Holistic Assessment of Process Mining in Indirect Procurement (BestMasters) 9783658414528, 9783658414535, 3658414529

Citation preview

BestMasters

Vaishali Pawar

Holistic Assessment of Process Mining in Indirect Procurement

BestMasters

Mit „BestMasters“ zeichnet Springer die besten Masterarbeiten aus, die an renommierten Hochschulen in Deutschland, Österreich und der Schweiz entstanden sind. Die mit Höchstnote ausgezeichneten Arbeiten wurden durch Gutachter zur Veröffentlichung empfohlen und behandeln aktuelle Themen aus unterschiedlichen Fachgebieten der Naturwissenschaften, Psychologie, Technik und Wirtschaftswissenschaften. Die Reihe wendet sich an Praktiker und Wissenschaftler gleichermaßen und soll insbesondere auch Nachwuchswissenschaftlern Orientierung geben. Springer awards “BestMasters” to the best master’s theses which have been completed at renowned Universities in Germany, Austria, and Switzerland. The studies received highest marks and were recommended for publication by supervisors. They address current issues from various fields of research in natural sciences, psychology, technology, and economics. The series addresses practitioners as well as scientists and, in particular, offers guidance for early stage researchers.

Vaishali Pawar

Holistic Assessment of Process Mining in Indirect Procurement

Vaishali Pawar Augsburg, Germany

ISSN 2625-3577 ISSN 2625-3615 (electronic) BestMasters ISBN 978-3-658-41452-8 ISBN 978-3-658-41453-5 (eBook) https://doi.org/10.1007/978-3-658-41453-5 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer Gabler imprint is published by the registered company Springer Fachmedien Wiesbaden GmbH, part of Springer Nature. The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany

Abstract

In today’s highly competitive world where Digitalization has become need of the hour, companies struggle to carry out digital transformation in most efficient manner. Process Mining is one of the techniques which enable proficient digital transformation and is gaining importance. Companies aspiring to implement Process Mining and profit from its benefits lack a step-by-step guide covering holistic view to judge the feasibility of Process Mining within their organization. Literature studies reflect the approach of creating “business case” to evaluate the value realization for generic project investments. No studies have been found specific to business case in relation to digitalization which implies a broad spectrum of technologies. Neither business case covering holistic view specific to Process Mining could be found within Literature review. The conventional business case approach has been found incapable to support prudent evaluation of digital technologies. This Master thesis focuses on preparing a business case template covering an integrated holistic concept specific to Process Mining, to evaluate its feasibility and consequent application to indirect procurement. This blueprint will enable companies to carry out a viability check of process mining technology within their organization in relation to indirect procurement. To achieve this, knowledge was first acquired through literature study on Process Mining as a technology, Indirect procurement as a business function, Business case for generic projects and business case framework for IT-investment projects. Resulting, a novel revised IT investment business case framework is prepared which could support organizations decide on right IT technology investment amongst alternatives. Further an integrated holistic concept consisting of four aspects (“As-Is” analysis, technical feasibility, Data quality assessment and

v

vi

Abstract

financial case) was specifically developed for Process Mining. The generated integrated holistic concept for assessment of Process Mining can guide organizations evaluate their readiness within indirect procurement business function. The master thesis discovered that it was extremely vital to carry out Process Mining readiness check before taking decision to implement the tool, to enable well-organized decision and avoid financial losses resulting from failed implementation. Subsequent business case template for Process Mining and integrated holistic concept for assessing process mining in relation to indirect procurement will provide companies a step by guide to prepare themselves and explore the limitless potential of process mining as a tool. This could be seen as a steppingstone towards achieving resourceful and effective digital transformation to prepare for the future.

Restriction notice

This thesis entitled “Holistic Assessment of Process Mining In Indirect Procurement”

Disclosure of this thesis, even partly, requires authorization of the author and of Prof. Dr. Erich Groher (International School of Management, München)

vii

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Motivation and Initial Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Aim of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 3 3

2 Process Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Process Mining Market Trend and Providers . . . . . . . . . . . . 2.1.2 Discussion of Related Terms . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.3 Limitations of Data Mining and BPM . . . . . . . . . . . . . . . . . . 2.1.4 Positioning Process Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Event Data and Event Logs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Characteristics and Structure of an Event Log . . . . . . . . . . . 2.3 Process Mining Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Process Discovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Conformance Checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Enhancement / Process Re-engineering . . . . . . . . . . . . . . . . . 2.3.4 Operational Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7 7 9 10 14 15 16 18 21 22 24 25 27

3 Indirect Procurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction to Procurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Direct and Indirect Procurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Direct Procurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Indirect Procurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Direct v/s Indirect procurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Excellence in Indirect Category Management . . . . . . . . . . . . . . . . . .

