​- pricing ​- the ​- next​-frontier ​- in​ -apparel​-and​-fashion​-advanced​-analytics. Gartner. (2020, July 27). Hype cycle for artificial intelligence, 2020. https://web​-assets​.bcg​ .com​/ 87​/ 01​/e5e​b10e​8 4d9​d a69​9182​b c3f76727​/ bcg​- how​- to​- grow​- revenue​- quickly​- and​ -sustainably​-in​-transformations​-aug​-2020 ​-rev​.pdf (client access required). Global Newswire. (2020, March 9). The revenue management market size is projected to grow from USD 14.1 billion in 2019 to USD 22.4 billion by 2024, at a compound annual growth rate (CAGR) of 9.6%. https://www​.globenewswire​.com​/news​-release​/2020​/03​/09​/1996962​ /0​/en ​/ The​-revenue​-management​-market​-size​-is​-projected​-to​-grow​-from​- USD ​-14 ​-1​-billion​-in​ -2019​-to ​- USD ​-22​- 4 ​-billion​-by​-2024 ​-at​-a​- Compound​-Annual​- Growth​- Rate​- CAGR​- of​-9​- 6​ .html. Gregg, B., Hazan, E., Khanna, R., Kim, A., Perrey, J., & Spillecke, D. (2020, May 7). Rapid revenue recovery: A road map for post-COVID-19 growth. McKinsey & Company. https:// www​. mckinsey​. com ​/ business​-functions ​/growth​- marketing​- and​- sales ​/our​- insights ​/rapid​ -revenue​-recovery​-a​-road​-map​-for​-post​- covid​-19​-growth. Inside BIGDATA. (2020, July 30). Using AI to optimize pricing. https://insidebigdata​.com ​/2020​ /07​/30​/using​-ai​-to​-optimize​-pricing/. McKinsey & Company. (2017, April). Smartening up with artificial intelligence (AI)—What’s in it for Germany and its industrial sector? https://www​.mckinsey​.com/~​/media ​/mckinsey​ /industries​/semiconductors​/our​%20insights​/smartening​%20up​%20with​%20artificial​ %20intelligence​/smartening​-up​-with​-artificial​-intelligence​.ashx. McKinsey & Company. (2019, November). Global AI survey: AI proves its worth, but few scale impact. https://www​.mckinsey​.com/~​/media​/ McKinsey​/ Featured​%20Insights​/Artificial​ %20Intelligence​/ Global​%20AI​%20Survey​%20AI​%20proves​%20its​%20worth​%20but​

2 56

C hapter

2 1

A I a n d I ts I mpact o n P rici n g T ech n o l o gy %20few​%20scale​%20impact​/Global​-AI​- Survey​-AI​-proves​-its​-worth​-but​-few​- scale​-impact​ .pdf. McKinsey Global Institute. (2017, June). Artificial intelligence: The next digital frontier? https://public​.tableau​.com ​/app​/profile​/mckinsey​.analytics​/viz​/Artificial​_ Intelligence​/ Impact​ _of​_ AI ​_ and​_ Analytics. McKinsey Global Institute. (2018, April 17). Estimated impact of artificial intelligence and other analytics by industry, function and business problem. https://public​.tableau​.com​/app​/ profile​/mckinsey​.analytics​/viz​/Artificial​_ Intelligence​/ Impact​_of​_ AI ​_ and​_ Analytics. Omale, G. (2019, October 2). 4 key insights from the Gartner Hype Cycle for CRM sales technology. Smarter With Gartner. https://www​.gartner​.com ​/smarterwithgartner​/4 ​-key​ -insights​-gartner​-hype​- cycle​- crm​-sales​-technology​-2019. Rey-Herme, B., & Milani, A. (2019). 5  Ways to guide the deal with CPQ. Accenture. https:// www​.accenture​.com/​_ acnmedia​/ PDF​-107​/Accenture​-5​-Ways​-guide​-deal​-with​- CPQ​.pdf. Rizzi, W., Wang, Z. M., & Zielinski, K. (2018, September 20). How machine learning can improve pricing performance. McKinsey & Company. https://www​.mckinsey​.com ​/industries​ /financial​-services​/our​-insights​/ how​-machine​-learning​- can​-improve​-pricing​-performance. Shyam, P. (2020, May 29). AI can solve maintenance and quality challenges for manufacturers. Capgemini. https://www​.capgemini​.com ​/us​- en ​/ insights​/expert​-perspectives​/ai​- can​- solve​ -maintenance​- and​- quality​- challenges​-for​- manufacturers/. Sicular, S. (2020, July 29). The hype cycle for artificial intelligence 2020 reflects the state of AI in the enterprise. Smarter With Gartner. https://blogs​.gartner​.com ​/svetlana​-sicular​/the​-hype​ -cycle​-for​-artificial​-intelligence​-2020​-reflects​-the​-state​-of​-ai​-in​-the​- enterprise/. Trenka, J., & Demmelmair, M. F. (2019, September 2). Pricing intelligently for competitiveness and growth. Accenture. https://www​.accenture​.com ​/us​- en ​/insights​/strategy​/pricing​-growth. Velez, J., Magnette, N., & Bagri, N. (2019, April 1). Mastering the art of the 80 percent: How to drive sustainable B2B pricing excellence with data and analytics. McKinsey & Company. https://www​. mckinsey​.com ​/ business​-functions​/growth​- marketing​- and​- sales​ /solutions​/ periscope ​/our​- insights​/articles​/ mastering​- the​- art​- of​- the​- 80 ​- percent​- how​- to​ -drive​- sustainable​-b2b​-pricing​- excellence​-with​- data​- and​- analytics.

C hapter

2 1 

257

C hapt e r 2 2

Why AI Transformations Should Start with Pricing Joël Hazan, Camille Brégé, Jean-Sébastien Verwaerde, and Arnaud Bassoulet

The acceptance and financial upside of AI continue to lag far behind the hype, except in one surprising and often-overlooked area: pricing. The speed, sophistication, and scale of AI-based tools can boost EBITDA by 2 to 5 percentage points when B2B and B2C companies use them to improve aspects of pricing that have the greatest leverage within their organizations. A large-scale global survey, conducted by the Massachusetts Institute of Technology (MIT) and the BCG Henderson Institute (BHI), revealed how successful AI-driven pricing transformations can be and how infrequently they are pursued. In the technology sector, for example, only 12% of the companies surveyed used AI to improve their pricing, but their initiatives succeeded twice as often as the efforts of companies that applied AI to other functional areas (see ‘About the survey’).

Why pricing is an excellent target

Success in boosting the top and bottom lines isn’t the only reason that companies should focus their AI initiatives on pricing rather than on other functions or processes. Pricing already relies on processes and tools. An AI transformation can vastly improve a company’s existing data flows and processes, and those related to pricing are ideal starting points because they are usually well-established but often lack sophistication. A pricing maturity assessment, conducted by BCG and the Professional Pricing Society, revealed that more than 50% of all industrial goods companies still use Microsoft Excel to build their primary pricing tools and that 25% of B2B companies use static, one-size-fits-all pricing with limited inputs and infrequent updates. These companies are ripe for the kind of step change that AI can provide (Brégé et al., 2020). Large companies often have a patchwork of pricing processes, and AI can enable them to raise and then scale their level of sophistication. The MIT-BHI survey showed that large companies (those with more than $10 billion in annual revenues) that undertook AI-driven pricing transformations achieved more than $100 million of revenue DOI: 10.4324/9781003226192-28

2 59

J o ë l H a z a n et al .

improvement 70% more often than companies that focused on another area (see the exhibit). Their struggle rate was also much lower. Only 13% that pursued AI-driven pricing initiatives failed to see any benefit, compared with 34% of companies whose AI initiatives did not address pricing. Companies can win many pricing battles with AI. Across industries, companies can tackle pricing’s complexity using advanced analytical approaches that take advantage of AI. Consumer packaged goods (CPG) companies, for example, can use insights from AI-based analytics to rethink pricing for their overall brand portfolio, refine their packprice architectures (Nieto et al., 2020), and improve their mix management by doubling down on products and channels that have higher margins. CPG companies can also use AI-based tools to improve the efficiency of their promotions and tighten their trade terms. B2B companies can employ AI-based analytics to mine rich transaction data to find quick wins in terms of incremental price differentiation and improved discounting. They can also use AI-based tools to determine price metrics, set price levels, and manage price implementation. Companies should identify the battles with the clearest and fastest upside relative to the investment—and begin their AI pricing transformation with those. A pricing transformation is easier for employees to accept. Some members of the pricing team may fear that AI applications will eliminate their jobs. However, such concerns are often quickly allayed because AI initiatives equip pricing teams with more powerful, user-friendly tools that require their subject matter expertise and business judgment (Chu, 2020) to deliver optimal performance. This human-machine or bionic approach not only addresses team members’ fears but also shows how such initiatives can accelerate or improve what they do, even when these initiatives create new areas such as demand centers or a yield management function. The bionic approach also frees up their time to take on more value-added tasks and to address the more far-reaching challenges that often get crowded out by acute short-term priorities. As a result, AI pricing transformations tend to cause less organizational disruption, even though they involve marketing, sales, finance, and revenue growth management (RGM).

A CPG company’s AI pricing transformation

At the explicit urging of its CEO, a large CPG company assessed various areas of its organization to determine which AI initiatives offered the best opportunities for securing quick returns, advancing tools and processes, winning battles, and gaining team acceptance and confidence. The company determined that pricing offered the best opportunities, and then it narrowed the scope to promotions—their performance, timing, and range. But even that scope was too broad because of the differences across regional, national, and local markets. The company finally chose to focus on promotions in established channels in two large markets. The plan was to start small, build a minimum viable solution, prove the value, and commence with change management. Only then would the company scale the initiative to more markets and teams. This intense, very specific focus helped the company build concrete solutions that delivered value in two ways: first, by providing faster, fact-based answers on how to improve existing promotions; and second, by testing and implementing ideas that the company had either lacked faith in or never considered. The company estimates that 260

C hapter

2 2

W hy A I T ra n sf o rmati o n s S h o uld S tart with P rici n g

after scaling up the initiative, the full impact will yield 2 percentage points of EBIT improvement. From the beginning, the change management effort focused on demonstrating that AI pricing is a bionic approach and that the advanced analytics would still require the team’s skills and experience applied to different tasks. This was critical because some team members feared that AI would diminish their roles or make them redundant. Early successes, however, proved to be the biggest motivators for the team to sustain its progress. Three key steps to success

The success of AI pricing transformations generally hinges on the quality of three factors: data, vision, and change management support. We built on this insight to create three recommendations for companies that want to take advantage of the significant upside potential of AI-based pricing. Focus on data early and sustain the effort. Conducting an up-front data assessment will help the company select the right focus area within pricing as well as the right region or market in which to pilot a program. A niche market, for example, may have rich accessible data and a large potential return but limited opportunities to scale. An established market may have solid data and fewer potential upsides locally but offer opportunities to scale a solution quickly. Once the initial AI pricing solution is built and launched, the company’s focus should shift to gradually improving both the quality and availability of data in order to unlock the full potential of the solution. Perfect data does not exist, so there is no need to wait for it or make it an excuse to delay an initiative. Have a clear target vision, but also invest in less advanced solutions. The AI vision for a pricing transformation will represent a step change for most companies. Although many companies still manage to get by with Microsoft Excel-based solutions and basic pricing strategies, competitive pressures and customer demands will eventually force a move to the faster, more dynamic, and more scalable tools and techniques that AI enables. Companies need to develop their aspirational AI vision early on. Those that have limited data availability or a lower level of pricing maturity in some markets should start their journey with a less advanced approach in those areas. But at the same time, companies should develop a longer-term vision to work toward, or they risk AI being viewed solely as an incremental change. Emphasize the change effort. The development and implementation of AI pricing initiatives will demand cooperation from day one between the pricing team and the sales force. They will also require an investment in capabilities to build up advanced analytics teams. The initiatives may also shift the balance of power with respect to pricing decisions because pricing has no natural owner in most companies. Responsibility may rest with finance, marketing, or sales, or an emerging area such as RGM. The explicit support of C-level executives will make it easier for teams to cooperate and develop a comfort level with new tools. Increasingly, companies are seizing the opportunity to apply AI to pricing and launch a transformation. They recognize that the improvements are not only financial but also strategic. Companies can free up resources to focus on longer-term issues rather than short-term, tactical firefighting, and they can build on this success to launch other AI initiatives. C hapter

2 2 

2 61

J o ë l H a z a n et al .

About the survey The Massachusetts Institute of Technology and the BCG Henderson Institute surveyed more than 3,000 managers in 29 industries and 112 countries to understand how organizational learning can enhance the effectiveness of AI initiatives. The study revealed that 60% of the companies whose managers participated in the survey have realized no revenue benefits at all from their implementations of AI. Only 10% of the companies have achieved significant results from their AI initiatives, but the change efforts required were extensive and sometimes painful. The companies that applied AI to strengthen their pricing processes made some of the greatest strides in terms of AI initiatives being accepted by employees and generating financial benefits.

Bios

Joël Hazan is a managing director and partner in the Paris office of Boston Consulting Group (BCG). You may contact him by email at hazan​.joel​@bcg​.​com. Camille Brégé is a managing director and partner in the firm’s Paris office. You may contact her by email at brege​.camille​@bcg​.​com. Jean-Sébastien Verwaerde is a managing director and partner in BCG’s Paris office. You may contact him by email at verwaerde​.jean​-sebastien​@bcg​​.com. Arnaud Bassoulet is a principal in the firm’s Paris office. You may contact him by email at bassoulet​.arnaud​@bcg​.​com. Acknowledgments

The authors thank Sara Benjelloun for her contributions to this article.

Key objectives 1. Introduce AI in pricing concepts 2. Explain the impact that AI can have on pricing and organizational transformations. 3. Position pricing as a primary target to the application of AI

Key summary points 1. Pricing is a good candidate to perform AI-based analysis on transactional data. Pricing is historically a very analytical function, and price optimization methods are common in the field of pricing. 2. Pricing should also be first because the impact of pricing science on a business has been demonstrated over decades. Most companies doing pricing analytics have already prepared the data for analysis and optimization. 3. The key success factors for AI pricing projects show that a balance is needed between technical and AI dimensions and organizational dimensions. Change and process management are essential for success.