31 31 32 32 33 35 35

ix

x

Contents

4 Business Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Digitalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Introduction to Business Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 A general Business Case Template . . . . . . . . . . . . . . . . . . . . 4.2.2 Advantages of a Conventional Business Case . . . . . . . . . . . 4.2.3 Limitations of Conventional Business Case for it Investments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Business Case Framework for it Investments . . . . . . . . . . . . . . . . . . 4.4 Revised Business Case Template for it Investment . . . . . . . . . . . . . 4.5 Business Case for Process Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Integrated Holistic Concept to Assess Process Mining in Indirect Procurement (Holistic Model) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 “As-Is” Indirect Procurement Process Analysis . . . . . . . . . . . . . . . . 5.2 Technical Feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Stepwise Guidance for Technical Feasibility Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Data Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 DQD for Process Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Defining the Six Primary DQD . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.1 Completeness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.2 Uniqueness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.3 Timeliness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.4 Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.5 Accuracy/Correctness . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.6 Consistency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Data Quality Assessment Levels . . . . . . . . . . . . . . . . . . . . . . . 5.3.3.1 Activity Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3.2 Timestamps Level . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3.3 Meta-attribute Level . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Financial Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Financial Measurement Method . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Cost of Investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 Gain on Investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.4 Intangible Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

39 39 42 43 44 45 46 51 53 57 58 60 65 72 73 76 76 76 77 77 78 78 79 79 80 81 84 84 85 86 90

Contents

xi

6 Final Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93 93 96 96

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97

Abbreviations

AI API BC BOM BPM BPMN BPR CDC CIS CRISP-DM CRM DAMA DIGIT DM DMTC DQD EDI EPK ERP ETL FI FTE HBR HR IM&S IoT

Artificial Intelligence Application Programming Interface Business Case Bill Of Materials Business Process Management Business Process Management Notations Business Process Reengineering Change Data Capture Computational Intelligence Society Cross Industry Standard Process for Data Mining Customer Relationship Management International Data Analytics Management Association EUROPEAN COMMISSION Directorate-General for Informatics Data Mining Data Mining Technical Committee Data Quality Dimensions Electronic Data Interchange Ereignisgesteuerte Prozesskette Enterprise Resource Planning Extract, Transform and Load Finance Module Full Time Employee Cost Harvard Business Review Human Resource Invest, Material and Services Internet of things

xiii

xiv

IRR IT JDBC JIT ML MRO MRP Non-PO NPV OTIF rate P2P PAIS’s PM PO ROA ROI ROIC TCO TQM VUCA WFM YAWL

Abbreviations

Internal Rate of Return Information Technology Java Database Connectivity Just in time Machine Learning Maintenance, Repair, and Operations Material Requirement Planning Non-Purchase Order Net Present Value On Time In Full Procure-To-Pay Process Aware Information Systems Process Mining Purchase Order Return on Assets Return on Investment Return On Invested Capital Total Cost of Ownership Total Quality Management Volatile, Uncertain, Complex, Ambiguous Workflow Management Yet Another Workflow Language

List of Figures

Figure 1.1 Figure 2.1 Figure 2.2 Figure 2.3 Figure 2.4 Figure 2.5 Figure 2.6 Figure 2.7 Figure 2.8 Figure 2.9 Figure 2.10

Figure 2.11 Figure 2.12 Figure 2.13 Figure 2.14 Figure 2.15

Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traditional Process improvements methods v/s Process Mining Value Proposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of market volume and Process Mining Products PEAK Matrix assessment . . . . . . . . . . . . . . . . . . . . . Global Process Mining software market share, by end user 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Process Mining as the bridge between Data science and Process Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The CRISP-DM Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The BPM life cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Process mining as a combination of data-driven and process-centric approaches . . . . . . . . . . . . . . . . . . . . . . . . Event data flow for Process mining . . . . . . . . . . . . . . . . . . . . . Class diagram of an event log . . . . . . . . . . . . . . . . . . . . . . . . . Positioning of the three main types of process mining: (a) discovery, (b) conformance checking, and (c) enhancement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An example of process discovery output based on event log of a P2P process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Four quality factors for assessing process model quality . . . Connecting event log and model (with extension) . . . . . . . . . Process mining techniques explained in terms of input and output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detecting violations at run-time . . . . . . . . . . . . . . . . . . . . . . . .

4 8 9 10 11 12 13 16 17 18

21 23 24 26 27 28

xv

xvi

List of Figures

Figure 2.16 Figure 2.17 Figure 3.1 Figure 4.1 Figure Figure Figure Figure Figure Figure Figure

4.2 4.3 4.4 4.5 4.6 4.7 5.1

Figure Figure Figure Figure Figure

5.2 5.3 5.4 5.5 5.6

Figure 5.7 Figure 5.8 Figure 5.9 Figure 5.10

Partial trace of a running case and some predictive models, providing a prediction . . . . . . . . . . . . . . . . . . . . . . . . . Model based on historic data, providing recommendations for running cases . . . . . . . . . . . . . . . . . . . . Indirect spend optimization through value-capture best practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Digital technologies, tools, and methods currently used by organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Golden rules for successful digitalization . . . . . . . . . . . . . . . . General/conventional business case template . . . . . . . . . . . . . Business case focusing on 5 aspects by Ward et al . . . . . . . . Business case framework by Maes et al . . . . . . . . . . . . . . . . . Integrated business case framework . . . . . . . . . . . . . . . . . . . . . Business case template for IT investment . . . . . . . . . . . . . . . . Overview of before implementation analysis as integrated holistic concept for process mining . . . . . . . . . Overview of technical feasibility assessment . . . . . . . . . . . . . Typical A-priori model for a Purchase-to-pay process . . . . . Scenario 1 of change activity . . . . . . . . . . . . . . . . . . . . . . . . . . Scenario 2 of change activities . . . . . . . . . . . . . . . . . . . . . . . . Data quality dimensions for raw event data assessment applicable to this thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Three levels for data quality assessment . . . . . . . . . . . . . . . . . Questions answered with the help of meta-attributes . . . . . . Template for cost of investment calculation for process mining implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gain on investment calculation method . . . . . . . . . . . . . . . . .

28 29 37 41 42 44 47 48 49 52 58 61 66 71 71 75 79 82 87 88

List of Tables

Table Table Table Table Table

2.1 2.2 3.1 5.1 5.2

Table 5.3 Table Table Table Table Table Table Table Table Table Table Table

5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14

Table 5.15 Table 5.16 Table 5.17

A fragment of an event log . . . . . . . . . . . . . . . . . . . . . . . . . . . . Process Mining Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differentiation between direct and indirect procurement . . . . Various Process mining application deployment strategies . . . Differentiation between on-premise and cloud deployment strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Different mechanisms used for raw process event data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of data quality dimension for process mining . . . . Data Quality Framework used in this thesis . . . . . . . . . . . . . . . DQD—Completeness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DQD—Uniqueness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DQD—Timeliness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DQD—Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DQD—Accuracy/Correctness . . . . . . . . . . . . . . . . . . . . . . . . . . . DQD—Consistency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Template for DQ assessment at activity level . . . . . . . . . . . . . Template for DQ assessment at various timestamp level . . . . List of standard indirect procurement process meta-attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Template for Data quality assessment at meta-attributes level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Typical ROI cost components for IT investment . . . . . . . . . . . Savings benefit assumptions for the first three years . . . . . . . .