262

C hapter

2 2

W hy A I T ra n sf o rmati o n s S h o uld S tart with P rici n g

Key questions 1. What type of data might you need to be able to run AI pricing projects? 2. Who should be leading the project when applying AI to pricing? 3. Why is change management so critical for the success of AI in pricing?

References Brégé, C., Bourgouin, L., Langkamp, D., Chu, M., Beckett, M., Poirmeur, P., & Niepmann, J. (2020, November 20). Debunking the myths of B2B dynamic pricing. Boston Consulting Group. https://www​.bcg​.com​/publications​/2020​/dynamic​-pricing​-b2b​-myths. Chu, M. (2020, March 3). For important decisions, listen to your AI, but retain responsibility. BCG Gamma. https://medium​.com ​/ bcggamma ​/for​-important​-decisions​-listen​-to​-your​-ai​ -but​-retain​-responsibility​-33954bd26208. Nieto, R., Snellenberg, R., Sinha, A., Dubrovina, A., Beckett, M., & Hazan, J. (2020, May 21). A revenue management reset in consumer goods. Boston Consulting Group. https://www​ .bcg​. com ​/publications ​/ 2020 ​/covid​-19 ​- causes ​- a​- revenue ​- management​- reset​- in​- consumer​ -packaged​-goods.

C hapter

2 2 

263

C hapt e r 2 3

Digitization of B2B Pricing A Fundamental Shift Required Craig C. Zawada

Introduction Pricing: The evolving discipline

Prior to the 1990s, pricing could be characterized as ‘everybody does pricing, nobody does pricing.’ Sure, pricing decisions had to be made, but very few companies viewed it as a discipline they invested in to be especially good at. In 1992 the seminal Harvard Business Review (HBR) article ‘Managing Price, Gaining Profit’ was the spark that began to change that. In this article, Marn and Rosiello had two big ideas. Big idea 1 was that profits were extremely sensitive to small changes in price. Indexing the average economics of the top US publicly traded companies showed that a 1% increase in price could, on average, lead to an 11.1% improvement in profitability. The key question raised by this analysis was, why do so few companies invest the time and resources to be good at making sure their prices are optimized when profits are so sensitive to this management lever? Big idea 2 was the price waterfall framework. Here, they argued that many companies did not have clarity around how margins and profits were impacted by the myriad of on- and off-invoice pricing leakages that were inherent in most B2B sales transactions. Of course, companies would measure and focus on their bottom-line profitability, but few had a clear and detailed view of all the profit leakages that happen along the price waterfall when initially pricing a product or negotiating a deal with a customer. While the ideas in this HBR article generated a lot of interest, it took some time for companies to get a handle on what exactly was happening with pricing in their organizations. It was typically a daunting task to get to the root of what was actually happening with pocket prices. For example, creating a transaction database—one that includes all the on- and off-invoice discounts, terms, and conditions—often took three to four months during pricing projects conducted at McKinsey & Company because all the discount elements typically resided in disparate information systems.

DOI: 10.4324/9781003226192-29

265

Craig C. Zawada Equipment X

$5,000

Rental equipment example

$4,500 $4,000

Pocket Price

$3,500 $3,000 $2,500 $2,000 $1,500 $1,000 $500 $0 $100

$1,000

$10,000

$100,000

$1,000,000

Customer Annual Revenue Figure 23.1 Pocket price scatter. Source: PROS Holdings analysis.

When companies began to get a handle on their systems and measure all the elements in the price waterfall, many started to realize the real mess of pricing within their companies—and the pricing profession really started to take off around the year 2000. A common analysis that spurred companies to take decisive action was a pricing scatter plot. By comparing net pocket prices across a variety of transactions, most companies found complete randomness in their pricing (Figure 23.1). Executives knew that there would be some variation in pocket prices, but they expected prices to follow a pattern whereby larger customers received progressively lower pocket prices because of their better negotiating power. However, when the scatter plot analysis was done to look at the variation in prices across customers for the same product or service, many were aghast at the breadth and depth of pocket price variation. With this insight, companies realized that they must be leaving significant profits on the table with their current pricing practices. How did companies respond when they had visibility into what was happening? Most viewed pricing decisions within the company as needing greater visibility and control. Thus, they viewed pricing policies and decisions just like any other management discipline problem and applied tried-and-true management techniques used in other functional areas to pricing. Typical actions taken include these: ■

266

Pointing analytics to the problem. Purpose-built analytics for pricing is the first step toward understanding the leakages and opportunities in pricing. Sure, every company measures margins, but few had a clear insight into things like pocket prices. The rise of ERP systems made it easier to understand some of the elements of pricing but still did not provide the full picture. Off-invoice elements like shipping costs, rebates, payment terms, and so forth often resided in separate systems

C hapter

2 3

D igiti z ati o n o f B 2 B P rici n g

■

■

that had to be allocated to sales transactions. Tools like Excel and MS Access were used by many companies to create this full picture of all the pricing elements. Centralizing control. During this period, many companies added formal pricing functions within their organizations. The tasks of these pricing groups varied but largely focused on coalescing the analytical experience to gain visibility on pricing decisions, administering price changes, and managing approval processes for price exceptions. Instituting formal pricing processes. Another tried-and-true strategy to get a process under control is to add approval processes to decision-making. The ‘price exception request’ became a hallmark during this period. For one large high-tech company, more than 80% of pricing decisions had to go through a pricing ‘deal desk’ before being presented to the customer. Often, these deals went through this process three or four times before ultimately coming to an agreement. Some companies had hundreds of people who managed and approved special price requests.

Despite the added administrative burden, all these things were good for the bottom line, and for the pricing profession. Many companies improved their profits by millions of dollars by investing in the management discipline of pricing. There was, however, one major problem with this. The contextual environment for how to do pricing well changed, and pricing increasingly became a customer experience issue in addition to a management discipline issue.

Changing B2B buyer needs

While the establishment and rise of the pricing profession have been a great investment for many companies, the B2B selling landscape began to change late in the 2010–2020 decade. These changes were further accelerated by the COVID-19 pandemic. Each of these market shifts, in turn, has important implications for a company’s pricing capabilities.

The sales experience is the new key to revenue growth

Offering frictionless buying experiences has been the hallmark of many great successes in consumer products and services (e.g., Amazon, Uber). This has permanently altered the expectations for B2B buyers when they go to work and seek suppliers to fulfill their employer’s needs. The Conference Executive Board found in a recent survey that the sales experience is eclipsing traditional differentiators such as company brand, product or service features, and value to price (Cosgrove, 2015) (Figure 23.2). Imagine that you have a very intuitive and elegant website that is easy to navigate and provides all the relevant information on product features, availability, usage advice, and so forth, but that when it comes to price, it is either not provided (e.g., ‘Call for Price’) or gives very high list prices that are far from the market price. Then, even if the potential customer is interested and contacts the company, getting a mutually agreeable and market-relevant price relies on the traditional cumbersome negotiationand-approval process. If B2B buyers are increasingly demanding a frictionless buying experience, how prices are delivered to customers also must be part of that experience.

C hapter

2 3 

267

Craig C. Zawada

Figure 23.2 Relative importance of key B2B supplier attributes. Source: Conference Executive Board. Increased premium on speed

A key part of a positive sales experience is expediency. The adage that ‘time is the new currency’ is increasingly true in the B2B buying experience. For example, a supplier of electrical components was getting qualitative feedback from resellers that they were slow relative to competitors in providing responses to RFQs. Before investing in technology to improve their quoting process, they did a study that looked at their win rate on deals relative to the time it took them to provide a quote. What they found was shocking. The company’s win rate fell by 42% if they responded more than 24 hours after receiving the RFQ compared with when they responded in less than four hours (Figure 23.3). Their initial thought was that it was taking them too long to quote because of the size and complexity of the products they sold. They explored this further and did a time-and-motion study on their quoting process, finding that 68% of the time spent on quotes that took longer than four hours was due to the company’s pricing approval process. 2500

2000

60%

Total Quotes 50%

50% 36%

1500

36%

40% 32% 28%

1000

20%

500

0

30%

10%

24 hours

0%

Figure 23.3 Example of win-rate correlation with quote speed. Source: PROS Holdings analysis.

268

C hapter

2 3

D igiti z ati o n o f B 2 B P rici n g

Time increasingly costs money, and organizations need to be able to deliver marketrelevant prices quickly to meet customer buying needs.

Demand for self-service

Traditional B2B sales were historically dominated by high-touch interactions between the buyer and salesperson. Often, pricing was revealed only after there was significant back-and-forth discovery. According to McKinsey & Company, B2B sellers are seeing 90 to 120% increases in the importance of self-service in the research and evaluation phases of buying cycles (Gavin et al., 2020). That is, buyers are increasingly shortlisting vendors without interacting directly with salespeople. Making prices transparent is increasingly important during this self-discovery phase undertaken by the buyer. Without market-relevant prices available through self-discovery, sellers will increasingly miss the shortlist.

Increased trust in algorithms

B2B pricing has traditionally relied heavily on negotiated prices and discounts. A whole cottage industry of research firms would often provide market-price information to buyers who, in turn, invested in purchasing processes and policies to make sure the company did not pay more than competitors. Pricing algorithms, which do not rely on human-to-human negotiation, have also been proven to be very effective in the B2B context. A 2019 survey of over 1,000 B2B buyers by Hanover Research found that 60% of buyers prefer pricing from suppliers driven by algorithms versus a prolonged, traditional price negotiation. There is a perceived objectivity associated with algorithmic price determination which buyers find fair and trustworthy (Figure 23.4). All these buyer changes have significant impacts on the demands for pricing from B2B sellers. Sellers will capture an increasing share of the market if they can provide prices that are frictionless, fast, self-service, and market relevant.

66%

65%

63%

63%

of buyers describe data science driven pricing as

of buyers describe data science driven pricing as

of buyers describe data science driven pricing as

of buyers describe data science driven pricing as

FAIR

BENEFICIAL

TRUSTWORTHY

PREFERRED

Sophisticated vendors sometimes use data science to price quotes to ensure that they have optimized your offer by accounting for current market conditions, inventory levels, your purchase history, and cost changes. Please rate your feelings toward a vendor’s use of this approach to price your desired products and services. (N=1053)

Figure 23.4 B2B buyers trust supplier pricing derived from algorithms. Source: PROS commissioned B2B buyer survey, 2019, by Hanover Research. C hapter

2 3 

2 69

Craig C. Zawada How does B2B pricing need to change to meet the needs of the new B2B buyer?

Buyers are increasingly seeking to conduct business digitally. That only further elevates the importance of pricing excellence in the digital world. However, many companies are ill-prepared for this new high-stakes environment that they’ve been thrust into. There are several customer-centric pricing capabilities that companies need to add versus solely focusing on improving the management discipline around pricing. 1. Shift more pricing to low-touch or no-touch pricing. Buyers today don’t want a relationship. They want an efficient transaction that they believe is fair and equitable. The outdated high-touch sales cycle is out of step with buyers who have increasingly grown up in the Amazon world of e-commerce. Instead of high-touch, they want low-touch or no-touch. McKinsey’s research has found a 52% decrease in the number of buyers who prefer to buy from a sales rep in person (Bages-Amat et al., 2020). Today’s buyers want automated, guided selling. They want self-service options that let them move through the funnel and have access to market-relevant prices unencumbered. The Hanover B2B buyer research also found that 60% of buyers will pay a higher price to buy faster/instantly and sidestep drawn-out price negotiation processes (PROS Holdings, 2019). We have seen many companies move over 80% of their pricing to low- or no-touch digital channels that have helped dramatically streamline the pricing process and deliver a much-improved customer experience (Figure 23.5). 2. Use science and algorithms to deliver market-relevant prices. Years ago, sellers could offer an SKU or service for a price that remained relatively static. Today, there is an expectation that prices will move more dynamically based on supply, cost, demand, and competitive situations. Companies need the agility to quickly reset market pricing based on current conditions and deliver the right price at the right time through digital channels. For instance, unprecedented recent cost and supply volatility has put tremendous pressure on pricing groups to adjust prices more frequently. In fact, 64% of B2B buyers also indicated they would switch to suppliers that offer personalized, market-based prices. Proven pricing technologies are now available to enable companies to deliver these more dynamic prices. 3. Remove complexity when ‘high-touch’ pricing is needed. Gartner found that 77% of B2B buyers reported that their latest purchase was very complex or difficult. As mentioned previously, many companies’ pricing processes are created to solve managerial control problems, such as variations in pricing or overnegotiation. That leads to processes ostensibly designed to protect the company rather than facilitate a purchase. When a high-touch pricing negotiation is needed, companies can use the same science and algorithms to create ‘no-touch’ pricing and to provide a negotiation range to sales and sales management. This range can further be supported by facts about how much similar customers are paying for comparable products and transactions. The goal is to make the high-touch process as meaningful and frictionless as possible when needed. 4. Provide a consistent channel experience. Buyers expect that whether they talk to a salesperson on the phone or check a page on a website, they’ll see the same pricing. That unified experience is important to buyers; thus, the new digital channels require that prices be synced across channels. This often requires an upgrade from using simple tools like Excel for managing prices across channels.

270

C hapter

2 3

C hapter

2 3 

SE tom s

(C

u

(A

no

T AN us) ST ymo N I n

e

Incentive Level

Medium

20%

HIGH TOUCH NEGOTIATION

Over-reliance on “high touch”

Low

E RV n) SE know LF r

Opportunity Size

Figure 23.5 Shift to more self-service/instant pricing.

Small

Medium

Large

High

80%

Small

Medium

Large

Low

INSTANT

Incentive Level

Medium

SELF-SERVE

High

HIGH TOUCH NEGOTIATION

D igiti z ati o n o f B 2 B P rici n g

271

Opportunity Size

Craig C. Zawada

Traditional – Price Management

Digital Commerce – Customer Focused Pricing

• Protects pricing/margin

• Enables nimble market pricing

• Creates controls to manage, monitor and execute pricing

• Helps transition to no-touch pricing

• Creates pricing models to match price to value

• Implements repeatable, largely self-running dynamic pricing model • Continuous ideation and testing of pricing strategy

• 80% price execution, 20% strategy

Figure 23.6 Build the right pricing organization.