20 25 36 62 63 64 74 75 76 76 77 77 78 78 80 81 83 83 85 89

xvii

1

Introduction

1.1

Motivation and Initial Situation

“Data is the new oil of the 21st century and analytics is the combustion engine”—Peter Sondergaard: Gartner Research aptly describes the vitality of data and how it is changing the dynamics of the world today. Digital evolution has led to availability of enormous amount of data which if utilized in a proper manner could be as an asset or as a deadlock otherwise. Businesses today are dwelling in a world which demands them to be disruptive and innovative. Competition within businesses mandates a more valuebased approach towards process excellence and deliveries. We can encapsulate that organizations today are living in a world best termed by a four-letter acronym called VUCA (Volatile, Uncertain, Complex, Ambiguous). The VUCA, inherent to today’s world has now become a reality because of the pandemic COVID-19 which is a Black Swan situation (cf. Ghosh 2020: 1). VUCA world compels the organizations to make continuous improvement as a part of their DNA. The main challenge companies face to bring about an improvement within their processes is the mindset change. Implementation of improvement opportunities when based on factual process data extracted from today’s information systems provide companies a stronger and reliable option which will help them overcome this mindset barriers. As processes today are fortified with varied diversity of data sources (IT business applications), it has become increasingly important to inculcate the capability of incorporating quickly rising volume of heterogenous data sources into crucial decision-making process. Advent of fact-based process improvement techniques like “Process Mining” is proving as a boon to achieve process excellence and much required agility. Addition of Artificial Intelligence (AI) and Machine Learning (ML) capabilities © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5_1

1

2

1

Introduction

when applied to process mining technology has led to realize automation within this continuous improvement initiative and achieve enhanced execution capacity of processes. As a result, companies using process mining are benefiting from improved lead times, better quality to products owing to waste reduction and resulting enhanced customer satisfaction (cf. Celonis Customer stories 2021: 1). Process mining tool is a key enabler to achieve process excellence and accelerate organization’s digital transformation journey. It has emerged as one of the fastest growing markets in the intelligent automation space due to increasing enterprise awareness of the technology and its benefits. It is rapidly gaining traction across geographies and finding a wide range of used cases across business functions and verticals. The adoption of process mining solutions does not only help enterprises achieve cost savings and operational efficiencies by identifying process optimization/automation opportunities, but also improves workforce productivity and enhance customer experience. Numerous authors including Dakic et al. (cf. 2018: 866), Hind et al. (cf. 2017: 1) and the Process Mining Manifesto itself (cf. Aalst et al. 2012: 170), have highlighted that in the last decade process mining has evolved rapidly in terms of its capabilities, features, and functionalities. It is gaining tremendous attention and is being accepted as one of the most important innovations in the field of Business Process Management (BPM) from both industry as well as research community. This growing relevance of Process Mining within evolving market has been a trigger for this master thesis.

1.2

Problem Statement

Implementation of any new tools within an organisation is a tedious and demanding task. It requires change of mindset, readiness from organizations (employees, infrastructure, business functions etc.,) and could need enormous investment in regard to time, cost and efforts. In such a scenario companies need to take wise decisions with respect to tool implementation by diligently analysing the potential benefit which could be generated against the incurred costs. For this companies use “business case” approach for making decisions in relation to innovations. However, when it comes to IT investments, the conventional business case approach cannot fully cover all the associated risks. This makes the assessment/decision making challenging. Below mentioned description clarifies two problem statements addressed within this Master Thesis. Problem statement #1: The currently existing conventional business case is incomplete and does not fulfil all requirements necessary to evaluate IT investment or process mining in specific.

1.4 Thesis Outline

3

Problem statement #2: Process mining as a budding technique with promising prospective does not have a blueprint or a step-by-step guideline which can direct companies and help them in mitigating risk factors in early stages. A step-by-step assessment can provide companies with a robust, organised and coordinated assessment flow to better anticipate opportunities and associated risks of process mining techniques within their organisational setup. Thus, enabling them to actively take mindful decision regarding Process Mining tool implementation. Currently, in absence of a step-by-step guideline for evaluating process mining implementation feasibility/readiness, companies struggle to make timely decisions and tap promising potential of process mining for achieving operational excellence and enhanced performance.

1.3

Aim of Thesis

In order to have clarity and precise segregation, aim of the thesis has been divided in three aspects: Aim #1: Develop a business case framework for IT investment and identify concepts that contribute to early assessment of process mining. Aim #2: Based on the assessment concepts identified in Aim #1 a guideline will be prepared that will support organizations execute assessment of readiness for process mining within indirect procurement. This guideline is denoted as an “integrated holistic concept to assess Process Mining in Indirect Procurement” in this thesis. Aim #3: Validation of the developed integrated holistic concept through application to case company in order to ensure operational viability.