5. Build a nimbler pricing organization. The good news is that all of the advancements in pricing technologies allow pricing groups to take a more strategic role in crafting and testing different strategies to gain insight into pricing’s impact on customer behavior. This allows pricing organizations to transition from a reporting and control-type function to one that opens up new growth and profit opportunities for the company (Figure 23.6).

Bio

Craig Zawada is the Chief Visionary Officer at PROS. A widely published author, Zawada is perhaps best known for co-authoring The Price Advantage, which has been recognized as one of the most pragmatic books available on pricing strategy. Prior to joining PROS, he was a partner and leader in the North American Pricing Practice at McKinsey & Company.

Key objectives 1. Learn about the evolution of the pricing discipline and the reasons why companies have made significant investments to improve pricing practices 2. Understand recent changes in B2B buying behavior and their impact on a supplier’s pricing 3. Learn about new capabilities required because of the increased digitization of B2B pricing

Key summary points 1. The pricing profession has seen a tremendous level of investment since the early 1990s. This has led to the subsequent development of practical tools and frameworks for capturing opportunities through better pricing.

27 2

C hapter

2 3

D igiti z ati o n o f B 2 B P rici n g

2. Most of the upside in pricing came from better management discipline (e.g., applying better analytics, centralization of pricing decisions, instituting approval processes). 3. While companies found value in applying these management disciplines, it came at the cost of speed and friction to the buying/selling process. 4. B2B buyers are now demanding a more frictionless buying experience, and pricing practices must adapt and become more customer-centric. 5. Recent advancements in pricing technologies help accelerate the demand for pricing to be increasingly self-service, frictionless, and market relevant while opening opportunities for pricing organizations to elevate their strategic impact on the business.

Key questions 1. What are the key tools and frameworks developed to help better manage pricing during the development and growth of the pricing discipline? 2. What are the key contextual changes in B2B buyer behavior? What impact do these have on traditional price management practices? 3. What skills and capabilities will become more important for price strategy/management in the future and why?

References Bages-Amat, A., Harrison, L., Spillecke, D., & Stanley, J. (2020, October 14). These eight charts show how COVID-19 has changed B2B sales forever. McKinsey & Company. https:// www​.mckinsey​.com​/ business​-functions​/marketing​-and​-sales​/our​-insights​/these​- eight​- charts​ -show​-how​- covid​-19​-has​- changed​-b2b​-sales​-forever. Cosgrove, O. (2015, February 25). Challenger selling [Slides]. https://www​.slideshare​.net​/ OliviaCosgrove​/challengerselling. Gavin, R., Harrison, L., Plotkin, C. L., Spillecke, D., & Stanley, J. (2020, April 30). The B2B digital inflection point. McKinsey & Company. https://www​.mckinsey​.com ​/ business​ -functions​/marketing​-and​-sales​/our​-insights​/the​-b2b​-digital​-inflection​-point​-how​-sales​-have​ -changed​-during​- covid​-19. Marn, M. V., & Rosiello, R. L. (1992). Managing price, gaining profit. Harvard Business Review, September–October. https://hbr​.org ​/1992 ​/09​/managing​-price​-gaining​-profit. PROS Holdings. (2019). What B2B buyers want: A survey of 1053 purchasing professionals. https://pros​.com​/ learn​/white​-papers​/what​-b2b​-buyers​-want.

C hapter

2 3 

273

C hapt e r 2 4

Value-Based Offers Assisted by Artificial Intelligence Stella Parks and John V. Colias

With more and more enterprises adopting digital applications arising out of the digital transformation, data-driven analytics has become the backbone of the purchase decision process. This trend highlights a baseline requirement, understanding customer needs and aligning these needs with industry-specific trends from multiple points of view, especially suppliers, as a baseline expectation today. The pandemic accelerated the adoption of digital technologies by several years. McKinsey predicts that many of these changes could stay for the long haul (Schneider et al., 2020). Forrester’s 2021 B2B Buying Study reveals how buyers responded to the ‘new norm’ engaging in virtual activity, seeking information about sellers and their offerings before making a purchase (Caplow, 2021). The emergence of this digital transformation has created an opportunity for businesses to capture relevant insights at the right time during the sales process. This chapter focuses on how artificial intelligence (AI) can assist B2B sellers in identifying and interpreting unmet customer needs and creating value-based offers that will delight customers and exceed their expectations. Gartner (2021), a prominent technology research and consulting company, predicted that ‘75% of B2B sales organizations will augment traditional sales playbooks with AI-guided selling solutions by 2025.’ The combination of AI with value-based selling would produce digital innovations that would facilitate offering to customers ‘the right solution at the right price at the right time,’ enabling a ‘blue ocean strategy.’ The Blue Ocean Strategy defines offering in terms of value and innovation and not in terms of market share or industry attractiveness. Also, each offering should be assessed from the point of view of buyers, not in terms of supplier’s other offerings. That is how would be buyers judge offerings. (Kim & Mauborgne, 2014) On the one hand, most sellers know their company’s offerings and understand how to connect their customers’ needs to offerings that best suit their requirements. Many DOI: 10.4324/9781003226192-30

2 75

S tella Parks a n d J o h n V. Co lias

sellers have strong abilities and can clearly show why their companies’ offerings fulfill customer needs better than their immediate competitors’ offerings. However, ‘IT buyers are unable to both quantify the value proposition and clarify its business impact as important supplier weaknesses’ (Hinterhuber et al., 2021, figure 6.2). On the other hand, sellers are challenged to keep up with this fast-paced, everchanging environment that needs to be curated and refined based on what customers value today and in the future. Product and marketing teams currently help to alleviate this task at an overall level, but it takes time for sellers to develop their offers into a value-based offer to meet their customer’s specific needs. In this chapter, we focus on how the use of AI complements the sellers’ primary role in creating the best value-based offers for their customers. AI has been, and continues to be, a tool for accelerating and improving decision making. No doubt, the companies who ask the right questions after effectively mining, transporting, and analyzing their data are the companies who will lead the pack. (McGaw, 2017) Digital innovations and methodologies that make data acquisition more robust are creating new technological advancements that produce meaningful customer-level data. According to Salesforce, a cloud-based software company, AI is creating new touchpoints that result in more data collection (DiSilvestro, 2021). Such data may be used within AI algorithms to develop enhanced tools to support value-based pricing and selling efforts. These value-based pricing and selling methods derive solutions that are highly valued by both the customer and the seller, resulting in greater satisfaction gains for the customer and attractive profits for suppliers. Toytari and Rajala (2015) defined customer value as the ‘difference between the perceived benefits received and the perceived sacrifices made by a customer’ and valuebased selling as ‘a sales approach that builds on identification, quantification, communication, and verification of customer value.’ Hinterhuber (2017) points out that to realize the benefit of value-based pricing and value-based selling, the quantification of value is essential. The ability to quantify value becomes even more crucial as premium pricing decisions are beginning to be made by machine algorithms (Hinterhuber & Liozu, 2018). In a literature review on value-based marketing for industrial smart services and data-driven services that use digital technology, Classen and Friedli (2019) conclude that more research is needed in analytic methods to quantify value and develop value-based prices. In the present study, we outline an AI methodology that uses a choice modeling method to measure customer willingness to pay for optimal features and premium prices that are designed to produce a solution valued by the customer and more profitable to the supplier. This methodology focuses on certain data requirements: ■ ■

2 76 

Customer-level data for solutions offered to customers cross-referenced with the outcome of the offer (purchase or no purchase). Multiple offers made to a statistically significant number of customers. C hapter

2 4

Value - B ased Offers A ssisted by A I ■ ■

Sufficient variation in features included in the solution and the offer price. Data engineering that automatically generates the required data.

Based on the data, machine learning and optimization are automated to produce suggested solutions in real time. Sellers can use the database to create the next-best offer for the customer while the AI generates the relevant questions that sellers need to ask customers. The methodology uses a mixed logit model to measure customer perceived value and predict responses to alternative future sales offers. The model is estimated via hierarchical Bayes and machine learning, delivering customer-level parameter estimates. The parameters provide intelligence about what customers value. This intelligence feeds the seller through an AI platform to create value-based offers. Within the AI platform, AI and sellers each play critical roles. AI and seller’s roles can be grouped into seven factors that make a difference in creating value-based offers (Table 24.1). AI methodology

The core of the AI methodology is automation which includes the following: ■ ■

■

Recording of the features, prices, and purchase outcomes of sales offers (descriptive). Machine learning to continuously ingest offer data, tune a customer choice model, and output customer-level value of offer features and price sensitivity estimates (predictive). Piping of feature values and price sensitivity estimates into an optimization algorithm that produces customer- and/or segment-level solutions (AI-assisted prescriptive).

Table 24.1  AI

and seller roles

AI and seller roles in an AI-assisted value-based selling platform AI’s role

Seller’s role

1 Ascertain the customer’s best available alternative options. 2 Predict the customer’s maximum willingness to pay and the extra value of each feature as perceived by the customer 3 Establish target prices before sales personnel enter negotiations with customers: 4 incorporating management-approved discounting rules 5 allowing flexibility to explore exceptions to the rules for strategic situations 6 Systematically monitor and provide a dashboard summary of closed price deviations to sales leaders and other decision-makers (pricing team) to adjust the price as needed.

1 Establish and build customer relationships to engage in the customer experience journey. 2 Justify and ask for management approval to deviate from list prices. 3 Quantify the value of the final offer with input from the customer: 4 understand customers' unmet needs 5 walk away from unprofitable deals 6 determine when additional free or discounted services could be offered as incentives to close a deal

C hapter

2 4 

27 7

S tella Parks a n d J o h n V. Co lias ■

Interactive access and use of customer solutions by sales personnel to assist in their value-based pricing and selling process (seller-enhanced prescriptive).

Stages of AI-assisted value-based selling automation​

Mixed logit customer choice models are recommended as the foundation of the predictive and prescriptive elements of the AI methodology (Figure 24.1). This type of model has become the gold standard in marketing research to measure feature value and price-value tradeoffs (Hensher & Greene, 2003; McFadden & Train, 2000). In contrast to mixed logit customer choice in most applications within the field of marketing research (Chrzan & Orme, 2000), the AI-assisted methodology would rely primarily on actual purchase-history data (revealed preference) rather than primarily on survey data (stated preference) but supplemented with stated preference when needed due to lack of variability or high correlation of price and feature attributes. When hypothetical choice data is used to supplement historical offer data, then the AI-assisted methodology would combine stated and revealed preference data to estimate the mixed logit model. The benefits of combining stated and revealed preferences to improve predictive accuracy have also been established in the academic literature (Adamowicz et al., 1994; Brownstone et al., 2000; Walker & Ben-Akiva, 2002). Machine learning can further improve predictive accuracy of the mixed logit model by tuning certain hyperparameters (Colias et al., 2021). Mixed logit customer choice models provide an economic theory-based measure of perceived value for product and service solutions and price sensitivity (Ben-Akiva & Lerman, 2018; Train, 2009; Train & Weeks, 2005). More specifically, two economic concepts are quantified, namely the marginal utility of adding or changing each offer feature and the marginal utility loss of paying an extra dollar. Dividing the former by the latter produces an estimate of customer perceived value. The quantification of feature value is essential to value-based pricing and selling (Hinterhuber, 2017), and the mixed logit estimates would complement the seller’s efforts to achieve such quantification as part of the value-based pricing and selling process. The mixed logit model not only is well integrated with economic theory through the random utility model (Hensher & Greene, 2003; McFadden & Train, 2000) but also approximates any choice model with any distribution of marginal utilities, that is, any distribution of preferences among features and prices. Econometric methods, such as hierarchical Bayes and maximum simulated likelihood, have enabled the mixed logit choice model to deliver unique estimates for each Stages of AI-Assisted Value-Based Selling Automation Prescriptive (What actions/solutions would be best?)

Descriptive (What happened?)

Predictive (What will happen?)

Figure 24.1 From predictive insights to prescriptive personalized recommendations.

278

C hapter

2 4

Value - B ased Offers A ssisted by A I

individual customer. Machine learning avoids overfitting historical data, thus improving predictive accuracy, and providing higher-quality value-based solutions. Digital innovations can automatically record customer-level sales offer data. Within the context of enterprise sales (B2B), customers are approached multiple times with offers of product and service solutions, providing sufficient information for hierarchical Bayes to produce high-quality individual-level utility estimates. In the event that variation in solutions and pricing within the sales-offer data prohibits high-quality individual-level estimates, then hierarchical Bayes automatically shrinks to a market average result. Alternatively, hypothetical tradeoffs among alternative customer solutions can be presented to customers, within the digital platform, to gather additional data that supplements the sales-offer data to enhance the quality of the individual-level utility estimates (Ben-Akiva & Morikawa, 1990; Hensher & Bradley, 1993). More recently, machine learning methods have been introduced to further optimize the predictive accuracy of mixed logit models without sacrificing their tie-in with economic theory (Aboutaleb et al., 2021; Colias et al., 2021; Van Cranenburgh et al., 2021). Machine learning (ML) techniques can be used to tune the shape of the distribution of marginal utilities across customers, enabling the model to better predict customer-level perceived value for each feature of the customer solution (Colias et al., 2021). As new sales-offer data become available within the digital platform, the mixed logit prediction models would flow into a nonlinear mixed-integer programming algorithm to simulate and optimize the solutions for each customer segment or each individual customer. Any desired definition of customer segment can be used. The customer segment is ingested into the AI platform and used within the nonlinear programming algorithm, automatically. While the mixed logit model does indeed produce customer-level values for each solution feature, segment-level solutions are often advantageous for the purpose of developing a consistent value-based pricing and selling strategy across sellers and their customers (Hinterhuber, 2004). The final deployment of the AI methodology would be seamlessly integrated into the digital sales platform and software, so that all the automation takes place under the hood, so to speak. Ultimately, the seller would be responsible for the final result. The AI methods would complement the seller’s implementation of value-based pricing and selling. For example, suppose the seller offers products A and B, along with optional features A1, A2, B1, and B2. Contract elements can be either C1 or C2. The seller would like to offer a new product feature B3 and contract element C3, both having never been offered before. The AI-assisted platform would include a dashboard with historical data on percentage of sales offers accepted for each combination of A1, A2, B1, and B2, along with the closing price (actual and optimal). The seller would enter a description of new feature B3 and contract element C3. Then, in conversation with the buyer, the seller would ask the buyer to review several potential offers, some of which include new feature B3 and contract element C3. The buyer would state preferences among hypothetical alternative offers via a web page, and stated responses would be automatically added to the

C hapter

2 4 

279

S tella Parks a n d J o h n V. Co lias

digital database. The AI-assisted dashboard would immediately be updated to include optimal offer prices for product offers that include the new feature and contract element. All these steps would be integrated into the seller’s conversations with the buyer, in real time.