1.4

Thesis Outline

This Master Thesis report is divided into 7 sections, Figure 1.1 illustrates the thesis outline. Chapter 1 displays the introduction presenting initial situation or background of the thesis. It primarily focuses on the problem statement which explains the reason for the chosen topic and main aim of the Master Thesis. Chapter 2 shows an in-depth literature and theoretical introduction to process mining technology and various aspects linked to the topic. Chapter 3 provides a

1

Figure 1.1 Thesis Outline. (Source: Own illustration)

4 Introduction

1.4 Thesis Outline

5

systematic description of the chosen process of “Indirect Procurement” with its importance and relevance to business process improvement initiatives. Chapter 4 describes a complete overview of digitalization as need of the hour, then further describing conventional business case and associated template. Moreover, this chapter describes “IT specific business case” along with its framework consisting of all necessary aspects required to make a robust and value realizing business case. IT business case is further extended to a “Process Mining Business Case” along with additional aspects specific to Process Mining. Chapter 5 presents the integrated holistic concept for assessing process mining feasibility within indirect procurement. This holistic model consists of four components, the “As-Is” analysis, the risk analysis factors mainly the technical feasibility, data quality assessment and financial feasibility. It is based only on the “beforeimplementation” components of a business case. Advancing, chapter 6 describes the empirical results and presents an in-depth outline of the implication of holistic model on the case company. The last chapter 7 concludes the thesis by displaying main learnings from literature study and practical experience gained during master thesis. Further it discusses about the validity of the produced integrated holistic concept, states the limitations of the integrated model, and defines future research work required as an extension to the business case.

2

Process Mining

2.1

Introduction

Business processes are core part of organizational design and act as a pulse of every organization. Business process can be defined as “a collection of activities that takes one or more kinds of input and creates an output that is of value to the customer.” – (Hammer & Champy 1993: 90–91)

or “a specific ordering of activities across time and place, with a beginning and an end with clearly defined inputs and outputs” – (Davenport 1993: 1)

By the above definition it is clear that business processes act as a base line for achieving operational performance. Hence, they should be continuously analyzed, monitored, and improved consistently to achieve efficient organizational performance and business-value. To do this companies should remove inefficiencies, bottlenecks and ensure compliance issues are improved within processes. Therefore, understanding the current “As-Is” process situation is necessary to set an improved “To-Be” process. Traditional quality management methods associated to business improvements, such as Six Sigma, Lean management, etc., were developed to understand As-Is process. However, according to several authors

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5_2

7

8

2

Process Mining

in the literature and business research institutes these process improvement initiatives show weaknesses/limitations associated with them, shown in Figure 2.1. Furthermore, the researchers have highlighted the way process mining overcomes these shortcomings as follows:

Figure 2.1 Traditional Process improvements methods v/s Process Mining Value Proposition. (Source: Authors own illustration based on (cf. Graafmans et al. 2020: 2); (cf. Aalst 2016: 48); Harvard Business Review (Davenport & Spanyi 2019: 1) and Apromore (cf. 2021: 1))

Process mining is an emerging technology and discipline that seeks to rectify critical flaws of traditional business improvement initiatives by using computational algorithms, AI and machine learning technologies. This enables to create objective process models solely on the factual information coming directly from ERP systems. Many authors Dakic et al. (2018): 866, Hind & Chivi (2017): 1, Aalst (2016): 21, 342 and the Process Mining Manifesto (The associated data can be viewed in Appendix 8 in the electronic supplementary material). It claims Process Mining to be given much attention in the last decade and is being accepted as one of the most important innovations in the field of Business Process Management (BPM) from both industry as well as research community owing to following reasons: ➢ The amount of data being stored about process executions is rapidly growing.

2.1 Introduction

9

➢ There is an increased need for aligned processes with available data to meet compliance, efficiency, and customer requirements. ➢ Process Mining provides data mining, big data opportunities ➢ The rise of commercial tools capable of performing Process Mining analysis and techniques.

2.1.1

Process Mining Market Trend and Providers

Most of the leading research and analysis firms such as Gartner, Everest Group and Forrester promote process mining as an endowment to many organizations. In 2018, Gartner’s report stated that the process mining software market for new product license and maintenance was valued at approximately $160 million in 2020. The global process mining market volume is further expected to triple or quadruple in the next years, Figure 2.2

Figure 2.2 Overview of market volume and Process Mining Products PEAK Matrix assessment. (Source: Gartner report and Everest group. (Source: Gartner Research and Everest Group (cf. Dilmegani 2020: 1)))

Figure 2.2 also features top 13 Process Mining vendors, shown on a PEAK matrix evaluated by Everest group, based on factors like market adoption, level of value delivered and portfolio mix (coverage of industries, geographies, used cases, etc.). Celonis has emerged as a market leader by gaining highest evaluation score for its market impact, vision, and capability dimensions.

10

2

Process Mining

Figure 2.3 shows facts related to application of Process mining within industry. After BFSI (Banking, financial services and insurance), the manufacturing industry is anticipated to rank second in the revenue share. This is due to increased awareness and adoption of automation within manufacturing sector. Process Mining appears to play a vital role in providing measures and solutions for better quality products. Hence, proving as a catalyst to drive market expansion for manufacturing industry.

Figure 2.3 Global Process Mining software market share, by end user 2020. (Source: Fortune Business Insights (cf. 2020: 1))

2.1.2

Discussion of Related Terms

Prof. dr. ir. Van der Aalst is the most cited process mining author in literature and is known as the “God father for process mining”. Aalst (cf. 2016: 12–16) through Figure 2.4 he has illustrated two main fields of a digitalized business: Data science dealing with gaining insights from enormous amount of data using scientific methods and Process science dealing with an aim to improve processes using scientific methods (The associated data can be viewed in Appendix 1 and Appendix 2 in the electronic supplementary material). Process mining is seen as a sub-disciple of both. Aalst (cf. 2016: 447) considers process mining as

2.1 Introduction

11

the “missing link” between data mining and traditional model-driven Business Process Management (BPM) initiatives.

Figure 2.4 Process Mining as the bridge between Data science and Process Science. (Source: Aalst (2016: 18))

Data mining is an important part of data science discipline and perhaps the most closely associated concept (cf. Foster et. al 2013: 4). In Figure 2.4 Data mining is seen partially overlapped by other data science sub-disciples’—statistics, algorithms and machine learning. Statistics is the foundation of data mining used for studying data, data relationships, as well as for dealing with numeric data in large databases. Algorithms or computer codes are the building blocks that constitute AI and are built to apply “human-thought-like” processing to the statistical problems. Machine learning is the combination of both: advanced statistical analysis and AI heuristics, which can be used for data analysis and knowledge discovery (cf. Girija et al. 2006: 591–592). (The associated data can be viewed in Appendix 3 of the electronic supplementary material). The data mining uses

12

2

Process Mining

CRISP-DM (Cross Industry Standard Process for Data Mining) method shown in Figure 2.5 consist of 7 phases (The associated data can be viewed in Appendix 4 of the electronic supplementary material)

Figure 2.5 The CRISP-DM Model. (Source: (Chapman et al. 2000: 10–11))

BPM current status origins from the concept of Process Re-engineering, Process Innovation approaches, Process Modelling and Workflow management with a broader scope using methods, metrics, software tools to achieve operational excellence (cf. Rosemann & de Bruin 2005: 1–3). BPM initiative is cyclic in nature

2.1 Introduction

13

because it is considered as a continuous improvement initiative and covers endto-end business processes to regularly keep up with the constantly ever-changing behaviour of customer needs, technology, and competition. (The associated data can be viewed in Appendix 5 in the electronic supplementary material). Dumas (cf. 2013: 21) proposed the BPM life cycle (Figure 2.6) which comprises of six central phases and a detailed explanation of each BPM phase can be further found in (The associated data can be viewed in #6 in the electronic supplementary material).

Figure 2.6 The BPM life cycle. (Source: Dumas (2013: 21–22))