Creating value-based offers assisted by AI: ML choice model with optimization Sequence to build AI-assisted value-based selling platform

1. assess the company’s current sales databases, including solutions offered with pricing 2. identify missing data required for mixed logit modeling 3. beta test the ML mixed logit modeling method and offer an optimization algorithm with the available data. 4. design data pipeline and consolidate relevant databases 5. automate data pipeline, ML, and mixed logit modeling, offer optimization 6. design and build a platform (digital innovation) for integration of AI and seller roles in AI-assisted value-based selling—coding/programmers 7. beta test new platform 8. deploy new platform

Making the value-based offers relevant and personalized

The AI-assisted value-based methods presented thus far would quantify perceived value (even for new and innovative attributes of offers); it would enable sellers to offer the right solution at the right price and rapidly develop new solutions. However, several additional elements must be incorporated to ensure that solutions are relevant, personalized, and offered at the right time. First, sellers must understand their customers’ external trends related to economic, technological, societal, geopolitical, and regulatory requirements as well as internal compelling events such as contract renewal/expiration. AI can assist sellers to move more quickly and seize opportunities, while the sellers’ deep understanding of customer needs enables them to manage customers’ expectations and relationships. Second, sales, product, pricing, and executive leadership must be aligned in focusing on a customer mindset. ‘The mindset of the customer as the center of all decision making is a critical component, as well as the skills and tools for dealing with Big Data’ (Tanner, 2014). In Figure 24.2, ‘hypothetical solutions evaluated’ identifies new attributes, as many of the attributes valued by customers in the past were based on past actions and outcomes with little variability among historical offers. The platform depicted in the figure can test many scenarios simultaneously: not just simple A3, B3, C3 testing but as many as a dozen different versions. By cycling through many tests, we can accelerate learning of the customer mindset and appropriately direct the conversation among sales and executive leadership. Third, as stated at the beginning of this chapter, our goal is to create digital innovations that would enable offering customers ‘the right solution at the right price at the right time.’ Up to now, our discussion has been focused mainly on how the AI-assisted value-based selling platform will deliver ‘the right solution at the right price’ portion

280

C hapter

2 4

C hapter

2 4  Customer Choice Model

YES

Right Offer? (Optimization Decision Tool)

NO

Automated & Continuous updates

Potential Actions

Value-based offers with an Optimal Pricing

Figure 24.2 Value-based offers assisted by a machine learning choice model.

Solutions Offered But Not Purchased

Hypothetical Solutions Evaluated

Add new attributes

Solutions Purchased

Machine Learning: Mixed Logit at Customer-Level

Feedback

Creating Value-based offers assisted by AI: Machine Learning Choice Model with Optimization

Value - B ased Offers A ssisted by A I

281

S tella Parks a n d J o h n V. Co lias

of the equation. For now, let’s dive deeper into the final portion of the equation ‘at the right time.’ One of the greatest sources of disagreement between marketing and sales is knowing the right time when customers and sellers are ready to talk. Today, digital marketing has been deployed to target not only who but also when to offer the right solution. For example, Bombora, a data cooperative of B2B publishers, helps identify the businesses that are ready to buy based on privacy-compliant intent data. Forrester claimed that ‘Bombora’s content consumption model has become the de facto standard in B2B marketing for third-party behavioral data to indicate intent’ (Bombora, 2021). Fourth, when trying to make value-based offers personal, more emphasis should be placed on customers’ share of wallet rather than short-term profit or market share. The ‘optimization decision tool’ in Figure 24.2 would enable sellers to adopt the right activities today to increase future profitability. By selecting a next-best offer that maximizes net present value (NPV) subject to constraints on prices and strategic product attributes, it enforces the focus on the customer instead of products as the driver toward profitability. The optimization decision tool would be used to promote retention, deliver ROI from acquisition efforts, and identify profitable solutions that customers value. Moreover, it would provide insight into the size, distribution, and profile of customers that can be used to develop strategies for increasing retention and loyalty programs. Personalized value-based offers would be optimized at the right price so that suppliers can quickly determine what works and what doesn’t work with a greater assurance of being right, rather than best-effort trials. In addition, the AI tool would predict the willingness of customers to pay a premium price with prescriptive loyalty program offers, especially if switching costs are high. Other available variables such as tolerance to forgive a bad customer experience, greater frequency of purchases, better net promoter score (NPS), and initiating referrals could be incorporated into the AI-assisted platform depicted in the figure as potentially significant triggers of new offers. These types of AI-driven insights would assist sellers to offer not only the right solutions and price but also better understand, predict, and engage customers at the right time.

Making the platform agile, aligned, and actionable

The AI-assisted platform presented here would enhance speed and agility in developing new solutions for customers. Agility is a necessary condition for a successful platform, but agility combined with an aligned shared vision among specific customers, the community of value-added supplier partners, and the selling company is necessary and sufficient to create a sustainable outcome. Indeed, a shared vision, purpose, and values through trust and collaboration are required for sustainability. By sustainable, we mean that customers’ needs are met or exceeded and that companies are profitable. To create collaborative and scalable solutions, the primary questions in Figure 24.3 can help structure a balanced digital innovation ecosystem. The company can leverage insights from the AI-assisted value-based platform to identify the distinct target customers and extend the right solutions to acquire new customers who are also attracted to the value-based offers.

282

C hapter

2 4

Value - B ased Offers A ssisted by A I Balanced Digital Innovation Ecosystem

Customer

Community

Company

How well the supplier positioned to meet their customer’s needs than alternative solutions?

How the supplier capture complementary innovation from digital platform?

How the supplier scale the solutions and stay relevant and differentiated in the marketplace?

Aligned on Shared Vision, Purpose, and Values through Trust & Collaboration

Figure 24.3 Balanced digital innovation ecosystem.

One way to produce sustainable solutions would be to use the AI-assisted valuebased platform to implement five steps to scale and acquire new opportunities: 1. 2. 3. 4.

Identify customers who like a given value-based solution. Find the top complement attributes, those relevant to the broader customer base. Determine which driving attributes really impact the supplier’s customers. Understand how the complementary attributes change what customers value and how they impact current offerings. 5. Develop value-based complementary solutions dynamically to be recommended for sellers. These five steps would ensure sustainability through scale, especially in those industry categories where short-term profit must be sacrificed initially in order to achieve sufficient volume (scale) to produce overall profitability. The AI-assisted platform would find broadly desired attributes (complement attributes) across all customers, force those attributes to be included in a profit-maximized, or loss-minimized solution, and pick the best initial customers to offer the solution. The presence of the complement attributes would ensure broad demand, enabling sales to a broader set of customers, and ultimately creating sufficient business volume to take advantage of economies of scale for future profitability. Thus, the true success of an AI-assisted value-based pricing platform cannot be accomplished in a vacuum but through trusted partnerships with initial customers and a community of value-added suppliers. For example, if the complementary attributes are not offered today, it is through conversation among sellers and customers that such attributes would be identified, and the AI-assisted value-based pricing platform would enable the rapid optimization of price. Successful AI-assisted value-based platforms must be the product of effective community collaboration to solve industry challenges together and create a win-win-win ecosystem for customer-community-company to open new ways of thinking. Just like the African proverb, ‘If you want to go fast, go alone. If you want to go far, go together.’

Bios

Over 27 years at AT&T, Stella S. Park has held a variety of leadership roles in sales, customer research and analytics, global marketing, strategy and execution, and currently in global business channel marketing. She provides customer-centric and actionable insights C hapter

2 4 

283

S tella Parks a n d J o h n V. Co lias

to assist sales executives in making informed decisions and executes strategic initiatives. She has published in the Journal of Marketing Analytics and received the award for Best Paper of the Year in 2021. She holds an MBA from Pepperdine University, a Master’s in Telecommunications from the University of Dallas, and a Bachelor of Science in Applied Mathematics from UCLA. She can be contacted at parkstella777​@gmail​.c​om. As a leader with both university teaching and business consulting experience, John V. Colias, PhD, focuses on predictive modeling and prescriptive analytics. As Senior Vice President, Research and Development, at Decision Analyst, John combines academic and business interests to help analytics professionals by offering cutting-edge analytic solutions tempered by business realism. He holds a Doctorate in Economics from the University of Texas at Austin, with specializations in econometrics and mathematical modeling methods. He can be contacted at jcolias​@decisionanalyst​.c​om.

Key objectives 1. Understand how AI can assist sellers in delivering value-based offers to their customers. 2. Explain the core of an AI methodology that can develop ‘the right solutions at the right price at the right time’ for B2B customers. 3. Define the respective roles of AI and sellers. 4. Make the value-based offers relevant and personalized. 5. Expand the platform to be sustainable and scalable (agile, aligned, actionable).

Key summary points 1. A rapid adoption of digital transformation has created an opportunity for business to capture insights at the right time during the sales process. 2. AI should augment, not replace, sellers’ roles and responsibilities, especially in quantifying value at the optimal price for a specific customer. 3. The true success of an AI-assisted value-based pricing platform cannot be accomplished in a vacuum but accomplished through trusted partnerships with initial customers and a community of value-added suppliers.

Key questions 1. What are the roles and responsibilities of AI versus sellers in an effective and collaborative sales organization? 2. What would be the competitive advantage of using an AI-assisted value-based selling platform? 3. How should we use digital transformation to capture customer preference insights to optimize valued-based offers? 4. How can we make valued-based offers to create new innovations to scale and lower the price?

284

C hapter

2 4

Value - B ased Offers A ssisted by A I References Aboutaleb, Y. M., Danaf, M., Xie, Y., & Ben-Akiva, M. (2021). Discrete choice analysis with machine learning capabilities. arXiv preprint. https://doi​.org​/10​.48550​/arXiv​. 2101​.10261. Adamowicz, W., Louviere, J., & Williams, M. (1994). Combining revealed and stated preference methods for valuing environmental amenities. Journal of Environmental Economics and Management, 26(3), 271–292. Ben-Akiva, M., & Lerman, S. R. (2018). Discrete choice analysis: Theory and application to travel demand. Transportation studies. MIT Press. Ben-Akiva, M., & Morikawa, T. (1990). Estimation of switching models from revealed preferences and stated intentions. Transportation Research Part A: General, 24(6), 485–495. Bombora. (2021). We’re not just the leader in B2B Intent data. We invented it. Retrieved December 10, 2021, from https://bombora​.com ​/our​-data/. Brownstone, D., Bunch, D. S., & Train, K. (2000). Joint mixed logit models of stated and revealed preferences for alternative-fuel vehicles. Transportation Research Part B: Methodological, 34(5), 315–338. Caplow, B. (2021, April 14). Three seismic shifts in buying behavior from Forrester’s 2021 B2B buying study. Forrester. https://www​.forrester​.com ​/ blogs​/three​-seismic​-shifts​-in​-buying​ -behavior​-from​-forresters​-2021​-b2b​-buying​-survey/. Chrzan, K., & Orme, B. (2000). An overview and comparison of design strategies for choicebased conjoint analysis (Sawtooth Software Research Paper Series No. 98382). https:// sawtoothsoftware​.com ​/resources ​/technical​-papers ​/an​- overview​-and​- comparison​- of​- design​ -strategies​-for​- choice​-based​- conjoint​-analysis. Classen, M., & Friedli, T. (2019). Value-based marketing and sales of industrial services: A systematic literature review in the age of digital technologies. Procedia Cirp, 83, 1–7. Colias, J. V., Park, S., & Horn, E. (2021). Optimizing B2B product offers with machine learning, mixed logit, and nonlinear programming. Journal of Marketing Analytics, 9, 157–172. DiSilvestro, A. (2021). Rise of the chatbots: How AI changed customer service. Retrieved December 10, 2021, from https://www​.salesforce​.com ​/products​/service​- cloud ​/ best​-practices​ /how​-ai​- changed​- customer​-service/. Gartner. (2021). Gartner predicts 75% of B2B sales organizations will augment traditional sales playbooks with AI-guided selling solutions by 2025. https://www​.gartner​.com ​/en ​/ newsroom ​/press​-releases​/gartner​-predicts​-75-​-of​-b2b​-sales​-organizations​-will​-augment​-tra. Hensher, D. A., & Bradley, M. (1993). Using stated response choice data to enrich revealed preference discrete choice models. Marketing Letters, 4(2), 139–151. Hensher, D. A., & Greene, W. H. (2003). The mixed logit model: The state of practice. Transportation, 30(2), 133–176. Hinterhuber, A. (2004). Towards value-based pricing—An integrative framework for decision making. Industrial Marketing Management, 33(8), 765–778. Hinterhuber, A. (2017). Value quantification capabilities in industrial markets. Journal of Business Research, 76, 163–178. Hinterhuber, A., & Liozu, S. M. (2018). Thoughts: Premium pricing in B2C and B2B. Journal of Revenue and Pricing Management, 17(4), 301–305. Hinterhuber, A., Snelgrove, T. C., & Stensson, B.-I. (2021). Value first, then price: The new paradigm of B2B buying and selling. Journal of Revenue and Pricing Management, 20, 403–409. Kim, W. C., & Mauborgne, R. (2014). Blue ocean strategy, expanded edition: How to create uncontested market space and make the competition irrelevant. Harvard Business Review Press. McFadden, D., & Train, K. (2000). Mixed MNL models for discrete response. Journal of Applied Econometrics, 15(5), 447–470. McGaw, S. (2017, November 15). Lead your company’s digital transformation with these 5 questions. LinkedIn. https://www​.linkedin​.com ​/pulse ​/ lead​-your​- companys​-digital​ -transformation​-5​-questions​-steve​-mcgaw/. C hapter

2 4 

285

S tella Parks a n d J o h n V. Co lias Schneider, J., Smaje, K., O’Toole, C., & LaBerge, L. (2020, October 5). How COVID-19 has pushed companies over the technology tipping point—And transformed business forever. McKinsey & Company. https://www​.mckinsey​.com ​/ business​-functions​/strategy​-and​ -corporate​-finance​/our​-insights​/ how​- covid​-19​-has​-pushed​- companies​- over​-the​-technology​ -tipping​-point​-and​-transformed​-business​-forever. Tanner, J. F. Jr. (2014). Analytics and dynamic customer strategy: Big profits from big data. John Wiley & Sons. Töytäri, P., & Rajala, R. (2015). Value-based selling: An organizational capability perspective. Industrial Marketing Management, 45, 101–112. Train, K., & Weeks, M. (2005). Discrete choice models in preference space and willingnessto-pay space. In R. Scarpa & A. Alberini (Eds.), Applications of simulation methods in environmental and resource economics (pp. 1–16). Springer. Train, K. E. (2009). Discrete choice methods with simulation. Cambridge University Press. Van Cranenburgh, S., Wang, S., Vij, A., Pereira, F., & Walker, J. (2021). Choice modelling in the age of machine learning. arXiv preprint. https://doi​.org​/10​.48550​/arXiv​. 2101​.119. Walker, J., & Ben-Akiva, M. (2002). Generalized random utility model. Mathematical Social Sciences, 43(3), 303–343.