14

2.1.3

2

Process Mining

Limitations of Data Mining and BPM

According to Aalst (cf. 2016: 15) on one hand data science approaches (like data mining) tend to be “process agnostic” i.e., although it is data-centric it does not take actual end-to-end process into consideration. Business processes are core of any business organization (cf. Dakic 2018: 866). Therefore, with data mining practice in use businesses face certain performance identification challenges like identifying associated bottlenecks, process variations, and process compliance etc. Additionally, it cannot foresee problems by streamlining processes in the form of “As-Is” situation (cf. Aalst et al. 2011: 174). And on the other hand, BPM is not data-centric in nature. To evaluate the functions supported by BPM tool, Netjes et al. (cf. 2007: 17–26) had conducted an extensive evaluation of top BPM tool Suite: FileNet P8. Ideally a new process model should be created, or an existing process model is automatically adapted in the Re-design phase of the BPM lifecycle. However, the authors during their research found that the BPM tool in reality does not support the full BPM life cycle and consequently has reflected two major problems: 1. The actual execution of the process was completely disconnected from the (re)design phase i.e., it was incapable of automatically regenerating process models based on analysis of the current situation. Therefore, in practical the BPM cycle was not fully closed (cf. Netjes et al. 2007:17). 2. To aid the redesign of current process, users had to recompose and reinitiate process models manually all over again with the help of graphical editors. As the system did not provide any ideas for process improvements, designers had to themselves come up with options. Resulting, the tool provided no support to perform diagnosis and extract knowledge directly from factual data originating from event logs. Therefore, Aalst (cf. 2016: 15) and Netjes et al. (cf. 2007: 17–26) do not consider BPM initiatives to be data centric approach. The authors claim that the only way to address the above problems is by connecting “design” and “reality” of the process models i.e., the design and re-design of the process model should be based on the reality observed by the system through evidence-based event data. BPM life cycle can only be closed when factual information about processes is used by the redesign phase. Hence, in order to overcome limitations of data mining and BPM initiatives Aalst (cf. 2016:) and Netjes et al. (cf. 2007: 19) emphasize use of three classes of process mining techniques: process discovery,

2.1 Introduction

15

conformance and enhancement that uses factual process event data in the form of transformed “event logs” to connect reality with design.

2.1.4

Positioning Process Mining

Despite of Process Mining being a new research field, it is well-established and defined by IEEE Task Force ((The associated data can be viewed in #8 in the electronic supplementary material)). Process Mining Manifesto defines Process mining as: “The idea of process mining is to discover, monitor and improve real processes (i.e., not assumed processes) by extracting knowledge from event logs readily available in today’s (information) systems. Process mining includes (automated) process discovery (i.e., extracting process models from an event log), conformance checking (i.e., monitoring deviations by comparing model and log), social network/organizational mining, automated construction of simulation models, model extension, model repair, case prediction, and history-based recommendations.” – (Aalst et al. 2012: 172)

The above-mentioned definition embraces the unique approach of process mining to allow the analysis of any business process to be based on digital traces captured in ERP systems, used by process mining in the form of transformed event logs. It is the only technology which makes optimum utilization of valuable event data. “Process mining is builds on two pillars: (a) process modeling and analysis and (b) data mining” (Aalst 2016: 89). Figure 2.7 illustrates how Process Mining connects data-driven and process-centric approaches and the way it uses hidden evidence recorded as event data in IT systems. It also assists to create current “As-Is” of business processes in the form of process models and further supports in identifying areas of improvements by answering various performance and conformance related questions (cf. Aalst 2016: 26). This makes process mining a holistic concept for a continuous improvement initiative. It is quite evident that an event log is the foundation and the backbone for process mining technology. Aalst (cf. 2016: 125) claims event log to be the primary input and starting point for any process mining analysis. Several authors like van Zelst et al. (cf. 2021: 4), Peters & Nauroth (cf. 2019: 25) have strongly supported the importance of event log for Process Mining. Hence, the next sections of this chapter describe event log in detail based on its origin, characteristics, and the way it supports different process mining techniques.

16

2

Process Mining

Figure 2.7 Process mining as a combination of data-driven and process-centric approaches. (Source: Aalst 2016: 26)

2.2

Event Data and Event Logs

With the rise of digitalization, organizations typically use various PAISs (Process Aware Information Systems) to perform their day-to-day business activities and are independent of each other. For example, PAIS’s like ERP system SAP for creation of purchase orders, CRM tools for customer order handling and Salesforce to manage company sales, shown in. Whenever a process activity is executed in any PAIS’s, raw event data is generated which stores information about the executed activity in a tabular format of PAIS’s data source tables. As shown in Figure 2.8, required event data typically originates from multitude of data sources due to technical and organizational conditions Aalst (cf. 2016: 126). The first step is to obtain the project relevant raw process event data by applying ETL (Extract,

2.2 Event Data and Event Logs

17

Transform and Load) techniques. Aalst (cf. 2016: 22) highlights that four V’s of “Big Data”; Volume, Velocity, Variety, and Veracity apply to raw process event data. Hence, extracting, merging of event data from different data source systems and converting it into a system interpretable event log format poses a challenge.

Figure 2.8 Event data flow for Process mining. (Source: Adapted by author, based on Aalst (2016: 127))

In order to ensure credible input for process mining techniques, event logs should have a specific standard format. This standard format defines the structure of an “event log” sometimes also known as an “activity table”.