286

C hapter

2 4

C hapt e r 2 5

Digital Transformation How to Convert a Buzzword into Real Bottom-Line Value Mitchell D. Lee and Darius Fekete

Demystifying digital transformation

Global enterprises are being stretched in previously unimaginable ways as they face massive disruptions in supply chains, customer demand, and geopolitical upheaval. Organizations have been forced to rapidly pivot, shift business priorities, and accelerate digital adoption. All eyes are on digital technologies—but the tech industry abounds with acronyms, jargon, and buzzwords. Industry professionals, technology vendors, and elite consultancies alike have used the term digital transformation to describe a plethora of concepts ranging from artificial intelligence, IoT, machine learning, digital commerce, and others. Its broad and generic use has caused digital transformation to join the ranks of the ever-growing list of technology buzzwords, and even lose its original meaning. But not all is lost. Yes—at face value, digital transformation does now feel like just another buzzword, but it is a straightforward business imperative that has never been more relevant. It just needs some historical context and an understanding of its original, concrete definition. The roots of digital transformation date back to the 1940s when Claude Shannon laid the foundation for digitization and digitalization in his paper ‘A Mathematical Theory of Communications’ (Menear, 2020). There are some important terminology distinctions at play here. Digitization, for example, is defined by Gartner (n.d.-a) as the conversion of analog technology to digital format. If digitization is about digital formats, digitalization is about using those digital capabilities. Gartner (n.d.-b) further defines digitalization as ‘the use of digital technologies to change a business model and provide new revenue and value-producing opportunities.’ Some business leaders use the terms digitalization and digital transformation interchangeably, but there are fundamental differences between the concepts. While digitalization is focused on the use of technologies, digital transformation insinuates a much deeper and broader adoption that involves cultural and behavioral shifts. Digital DOI: 10.4324/9781003226192-31

287

M itchell D. L ee a n d D arius F ekete

transformation is not just about technology. It involves making significant organizational shifts around your people, processes, and business models. Successful digital transformation unlocks extraordinary business value. Empowering B2B buyers and sellers with digital commerce

The COVID-19 pandemic ushered in a new era of digital transformation, urgently forcing organizations around the globe to speed up digital initiatives to drive their response to the crisis. McKinsey & Company (n.d.) affectionately refers to this phenomenon as ‘the quickening’ and reported that 10 years of e-commerce adoption was compressed into three months during 2020. Now the C-suite is focused on prioritizing investments in the right initiatives that will fuel future growth. While digital transformation impacts areas like IoT, supply chain, logistics, and production, digital commerce should be a top priority. COVID-19 was not just a catalyst for the digital acceleration of internal business processes. It represented a fundamental shift in buying behavior and expectations. Today’s buyers are omnichannel creatures who expect to be able to weave in and out of channels without interference or interruption. Research shows that 82% of B2B buyers now expect a B2C-like buying experience; yet only 27% of B2B buyers say suppliers excel at meeting these expectations (BusinessWire, 2019). To remain competitive, B2B sellers need to provide highly personalized and accurate offerings—the right product, at the right time, at the right price. In short, B2B buyers must be empowered to self-serve across all the ways that a company meets its customers in the market. But remember, deploying digital commerce is not just about enabling your customers; it is also a win for your employees. In B2B, digital commerce empowers your sellers to shift from catalog-explainers and ordertakers to innovation experts and value-drivers. If you are looking for ways to enable your B2B sellers to partner with your customers for value creation and delivery, it is time to put digital transformation at the top of your priority list.

Improve the omnichannel experience with deal negotiation

If the last few years have taught us anything, it is that virtual and online sales are the new standard. Unfortunately, for many B2B organizations, e-commerce often fails to deliver a better customer experience than traditional channels. Insufficient product descriptions, lack of transparency into product availability, and slow response times are among the most-cited reasons (Sana, 2022) for buyer reluctance to purchase online. Additionally, most e-commerce sites do not allow for the same quality of engagement as a phone call to a sales rep. Buyers cannot cite a competing offer or try to negotiate commercial terms to achieve a better deal. E-commerce sites often function as an order-entry system rather than as solutions that aim to make the buyer’s job easier. The good news is the current shift to virtual sales creates a massive opportunity to delight B2B buyers and generate higher engagement. The challenge for B2B e-commerce platforms and marketplaces is to improve the customer experience without adding complexity to the purchasing process. Modern solutions need to deliver relevant product information as well as contextual and personalized pricing to the customer. Additionally, incorporating the last price paid into 288

C hapter

2 5

D igital T ra n sf o rmati o n

the offer is a standard expectation from B2B buyers in most industries. Reimagining negotiations for digital must go beyond simply modernizing the digital selling technology. But before we get into digital negotiation, let us explore a few additional cuttingedge technologies that can improve the digital B2B experience if planned properly. These technologies may have once been considered buzzwords but are very clearly moving from their infancy to maturity, driving productivity, efficiency, and actual business outcomes in the enterprise space. AI governance and transparency, for example, need to be taken seriously. Every day we create 2.5 quintillion bytes of data. This massive amount of data is difficult to unify, manage, and govern. But AI is growing more sophisticated and being applied to analytics, customer data, and business processes. Your organization should set guidelines for responsible, ethical, empathetic, and strategic use of data for AI. We share more on this topic in an upcoming section. Extended reality (XR), like augmented reality and virtual reality, is also drastically impacting the business landscape. In 2020 Gartner said that augmented reality had matured so rapidly that it was no longer considered an emerging technology. In fact, by 2021 a third of all large organizations were predicted to deliver multidimensional experiences, including extended reality (Herdina, 2020). Clearly, XR will be an important differentiator in customer engagement strategies over the next few years. It is already common to see AR usage in CPQ solutions, helping sellers meet their customers’ needs with easily viewed and manipulated images of the offerings. Advanced technologies like robotic process automation (RPA) and machine learning (ML) continue to automate and simplify human work. Often, business processes are being automated. When it comes to pricing and selling, automation empowers teams to look ahead into the sales and operations planning process, to finally integrate the impact of pricing changes on demand and to balance and plan capacity utilization. Now, back to digital negotiation. Omnichannel buyers require you to have an aligned sales approach throughout the entire buying journey, especially when discussing commercial terms. Negotiating typically requires several iterations between sales and customers. Back-and-forth discussions, offer reviews, deal desk reviews, and approvals will undoubtedly lengthen the deal cycle. To mitigate associated risks, and improve deal velocity, consider digital negotiation. Digital negotiations add a controlled dialogue to the virtual selling process and allow you to offer a win-win environment to your customers. Buyers have a chance to use the e-commerce platform to control spending, while sellers remain in control of the process and can manage the potential trade-offs. Digital negotiation provides customers the option to request a different price or different deal terms. Sellers enable a controlled bargaining process online to facilitate the sales conversation. Automation supports repeating requests from buyers, while predefined approval levels help manage the long tail. Digital negotiation frees up sales rep time to focus on higher value and more complex deals. Let us look at an example of how digital negotiation works (see Figure 25.1). As a buyer, I am purchasing an item from a distributor’s website. My selected product sells for €6.83 for 100 units. The price is individual, already considering my relationship with the seller. However, I am not happy with the price, and I want to pay less. The platform enables me to make an alternative offer, and €6.50 seems about right. Once

C hapter

2 5 

289

M itchell D. L ee a n d D arius F ekete

€ 6.83 / unit For 100 items

Make us an offer

€ 6.50 / unit

Purchase 29 items more

Figure 25.1 How deal negotiation works.

I enter my proposal, the system informs me that I can buy at that price if I purchase 29 more items. ‘Increase the quantity, and we can support the price.’ I agree and make the deal. Easy, frictionless, and without the need to involve a sales rep for a trivial trade-off. What is important? That the whole negotiation is conducted through a platform, without making phone calls or waiting for email responses. Engagement created by allowing me to make an offer increases the likelihood to transact. Plus, the counteroffer raised the overall sales value. Digital negotiation also allows faster deal closures, as it supports prompt responses and shortens the sales cycle. Deploying a solution for bargaining and sales conversations within self-service channels improves deal velocity.

Where human intelligence and artificial intelligence converge

Artificial intelligence is one of the most transformative and powerful technologies of our time, and it has been growing more sophisticated by the day. One of the earliest and most astounding milestones in the evolution of artificial intelligence occurred in 1997 when Gary Kasparov, a grandmaster, lost a game of chess to IBM’s Deep Blue computer. Of course, Kasparov demanded a rematch, which IBM declined. However, in 1998 Kasparov held a ‘Centaur Chess’ game where teams were formed by combining humans + AI. The combination was a nod to the mythological creature that was half-horse, half-human. The results? A human + AI ‘centaur’ beat the solo computer. This landmark event represents an important distinction about the power of AI: on its own, artificial intelligence is strong, but in many cases, it needs to be combined with human intelligence to truly work its magic. AI is strong on its own in situations that are well-defined, with clear rules and boundaries for operation—think of the handsfree parallel parking feature available in many automobiles. But in situations that are not as clearly defined—such as running a business—AI should be viewed as a foundation, with human intelligence as a supervisory leader. When organizations harness the

290

C hapter

2 5

D igital T ra n sf o rmati o n

power and scalability of AI alongside the intelligent decision-making of an experienced, business-savvy, sentient human being, that is when they win. Automation and AI have changed the pricing game for good. AI can comb through massive volumes of data sets to deliver a more accurate analysis of information and give you better visibility into essential information like margins, revenue, and profit. Of course, one of the best ways an organization can improve its margins is through pricing. But many organizations are inundated with an abundance of pricing information and data. Pricing automation that utilizes artificial intelligence can help dramatically cut through the massive volume of information and deliver a concise synthesis of the dataset. Pricing automation uses software to automatically manage the pricing process. Computers can find pricing data, analyze it using AI, and develop price guidelines much quicker than even the best price strategist. But that does not mean humans do not have a place in automated pricing. In fact, the process behind automated pricing is remarkably similar to the process a human being would follow. Here is an example: 1. compile price data from the web (i.e., from internal transaction data, competitors’ websites, validated market information, and other credible sources) 2. factor in business rules, costs, etc. 3. analyze the data 4. make pricing recommendations 5. monitor the market and competitors’ actions and adjust as necessary The difference is, in pricing automation, the entire process is done by pricing software. This is especially helpful when it comes to gathering price data and monitoring the market; machines can do this constantly and instantly, which makes the entire process faster and more dependable. The problem, however, is that many commercial AI and machine learning pricing solutions do not explain what is being analyzed or how the results get calculated. To add to the frustration, many of these tools do not allow users to specify exactly what criteria they are looking for from the data or to customize how the data is processed. There is often no way for a user to override the recommendations and impart human-influenced nuance into the way the technology processes the inputs. In black-box and even glass-box environments, the user is forced to sacrifice control and rely on the predetermined methodologies embedded in the technology. You can imagine how this creates challenges for a sales team that needs to know x, but the technology that supports them only knows how to produce y. And this problem persists whether the solution is black box or glass box, since both lack an element of human control. A black-box artificial intelligence solution, or any type of technology solution for that matter, refers to a piece of technology that does not reveal the inner workings of its process to its users. The user delivers inputs to the technology, and the technology spits forth an output, but what happens between those two steps is a mystery. Meanwhile, a glass-box solution enables you to view the steps in the middle. Black-box pricing solutions are the most common in the market but also the least transparent. Glass-box pricing solutions are slightly better than black box because they

C hapter

2 5 

291

M itchell D. L ee a n d D arius F ekete

provide the target price but also show the calculations—how the decision was made—a complete picture of all the variables that affected the final price that a customer was willing to pay. But even glass-box technology has its downsides. Witnessing something firsthand is incredibly helpful in deepening your understanding of a concept—but just standing on the sidelines gets old fast. Best-in-class intelligent pricing solutions go one step further than a glass box. Centaur intelligent pricing provides all the visibility of a glass-box solution with the added ability to apply human intelligence and thus maintain control. Even with AI, technology by itself is not always smarter than your sales team. That is why the human component is so critical to intelligent pricing. The centaur model enables you to adjust a price based on specific information that was not included in the original AI-driven pricing model. Chances are, someone on your sales team knows something—a particular fact about a customer, for example—that a machine, no matter how advanced, could simply never know. Therefore, make sure the technology they are using daily allows any necessary changes or updates to be made (Figure 25.2). In the sales process, a centaur model gives you an added layer of control. You can manually prune out a variable that is affecting pricing, such as geography, supply levels, or previous discounts. For example, a buyer at a company may have been grandfathered into a discount that is longer offered. When that buyer leaves the company, the salesperson can change that pricing model to remove that discount going forward. An AI-driven pricing algorithm might not know these specific variables. But a centaur approach allows for human intervention. Manual adjustments from your team can have a significant impact on margins and the bottom line, especially when your sales reps are bringing a wealth of experience and nuance about specific customers to the table. Centaur solutions, like intelligent pricing solutions from Vendavo, allow you to be confident in your pricing every time. Your sales team has the flexibility to consider continuous market dynamics and offer supervisory control over the AI that lets you employ your business judgment where appropriate. A win-win-win. AI Pricing Solution Features

Black Box

Glass Box

Centaur AI

User Adds Inputs

√

√

√

AI Generates a Price Recommendation

√

√

√

User Sees Outputs

√

√

√

√

√

User Sees “Under the Hood” Details of the Process User Can Make Changes Based on Business Judgement

√

Outputs Accurately Reflect the Nuance of Each Customer

√

Figure 25.2 Glass-box, black-box, and centaur AI pricing solution features.

292

C hapter

2 5

D igital T ra n sf o rmati o n

By using technology that combines human understanding, experience, and contextual insight with AI’s tireless processing power, decision-makers can leverage their data to act promptly and decisively. They can adapt to current circumstances as well as plan for strategic improvements. Relying solely on machines or solely on humans will not win the day. But combining them—and letting each one play to its strengths—will create a powerful advantage over companies who use a single approach to pricing. Achieving and measuring commercial excellence

As we see every day, the digitalization of business processes is well underway—the pandemic only accelerated it. But we must remember: the goal is not digitalization itself but the ability to future-proof your business, drive agility, and differentiate yourself from competitors. Successfully prioritizing commercial excellence capabilities means digitalizing the sales and business processes connecting front-office systems like your CRM with back-office ERP systems. To thrive amidst disruption, you must continue to ensure that you are delivering the right product, to the right customer, at the right price, at the right time. Commercial excellence encompasses the set of business processes that reside between your back-office ERP system and your front-office CRM. It is the articulation of what should occur between a sale and order fulfillment and digitalizing that process for optimal efficiency, accuracy, and profitability. Technically, ERP systems can handle quoting and pricing, and CRM systems can provide an exceptional customer experience. However, when it comes to making sure the right product, at the right price, at the right time, is presented during the quoting process, most ERP and CRM solutions come up short. That is where digital middleoffice capabilities that link your front and back office come in. Commercial excellence is an ever-evolving practice that forces companies to continually rethink the way they approach and measure success in the space. One of the first steps in achieving commercial excellence is to understand your organization’s current estimated level of commercial excellence maturity. The more sophisticated your organization can get around commercial excellence dimensions, including customer data, product selection, pricing and optimization, sales and channels, commercial analytics, and process integration, the more value you can deliver and capture in your markets. One of the most critical components of your middle office is pricing. The right price, in real time, placed in the hands of your sales team leads to maximum revenue and an optimal margin. Vendavo has built a commercial excellence maturity model that helps organizations gauge performance across key commercial excellence disciplines and identify where they fall across four stages of progressing maturity: unstructured, disciplined, adaptive, and optimized (Figure 25.3). Your maturity can be considered across six key business-process capabilities that make up your middle office: ■ ■

customer accounts product/Service

C hapter

2 5 

293

294

Capabilities

Figure 25.3 Commercial excellence model.

Process Integration

Commercial Analytics + Insights

Sales/Channels

Pricing + Optimization

Product + Service Selection

Customer + Account Data

Parent/child relationships unclear

Accounts are disorganized

UNSTRUCTURED

UNSTRUCTURED

Categorization based on basic attributes

Account management is clear and consistent

DISCIPLINED

DISCIPLINED

ADAPTIVE

ADAPTIVE

Classified/segmente d into groups with considerations of multiple attributes

STAGES

Segmented into microsegments beyond adaptive strategy

OPTIMIZED

OPTIMIZED

M itchell D. L ee a n d D arius F ekete

C hapter

2 5

D igital T ra n sf o rmati o n ■ ■ ■ ■

pricing sales/channel analytics capabilities and insights process integration

This model has been used to facilitate more than 650 self-assessments of commercial maturity, through value consulting engagements with organizations, and with academic partners through the Vendavo Center for Commercial Research. As you might guess, organizational maturity for commercial capabilities varies widely, both within and between companies. By answering a concise list of questions for each of the business capabilities, you can quickly determine your maturity in that area and in overall commercial excellence. With your organization’s maturity status in hand, you can then compare your overall status with other B2B organizations and frame your next steps for improvement. You can measure your own commercial excellence maturity by visiting www​.vendavo​.com ​/ commercial​-excellence​-maturity​-assessment. In addition to knowing your own maturity ranking, it is also helpful to know how you stack up against other B2B organizations. As you can see from the averages across B2B industries in Figure 25.4, scores cluster in stage 2, or the disciplined stage. That means that almost every organization has an opportunity for growth as the surrounding market landscape continues to evolve. Finally, it is important to understand that maturity levels across these capabilities will vary within large organizations, from business unit to business unit, across regions and geographies, and even between various levels of organizational structure. Gathering input from a wide range of representative stakeholders on a diagnostic assessment allows for a clear discussion of performance gaps and thus agreement for prioritized action. In many cases, commercial excellence emerges from areas that have not been prioritized in the past. A practical example: World-leading water technology company Xylem enables value-based pricing for 1 million product configurations

To improve profitability, Xylem, a global leader in water technology, knew they needed to move away from a simple cost-plus approach and price their products by the value perceived. With more than 1 million product configurations across more than 150 countries, however, that seemed a stretch goal. To price products that range from available, common components to specialties available only from the company based on customer perceived value across all markets, this leading innovator implemented a pricing solution that enabled value-based pricing but also global alignment across regions, product lines, and business units as well as timely, accurate generation of quote pricing for new configurations. The company chose to launch the project within its spare parts business. They began by dividing their spare parts into three segments: 1. key parts designed by the company 2. general parts designed by the company but less specialized 3. standard parts C hapter

2 5 

2 95

296

2.18

2.15

2.05

1.90

1.63

Product / Service

Pricing

Sales / Channel

Analytics

Process Integration DISCIPLINED

ADAPTIVE

Commercial Excellence Maturity

OPTIMIZED

Caution: your level of Process Integration, if low, can act as a limit to value realization even if you have strong capabilities in other business process categories.

UNSTRUCTURED

Figure 25.4 Commercial excellence maturity: B2B averages.

VENDAVO COMMERCIAL EXCELLENCE MATURITY ASSESSMENT

2.24

B2B Averages

Customer

Capability Category

M itchell D. L ee a n d D arius F ekete

C hapter

2 5

D igital T ra n sf o rmati o n

Key parts are the most important components to price correctly for profitability. To ensure accuracy, market surveys were conducted with customers and internal employees to develop perceived value for key parts across eight countries. Three levels of key price drivers were identified and used to develop using AI/ML pricing techniques. Customer perceived value was considered along with product attributes such as weight, diameter, and power. The repair threshold for each piece of equipment was also included for pricing evaluation. Products that are economical to repair instead of being replaced are reflected in the customer’s options in pricing. These and other insights were used to build a framework to support new parts pricing, with easy price maintenance and revisions—all tied to the company’s value-based and market-driven pricing strategies. ‘I receive a lot of questions about why pricing is the way it is, and now I can quickly look up the product and explain the logic of the pricing. Sales now agrees and can defend that price,’ says Niklas Lindstrom, pricing manager at Xylem. Based on the value Xylem was able to generate within spare parts pricing, the company has chosen to support pricing for the company’s portfolio of 300 base products, which in turn has millions of product configurations. The challenge was to price a configuration for quoting, without having to create an identity/part number for each variation—especially if it were not actually sold. This would have generated millions of unused records—unnecessarily burdening the resource planning systems. Pricing methodology was built by pricing the components, as shown in Figure 25.5. These calculations can now be quickly calculated to enable the sales team to quickly build bespoke quotes for each of their customers’ requests. This delivers impact across three categories (Table 25.1).

Base Unit Price

Price

Complexity Factor

Added Components

Figure 25.5 A practical example of value-based pricing.

Table 25.1  T hree

pricing categories

Pricing quality ■

■

■

Pricing efficiency

Market pricing instead of cost based Logical, aligned pricing throughout the portfolio Elimination of manual quoting and pricing errors

C hapter

2 5 

■

■

Value as the pricing driver, instead of one-off, uncoordinated product pricing Prices that are easy to explain with appropriate context

Single source ■

ERP, configuration, and quotation tools all use pricing from one source

297

M itchell D. L ee a n d D arius F ekete Not a project, a transformation

Pricing improvement and commercial excellence are not one-and-done or set-it-andforget-it projects. They require an eye toward sustained operation and delivery of improvements at a reasonable level of effort. It is easy to generate a snapshot set of target prices or to generate a price-increase file for products or customers. It is another question to think about a sustaining capability to do these things. But these are exactly the considerations that need to be top of the mind as businesses engage in hiring data scientists or consultants to work on pricing. Today’s economy is ripe with disruptive forces like supply chain shortages, labor shortages, rising wages, material shortages, M&A activity, volatile changes in input costs, pandemic demand winners and losers, and economic sanctions. Unfortunately, many businesses are being caught flat-footed in their ability to simply execute the changes they desire to implement. Pricing professionals are certainly putting in overtime as businesses urgently invest in efforts to react to these disruptions to keep up with cost or market price changes. Businesses need to look hard at the current state of the world and plan for capabilities beyond a one-time reaction. Today’s uncertainties only reinforce that a business makes money—or does not—based on its standing capability to change course whenever it needs to adjust. In other words, to be successful you must be agile. Investing in a sophisticated enterprise capability for continuing price execution or pricing guidance may be a level above efforts to execute a project. Yet, organizations who think about proper B2B pricing improvement capabilities and then expend the effort to ensure their implementation are highly likely to exceed five to 10 times the value compared with one-time consulting-driven or project-focused events. Successful pricing transformation is not a project; it is an ongoing process relying on proven people and process expertise, and purpose-built, enterprise-ready capabilities. A practical example: Molex’s commitment to commercial excellence

Molex, a multibillion-dollar global supplier of advanced electronic solutions, has firsthand experience executing an ambitious commercial excellence plan. Molex’s customer excellence journey began when it transitioned from a regional-based organization structure to a product-based structure. The restructuring uncovered several discrepancies within their pricing models, including rampant price erosion. Molex had a lack of true visibility in their pricing to end customers and ensuring that they were getting the right prices to the right customers was a complex challenge. Molex made a fundamental organizational restructure and invested in a pricing solution. In less than one year, they saw a total return on their investment. Molex has been committed to commercial excellence ever since and has built out an entire pricing organization to ensure its continued success.

So, who owns pricing?

Because of its centralized purpose in your organization’s success, pricing does not always have an obvious owner. The pricing function sometimes sits with sales, sometimes with finance, and sometimes with marketing. This persistent ambiguity of the pricing function offers an opportunity to demonstrate leadership, champion change, 298

C hapter

2 5

D igital T ra n sf o rmati o n

and be the glue that connects your organization together—after all, pricing is a team sport that exists at the intersection of all your commercial processes. So how should this work be delegated to your team? There are a couple of frameworks for delegation. The first, vertical delegation, is characterized by the following: ■ ■ ■ ■ ■ ■ ■

the amount of decentralization and authority given to the salesforce higher levels of authority often require more resources with objectively less gain lower levels of authority require more central support, which can be more efficient horizontal delegation, by contrast, follows these attributes the amount of strategic authority across functional groups in business units sales has the highest influence, followed by marketing, finance, and R&D finance is often considered the greatest roadblock, followed by accounting, sales, and production

Ownership of Pricing Process

Marketing scholars report that there is an inverted U-shape to vertical pricing delegation with sales; pricing performance was positively related to the degree of horizontal strategic pricing delegation and there was a positive interaction between the two. If pricing is a team sport, adding a modern CPQ empowers your salesforce to win. It will enable speed and efficiency through better quoting; pricing accuracy, and power; as well as higher professional output. Beyond weathering the storm of volatility, pricers and the pricing function can help the organization and drive bottom-line value. When processes are established and accountability is assigned, pricing moves from a finger-pointing exercise toward a meaningful, measurable motion of true digital transformation. Pricers should seize the opportunity to get the pricing function a seat at the leadership table. Figure 25.6 illustrates the extent to which pricing can become a strategic partner in your business. The more you own the pricing process and related pricing decisions, the closer you get to becoming a true commercial partner in your organization. Meanwhile, if your ownership of the process and related decisions is low, your potential is limited to an expert resource at best.

Functional Coordinator

Commercial Partner

Figurehead

Expert Resource

Ownership of Pricing Decisions

Figure 25.6 Process versus decision: Is pricing a partner in your business? C hapter

2 5 

299

M itchell D. L ee a n d D arius F ekete A practical example: How a $150 billion+ integrated health care services company reimagined its strategic outlook toward pricing

A $150 billion+ global integrated health care services and products company that provides clinically proven medical products and pharmaceuticals needed to reimagine its strategic outlook toward pricing. With over 50,000 employees, and over 100,000 products and offerings, the company was struggling with contract management, channel optimization, and product pricing analysis. To reimagine its strategic outlook toward pricing, the health care services company identified a few goals: ■ ■ ■

gain executive engagement and a commitment to pricing get pricing seated at the center of the organization, instead of as an afterthought encourage conversations and drive pricing changes based on recommendations

After they aligned their team and a modern pricing solution was introduced, the company achieved 33% more than their expected return. That translates to $30 million in margin over 12 months. What is important to note here, is that this organization did not wait until the system was ‘perfect.’ (Indeed, many companies, especially at the onset of their pricing journeys, suffer from analysis paralysis.) Instead, they began looking for data on day one—and even with imperfect data, they were able to get their entire team rowing in the same direction while improving and capturing value. Conclusion

Recent digital transformation trends have accelerated the need for process innovation, strategic business prioritization, and interdisciplinary collaboration in B2B organizations. Pricing has evolved significantly in the last two or three decades. Initially regarded as a marketing subfunction, or even an afterthought, today, pricing is recognized as a critical component of value creation. To unlock the potential of digital transformation, organizations must take a comprehensive approach to commercial business processes, aligning experts and best practices with purpose-built, enterprise-ready, technology capabilities to achieve commercial excellence. Part of the comprehensive approach is a pragmatic recognition of the need for quick wins to fuel the enthusiasm for the required changes—both in improvements in processes and in measured improvements in top- and bottom-line benefits. For this reason, a phased approach that addresses the fundamentals first—such as appropriate and consistent analytics—is key to long-term improvements (Figure 25.7). With early successes to point at, there is genuine engagement for what comes next on the commercial excellence journey—control and management of list price setting, negotiation guidelines, and approval processes for all sales channels, and developing the capability to understand the willingness to pay at very granular levels. Introducing and constantly referencing the concept of ‘crawl, walk, and then run’ is useful to set expectations and gain long-term commitment to the effort needed to transform—such that the new motions become the standard motion. The deliberative and ongoing evaluation of your commercial excellence maturity will empower you to continuously iterate and rethink your commercial strategies. Remember, digital transformation is not a project but an ongoing strategic business initiative. 300

C hapter

2 5

C hapter

2 5 

Segmentation & Optimization

Analytics / General

3

4

Optimized Target Prices

Automated Optimized Workflow

Leverage Peer Comparisons

Price Increase Effectiveness

Improved Deal Cycle Time

Improved Invoice Accuracy

Enforce Customer Compliance

Figure 25.7 Typical benefits as a function of capability improvement.

Channel Satisfaction, Avoid Rebate Penalties

Rebate Overpayment Prevention

Recover Value-Added Services

Eliminated Negative Margin Transactions

Improve Customer/Product Mix

Minimize Unwarranted Price Variation

Deal Management: Efficiency Deal Management: Effectiveness

2

Improve Sales Compliance

Mass Price Updates

Value-Based List Price Setting

Ensure Appropriate Price Premiums

Efficient List Price Setting & Publishing

Price Management

1

0%

1%

2%

3%

Benefits as a % of Revenue 4%

5%

D igital T ra n sf o rmati o n

301

M itchell D. L ee a n d D arius F ekete Bios

Mitch is VP Product Marketing and Profit Evangelist at Vendavo with more than three decades of experience in improving technical, operational, marketing, and commercial processes. Prior to Vendavo, Mitch was with BASF and Orica in product marketing and business management, driving operational optimization, pricing excellence, and margin improvement. Darius is Managing Consultant for Vendavo’s Value Consulting team, with more than a decade of professional experience in pricing and digital transformations. Before joining Vendavo, he delivered several complex business-transformation initiatives in B2B manufacturing, distribution, and financial services. Darius also worked at SimonKucher & Partners, advising clients on top-line growth, price optimization, and margin improvement initiatives.

Key objectives 1. Demonstrate the actual meaning and value of commercial digital transformation, beyond a buzzword. 2. Provide insights and best practices for leveraging digital commerce, self-service capabilities, data and AI, effective selling practices, and price optimization to unlock commercial excellence. 3. Illustrate how to use digital transformation initiatives to get pricing a seat at the executive table and champion pricing and selling change in your organization.

Key summary points 1. Successful digital transformation goes far beyond deploying technology for basic data collection or simple historical reporting. Though often poorly understood, a critical way to unlock the value of digital transformation is through a holistic application to your commercial business process—aligning your experts and best practices with purposebuilt, enterprise-ready technology capabilities to achieve commercial excellence. 2. Commercial excellence is an ever-evolving practice integrating capabilities across your pricing, selling, products, channels, and customer knowledge. The deliberate evaluation of the maturity of your capabilities, and your level of integration, forces you to continuously iterate and rethink your commercial strategies—it is not a project but an ongoing strategic business initiative. 3. The pricing function exists at the intersection of all the commercial processes. That intersection means that pricing sometimes sits with sales, finance, marketing, or even, in some cases, with no ownership at all. The persistent ambiguity offers an opportunity to demonstrate leadership—for those that choose to pick up the challenge.

302

C hapter

2 5

D igital T ra n sf o rmati o n

Key questions 1. What does commercial excellence encompass, and how do you measure it in your B2B organization? 2. How have B2B organizations driven success and business value through digital transformation? 3. In a B2B organization, what is the goal of the pricing function, where does ownership sit, and how do you make it a strategic priority with continuous improvement?

References BusinessWire. (2019, April 10). Vendavo introduces new artificial intelligence deal price guidance solution. https://www​.businesswire​.com​/news​/ home​/20190410005193​/en​/ Vendavo​-Introduces​-New​-Artificial​-Intelligence​-Deal​-Price​- Guidance​- Solution. Gartner. (n.d.-a). Digitalization. https://www​.gartner​.com​/en​/information​-technology​/glossary​ /digitalization. Gartner. (n.d.-b). Digitization. https://www​.gartner​.com​/en​/information​-technology​/glossary​/ digitization. Herdina, M. (2020, September 25). Augmented reality disappeared from Gartner’s Hype cycle— What’s next? AR Post. https://arpost​.co​/2020​/09​/25​/augmented​-reality​-gartners​-hype​- cycle/. McKinsey & Company. (n.d.). The quickening. McKinsey Quarterly. https://www​.mckinsey​.com​ /business​-functions​/strategy​-and​- corporate​-finance​/our​-insights​/five​-fifty​-the​-quickening. Menear, H. (2020, May 17). The history of digital transformation. Technology. https:// technologymagazine​.com ​/data​-and​-data​-analytics​/ history​-digital​-transformation. Sana. (2022). 2022  B2B buyer report. https://www​.sana​- commerce​.com ​/ b2b​-buyer​-report/.

C hapter

2 5 

303

Index

Page numbers in italics represent figures while page numbers in bold indicate tables. A/B testing 139, 154–55, 241 ‘A Rake Too Far’ (Gurley) 225 ABC Ltd example 109, 110–11, 112, 113–14 Accenture 187 accounting 174–75 advertising 226 advertising revenue 209 agility 282 Agriniser 206 Airbnb 215, 217–18, 225, 229 airline industry 20 airline sales 237 Albert Heijn supermarket 29 Allegro 209–10 AllState 102 Alsense Cloud Services 130 Amazon​.c​om 101, 226, 227; Amazon Marketplace 226; and dynamic pricing 28, 235, 236 analysis paralysis 300 analytics-based differentiators 53 Andreessen Horowitz 22, 72 Andreessen, Marc 230 annual recurring revenue (ARR) models 39–40, 78–79 Apple 229 artificial intelligence (AI) 1, 187, 290; and analytics-based differentiators 53; and capitalization 191; and chess 290; choice modeling method 276–80; and data cleaning 189, 192, 291; and digital transformations 

6; and dynamic pricing 31, 154, 253; and economic value analyses 89; economics of 187; governance of 289; and human intelligence 290–91, 292, 293; machine learning 31, 155, 190, 198–99, 252–54, 279; narrowing initiatives with 260–61; pricing revolution 251, 259–61; and seller roles 277; user interfaces 193; and value-based selling 275–76, 278, 280, 281, 282–83; and Web 3.0 218, see also enterprise AI automated pricing 252 average revenue per user (ARPU) 66, 68 B2B Buying Study (2021) 275 B2B industries 39 B2B selling landscape, changes in 267–72 Baculum 206 Bain survey (2020) 251 balanced digital innovation ecosystems 282, 283 Balfour, Brian 230 Bassoulet, Arnaud 259–63 BATNAs (best alternatives to negotiated agreements) 225, 228 beer trial 101, 102, 103 bex 212 bias: and price changes 237; in price experiments 242; selection bias 243; success bias 244 big data 176–77, 178 billing analytics 220

305

I n dex black-box pricing solutions 291, 292 blockchain 218 Blue Ocean Strategy 275 Bojinov, I. 241 Bombora 282 Bonoma, T. V. 161 Booking​.c​om 206 Boston Consulting Group (BCG) 252 Boxmotions 209 Brainly 208–09 Brand24 65–67; discounted accounts with 67–68; internal data diagnostics 66 Brégé, Camille 259–63 ‘Bringing Order to Discounts Gone Haywire’ (Bain & Company) 252 business billing departments 215 business initiatives 182 business model-based differentiators 53–54 business stakeholders 181 buyers, and COVID-19 288 buying firms 161–62 California Consumer Privacy Act (CCPA) 33 Campbell, D. T. 236 Campbell, Patrick 227 capability improvements 300, 301 capitalization 191–93 CGTrader 206 change 276; continuous 176–77 chess 290 Choco 212 choice modeling method 276–80 churn 23, 40–41, 65–68, 254 Cicero, Simone 6, 223–33 Classen, M. 276 ClickClickDrive 207 club sports example 242, 243 Coca-Cola Company 32 Coffey, Brad 22 Cohosting​.​io 207 Colias, John V. 6, 275–84 Collection Hub 206 Columbus, Louis 6, 251–55 commercial excellence 293, 294; maturity 293, 295, 296, 300 commissions: and marketplaces 205–07, 209, see also take rates company valuation 45 competitive intelligence 57, 59, 135; importance of 49 competitors 51; customer internal solutions as 51; direct 50, 51; discussing 50, 51; ‘do nothing’ options 52; indirect 50, 51; and pricing 79, 135; and transparency 24; underestimating 50; understanding 49 concurrent cannibalization 241, 243–44 conjoint analysis 140–42, 143–45, 146 consumer data collection 28, 33–34

306

consumer packaged goods (CPG) companies 260 consumer-packaged goods (CPQ) systems 253–54, 289, 299 Cook, T. D. 236 cookies 28–29 cost-plus pricing 11, 21, 24, 79, 97, 126, 135, 295; and manufacturing 119; moving away from 97 cost to acquire customers (CAC) 21–22 costs: of data 175, 177, 188–89, 198; of enterprise AI 188–90, 192; of infrastructure 190 Covey, Stephen 180 COVID-19 pandemic 267, 288, 293 Crella 208 CRM (Customer Relationship Support) 293; compared with CVM 45; and CSMs 40; evolution of 43–44 CSMs (Customer Success Managers), and CRM 40 Currier, James 223 customer acquisition cost (CAC) 106, 229–30 customer benefits, and customer value modeling 57, 58 customer-centricity 23–24, 40; and CRM 44; pricing techniques 270, 272; of value-based pricing 100 customer demographics 228 customer experience management 77–78 customer intimacy 123, 124, 125 customer lifetime value (CLV) 22, 39–40, 87, 98, 100, 103, 106, 226, 229–30, 232, 253–54 customer-randomized experiments 240 customer retention 39–41 customer success management 78 customer success teams: and CVM 42; and discovery 42–43 customer value, quantifying 76 customer value management (CVM) 12–13, 39–41; compared with CRM 45; evolution of 44; strategies for 42 customer value modeling 57, 58 customers: approaches to 279; changing dynamics of 40; collected data about 28, 33–34; concerns with value pricing 13–14; defining 106, 107; and dynamic pricing 28–30, 32, 32, 34; perceptions of value 77, 97, 126, 278; suppliers as 229; of X-as-aService businesses 121, 123 Cuvama 4 CVM platforms 13; and Verint 15 Danfoss 130 data 53, 173, 251, 253, 289; capture of 163–64, 178, 197, 276; cleaning 189, 192; 

I n dex



as cost 175, 177, 188–89, 198; external 154; generating 57–58; internal 66, 154; mindsets about 197; monetization of 123; and oil 173, 174, 179, 196; and PSM 136– 37; requrements for the choice modeling method 276–77; second use 198–99; transformation of 163; value of 174, 195, 200, see also big data; value-driven data data analytics 180; archtypes of 163, 164, 165–67 data-based differentiators 53 data industry 200 data lake metaphor 179, 180 data marketplaces 195, 199–200; internal 195; and privacy/security concerns 195–98 data marketplaces; transaction engine, see also marketplaces data monetization 173, 175, 176, 196, 210; predictions and questions 182–84; strategies 177–78; vs. acquisition 177 data pricing 197–201 data products 121, 122, 129, 153 data science 192 data science value engineering 180, 181 data scientists 190 data value assessment 183 data value chains 49 data warehouses 66, 199–200, see also data marketplaces Dataiku, 2019 Maturity survey 189 decision fatigue 103 decision-making 183–84; and value engineering 180, 181–82​ decision support experiments 238, 239 decoy effect 101, 102 delegation frameworks 299 Deloitte Digital 51 Dholakia, Utpal 25 difference-in-differences (DiD) 242–43, 244 digital add-ons 121, 122 digital commerce 288 digital negotiations 289, 290 digital pricing framework 84, 86; enablement 84, 85, 86; execution 84, 85, 86; offerdesign 84, 85, 86, 87–88, 89, 90, 91, 92–93, 94, 288–89; pricing ecosystem (PECO) 84, 85, 86; subprocesses 85, 86 digital pricing models 155 digital pricing transformations 156, 260–61; and infrastructure 154; and integration 154–55; successful implementations of 153; and user interfaces 154 digital products 121, 122, 129–30 digital technologies 287 digital transformation frameworks 163, 164 digital transformation projects 1, 51, 152–53 digital transformations 152, 252, 275, 287–88, 300; and pricing optimization

151; projects vs. products 152–53; and value data 15 digitalization 119 digitization 287 discounting 265; and price setting 74, 79; sales team use of 14–15; and Verint 15 disruptions 298 ‘do nothing’ options 52, 58 Dobry Mechanik 208 dollarization 58–59; challenges with 56–57; and customer value modeling 57, 58 dollarization techniques 54, 55 Dollarizing Differentiation Value (Liozu) 52 Dot Residential 209 Drucker, Peter 108 dynamic ecosystems 195 dynamic pricing 27, 32, 33–36, 198, 212, 231; algorithmic 154; and Amazon​.c​om 28, 235, 236; approaches for 30–32; and internal incentive schemes 34–35; omnichannel 35; personalized 28–30, 32, 33–34, 240–41; personalized coupons 29, 31; and price experiments 237; process 29, 30; strategies for 30; time-based 28, 31, 32, 33 e-commerce 288, see also marketplaces EBITDA 259 ‘Economic Digital Asset Valuation Theorem’ 176 economic value analyses 89, 90 economics experiments 197 enterprise AI 188; costs of 188–90, 192; and ROI 188 Epic Games 229 ethics, and transparency 19–20 Evans, Ben 226 EVE 55, 56 Eversports 207–08 Expedia 206 expediency 268, 269–70, 271 experiment-specific metrics 239 exploratory experiments 238, 239 extended reality (XR) 289 extensions 223, 224, 228–29 Fachmistrz 208 Fekete, Darius 6, 287–303 finance teams 152, 215 financial analyses 94–95 fintech 210–12 fixed subscription fees 129–30 Fixly 209 Forrester 275, 282 Fotolia 206 Fournier, Alex 5, 187–94 frameworks 83–84, see also digital pricing framework

307

I n dex freemium software designs 14, 23, 112, 227 Freidesk 207 frictionless buying experiences 267, 269, 290 Friedli, T. 276 Gartner 270, 275, 287, 289 GastroJob 209 Gates, Bill 230 GE Aviation 188 General Data Protection Regulation (GDPR) 33–34 Gilmore, J. H. 77–78 globalization 176–77 good/better/best pricing models 14, 23, 65–66, 91, 93, 97, 101, 103, 104–05, 112, 113–14, 116, 129–30, 227; 3D P&P framework 105–06, 115; beer trial 102, 103; deploying 114–15; and metrics 109, 110–11; and optional add-ons 91, 111, 112, 113; and transparency 23–24 GoPro 101 Gratka 209 Gubian, Claire 5, 187–94 Gurbaxani, Mrinal (MG) 39–47 Gurley, Bill 225, 230 Hall, J. 237 Hayes, R. J. 244 Hazan, Joël 259–63 headline metrics 239–40 Heidelberg subscription 130 Heidelberger Druckmaschinen 130–31 Hinterhuber, Andreas 1–7, 71–80, 276 Hockenmaier, Dan 230 Holst, Lennard 119–32 HubSpot 79 IDC, on digital/technological investments 1 Inc. Magazine 24 indifference price point 107–08, 111, 137 industrial smart services 159, 160; and performance management 162; vs. product-centric businesses 160 infrastructure: costs of 190; and digital pricing transformations 154 intangible assets 174, 185n1, see also data integration, and digital pricing transformations 154 internal data diagnostics 66 intertemporal cannibalization 241 IoT 218–19 IoT architecture 184 Johnston, Matt 5, 135–48 Kakkanatt, Chris 189 Karrer, B. 244 Kasparov, Gary 290

308

Katoo 212 Kendrick, C. 237 key-value items (product heroes) 33 KeyBanc 75 KPIs 39–40, 42, 45, 56–57, 66, 78–79; and AI 89, 253, see also specific metrics Krone Trailer GmbH 129 LaunchWorks 230 learning engine 223, 224, 228 Lee, Mitchell D. 6, 287–303 Leiting, Tobias 119–32 Levine, D. M. 241 lifetime value (LTV). see customer lifetime value (CLV) Liki24 209 Lindstrom, Niklas 297 LinkedIn 101 Liozu, Stephan 1–7, 49–60 listing fees 209 location-based pricing 29–31 Lu, A. 241 Łubiński, Jacek 6, 205–13 Lunching 206 luxury brands 33 machine learning models 190 Main, Kelly 24 managed service providers (MSPs) 73 ‘Managing Price, Gaining Profit’ 265 Mansard, Michael 5, 97–118 manufacturing: and data-intensive application 197; and digitalization 119, see also Original equipment manufacturers (OEMs) ‘The Marc Andreessen and Bill Gurley Schools of Pricing’ (Hockenmaier) 230 market reach, expanding 103 market research 139–40, 278 market segmentation 160, 161 market segments. see target audiences marketing 72; and pricing 71 marketing/product teams 276; and CVM 42 marketplaces 205, 211, 212, 216–17; and add-ons 208–09; and advertising revenue 209; asset management fees 210; and commissions 205–07, 209, 224–25; competition in 225; data monetization fees 210; and digital negotiation 289–90; future 218–20; and listing fees 209; pay per lead fees 209; payment fees 207–08; platform-marketplaces 223, 226–30; premium listings/promotion fees 209–10; setting prices 225–26; tasks of 224–25, see also data marketplaces; transaction engine markets 159 metaverse 7 

I n dex metrics 76, 91, 92, 93; and price experimentation 239–40, 242; for pricing 77, 126, 127, 129; and SaaS 108–09; subscription-specific 106, 120; for value determination 123; value metrics 108–09, 227–28, see also specific metrics Microsoft Excel 259 Miller, Scott 5, 83–96 Millman, Debbie 20 MIT-BHI survey 259–60, 262 mixed logit models 278–80 Molex 298 monthly recurring revenue (MRR) 66, 68 Moulton, L. H. 244 Müller, G. 77 MVPs 56–59 Nagle, T. 77 national brands 33 net revenue retention (NRR) 44 Netflix 101, 226 The New Invisible Hand: Five Revolutions in the Digital Economy (Westra) 19 Nosko, C. 237 offer negotiations 288–89 oil prices 173 OLX 209–10 one-time transactions 125, 126, 128, 129, 216, see also perpetual licenses operational variables 163–64, 167–68 operationalization, processes of 189–90 ‘optimization decision tools’ 281, 282 Orbitz Worldwide Inc. 29 Original equipment manufacturers (OEMs) 159–60, 164, 165–67, 170; pilot project 167, 168, 169; and target audiences 160– 62, see also industrial smart services OtoMoto 209



Parks, Stella 6, 275–84 pay per lead fees 209 payment intervals 127 Pei, Jian 6, 195–202 performance management 162 perpetual licenses 5, 65, 74–75, see also one-time transactions personalized coupons 29, 31 Phillips, R. L. 240 Phillips, Robert 6, 235–47 Pine, B. J. 77–78 pocket prices 265, 266 Porter, John 4, 11–17 Poundstone, William 102 Prato, Luis 5, 159–71 predictions, vs. questions 182–84 predictive maintenance 52, 58 price, and profitability 265

price elasticity 146–47, 252–53 price experimentation 235–36; evaluating results 242–43, 244; and metrics 239–40, 242; natural 237; planning 238, 244–45; running 242, 245; types of 237, 238, 239– 41, see also specific types of experiments price fencing 231 price lists 93, 94 price management 63, 272, 301; and discount policies 66–67, 74, 265 price realization 65; and value creation 63, 64 price-response curves 236–37 price steering 29 price-value trade-off analysis 89, 90 price waterfall framework 265–66 pricing 71, 79, 100, 223, 224, 226, 230, 259, 298, 299, 300; A/B testing 155; algorithmic 154, 269, 270; analytics 266–67; automated 252, 291; complexity of 83; data 197–98, 200–201; and data marketplaces 195; and discount policies 66, 74, 265; framework for smart-productservice offerings 120, 121; indifference price point 107–08, 111, 137; key contexts for 223; as management problem 266–67; natural variations in 236–37; online vs. offline 235; platform pricing 231–32; pressures on 65–66; price getting 71–72, 74; price setting 71, 73; scatter plots 266; statistics 2; and take rates 225–26; unit economics 229–30; and WTP 63, see also discounting; dynamic pricing; WTP (willingness to pay) pricing-based differentiators 53–54 pricing errors 253 ‘Pricing for Platform Powered Businesses’ 230 pricing initiatives 68; failures of 151 pricing models 14, 23, 24, 65–66, 74–75, 79, 91, 93–94, 101, 231–32; customeroriented 125; and digital pricing innovation 154; one-time transactions 5, 65, 74–75, 125, 126, 128, 129; resultsoriented 125, 126, 127; for smartproduct-service offerings 125, 126, 127; success-based 126, 127; usage-based 75, 76, 77, 79, 125, 126, see also good/better/ best pricing models pricing policies 236 pricing software 2 pricing teams 152, 260, 299 pricing transformations 260–61 prioritization matrixes 181, 182 privacy/security concerns, and data marketplaces 195–98 Private-label brands 33 product-centric models 98, 99, 125, 160 ‘Product Channel Fit Will Make or Break Your Growth Strategy’ (Balfour) 230

309

I n dex product channels 229–30 product development: conjoint analysis 140–42, 143–45, 146; and value drivers 12 product packages 90, 91, 101; industryspecific 103; mistakes with 100; recurring revenue models 100, 101; tiering 93, 103, 105, 110–11, 112, 115, see also good/ better/best pricing models product-randomized experiments 240–41, 243 profit, and dynamic pricing 27 Profitwell 229 propensity models 252 purchasing variables 163–64, 167–68 PwC 218 ‘quantified buyer personas’ (QBPs) 229, 232 Rajala, R. 276 Rekki 212 renewals, and upselling 15 reuse 191–93 Rix, Calvin 119–32 Roberge, Mark 79 ROI (Return on Investment) 68; and AI 188 Rynek Pierwotny 209 SaaS (Software as a Solution) 40, 99, 227; and cost-plus pricing 21; customers’ value to 68; and feature/function competition 11; growth of 97–98; and metrics 108–09; success 39; and transparency 19, 21–22, 24; and value-based pricing 21, 72, see also specific companies; subscription business model; subscription businesses; subscription revenue; X-as-a-Service businesses sales, self-selecting 22–23 sales cycles 290; and value conversations 14 sales teams: and automated pricing 291–92; and CVM 42; and digital negotiations 289, 290; and discovery 43–44; shifting jobs of 288; and value pricing 14 Salesforce 39, 65, 217–18, 229, 276 Saravu, Murali 6, 215–22 Schmarzo, Bill 5, 173–85 Schrank, Regina 119–32 Schüttflix 212 second use data 198–99 secondary metrics 239 self-service 269–70, 271, 288 service revenue 159 The Seven Habits of Highly Effective People (Covey) 180 Shadish, W. R. 236 Shannon, Claude 287 Shapiro, B. P. 161 sharing economy 216–18

310 

Shutterstock 206, 208 Shypple 209 Simchi-Levi, D. 241 Simon, Hermann 63 Simplo 212 Sinofsky, Steven 72 SKU companies 39 smart-product-service offerings 119, 131; alternative framework for 163, 164; designing 121, 123; monetary improvements through 127; pilot project 167, 168, 169; price models 125, 126, 127; pricing framework 120, 121, 131; and value determination 123, 124, 125; vs. product-centric businesses 160 smart products 121, 122, 129 Smith, Alex 39–47 social media monitoring 65 Sonar Home 207 Sonpar, Gaurav 5, 97–118 Southwest Airlines 20, 24 Spirit Airlines 20 Spot a Wheel 207, 209 Stableton Financial 210 Staples​.c​om 29 Stephan, D. F. 241 stockpiling 241 strategies: and data 177; deploying 114–15; and the digital pricing framework 84, 86, 87; diversifying 103; monetization strategies/tactics 87, 88; for platform pricing 231–32, see also pricing models subscription business model 39, 41; and customer integration 119–20; growth of 97–98 subscription businesses 41, see also SaaS (Software as a Solution) subscription fees 208, 215; fixed 127, 128, 129–30, 226; variable 127, 128, 129 subscription revenue 23; and manufacturing businesses 119 Super Prawo Jazdy 209 switchback experiments 241 Szabat, K. A. 241 take rates 145, 205–07, 223–26, 229, 231–32, see also commissions target audiences 143, 145, 160, 169, 279; defining 106, 107, 167; and discounts 252 taxes, and Web 3.0 219 TBO (Total Benefit of Ownership) 54, 55 TCO (total cost of ownership) 54, 55 technological evolution 1, 83; and AI 190 technology, vs. information 197 Telematics Data API 129 tools, monetization of 207 

I n dex total addressable market (TAM) 103 total cost of ownership (TCO) 162 Toytari, P. 276 transaction databases 265 transaction engine 223, 224, 227–28 transfarency 20 transparency 289; in automated pricing 291, 292; and CAC 22; and competitors 24; and customer self-selection 23; and ethics 19–20; internal 191–92; lack of 24; and pricing 19–20; and SaaS 19, 21–22, 24–25; and value pricing 21 truth-seeking experiments 238, 239 ‘$21 Trillion in Intangible Assets Is 84% of S&P 500 Value’ (Berman) 174 two-part tariffs 75 two-sided transactions 219–20 Uber 212, 226, 237 Ulevitch, David 22 upselling 14, 41, 78, 88, 252; renewal periods 15 usage-based pricing 75, 76, 77, 79, 125, 217, 220, 227, 231; implementing 77–78 use cases 183, 188; and data cleaning 189; operationalizing 153–54, 187 user interfaces, and digital pricing transformations 154



value 22, 44, 97; customer perceptions of 77, 97, 126, 278; from data 174, 195, 200; data determining 15, 183; determining 13, 123, 124, 125, 129; quantification of 120, 276; technological solutions for 12–13; understanding 12; unseen 63 value-based offers 280, 281, 282, see also value selling value-based pricing 11–12, 79, 100, 135, 198; analyses of 87, 89, 90; and CVM 13; developing 84; and SaaS companies 21, 72; and time 13–14; and transparency 21; and value selling 14–15; and Xylem 297 value communication 77 value constellation differentiators 53–54 value conversations, and sales cycles 14

value creation 63, 120, 122; data as 174–75; opportunities for 152; and price realization 63, 64 value-driven data 178–79 value drivers 12 value engineering 180; framework 179, 181 ‘value’ professionals 43 value propositions 226–27, 228, 276 value selling 15–16, 43; and AI 275–76, 278, 280, 281, 282–83; and value pricing 14–15 Van Westendorp’s price sensitivity meter (PSM) 108, 109, 136, 137, 139; Newton– Miller–Smith extension 138, 139 variable subscription fees 127, 128, 129 Vendavo 293, 295 Verint 15 Versum 207 Verwaerde, Jean-Sébastien 259–63 Viantro 206 virtual businesses 197 Vomberg, Arnd 4, 27–37 VRIO model 52, 53, 54 Wadhwa, Lalit 5, 151–57 Wagestream 208 Wall Street Journal 27 Web 1.0 215–16 Web 2.0 216–18 Web 3.0 215, 218–21 Westra, Kyle T. 4, 19–26 Wilczynski, Maciej 5, 63–69, 135–48 WTP (willingness to pay) 63, 67, 106, 108, 111, 116, 125, 135–36, 139–40, 282; conjoint analysis 140–42, 143–45, 146; and economic value analyses 89 X-as-a-Service businesses 40–41, 43–45, 88, 121, 122, 123, 130; price metrics 127, 129, see also SaaS (Software as a Solution) Xylem 295, 297 Zawada, Craig C. 6, 265–73 Zhao, J. 241 Zoom 103, 111 Zuora 101

311