18

2

Process Mining

An event log can be defined as below: “An event log is a collection of events extracted in the context of a process that indicates which activity has happened at a specific time.” – Sani, M.F. (2020: 1)

As Process Mining assumes “event log” as a collection of events, each event exhibits some typical elements and characteristics. To explain the association and cardinality of event log elements, “class diagram” illustrating different event elements and event log characteristics is displayed in Figure 2.9.

2.2.1

Characteristics and Structure of an Event Log

Figure 2.9 Class diagram of an event log. (Source: Aalst (2016: 147))

2.2 Event Data and Event Logs

19

Figure 2.9 can be summarized as below: (cf. Dakic et al. 2018: 867); (cf. Aalst 2016: 128 f.) • There is one event log per process and all events in the event log can be related to that particular process. • Each process consists of any number of activities, but each activity is assigned to exactly one process. • Every case belongs to a process. A case is a process instance. • Each activity instance relates to exactly one activity. • Each activity instance belongs to exactly one case. • Each event relates to exactly one case. • Each event belongs to exactly one activity instance. • Each case attribute relates to exactly one case; every attribute has one name and a value “(birthday, 12-02-1987)”. • Each event attribute belongs to exactly one event and consists of a name and a value together, such as “(costs, e200)”. • There are various sub-categories of event attributes, such as the timestamp, the executing resource, or the costs of an event etc. • Multiple events belonging to a single line-item are linked together in a process instance or case. • The events are arranged based on timestamps—forming a sequence of events. • An event log contains three mandatory attributes without which process models cannot be visualized and are as follows: 1. Case Id It is a unique identifier which enables identifying one single case of the process, for example, in a Purchase-to-Pay (P2P) process the handling of one purchase order line-item number forms one case ID (usually combination of purchase order number and purchase order line-item number). It influences the process scope as it solely determines where the process starts and where it ends for each case. Therefore, it is mandatory to know for every event, which case it refers to, so that the process mining tool can compare several executions of the process to one another. 2. Activity name Corresponds to the name of the process event performed in the ERP system. Activity names are assigned at the time of event log creation by data scientists. The aim is to describe reference of each activity within a single row of an event log

20

2

Process Mining

3. Timestamp The third most important prerequisite is to have at least one timestamp column that indicates at what date and time each of the activities took place. Timestamp for process mining follows a particular representational format, for example: dd/mm/ yyyy, hh/mm/ss. With the help of timestamps, performed activities are ordered in the sequence they took place. This supports in identifying delays between activities, bottlenecks in the process and measure the compliance performance. Table 2.1 shows a small fragment of a larger set of event log that illustrates typical information with events related to the so-called P2P process. Each row in the table represents to an event that took place, which captures a specific execution of an activity within a case. Different columns in the table are associated attributes of the events (van Zelst et al. 2020: 2–3). Table 2.1 A fragment of an event log

Source: Own illustration adapted from Aalst (cf. 2015: 5)

Once an event log is appropriately created the base for producing automatic process model becomes concrete to understand the “As-Is” process model. Further with creation of appropriate event logs different process mining techniques can be applied. Process mining techniques supports to automatically discover the “As-Is” process models based on information available in the event log (cf. Aalst 2016: 132) (Figure 2.10)

2.3 Process Mining Techniques

2.3

21

Process Mining Techniques

Figure 2.10 Positioning of the three main types of process mining: (a) discovery, (b) conformance checking, and (c) enhancement. (Source: (Robledo 2018: 1))

Figure illustrates different components of process mining techniques. It represents a real “world” that consists of organizations where business users carry out day-to-day business activities supported by various Process Aware Information Systems (PAIS’s) like ERP, CRM (Customer Relationship Management), Salesforce etc. These systems record event data about process activities. Process mining techniques use typically three artifacts described below: • The inputs: 1) an “event log”

22

2

Process Mining

2) an “a-priori model” (“Should Be”/Ideal Model): It illustrates optimal business process flow i.e., the most efficient and effective process route from initial input to final output for the defined primary goals. By default, an “a-priori model” only displays activities and connections. • The output: “process model”, that graphically describes the control-flow of a process based on the event log data. With the help of these artifacts, three types of process mining techniques can be applied defined as below:

2.3.1

Process Discovery

Process discovery is defined as the act of gathering information about an existing process and organizing it in terms of an “As-Is” process model. This technique takes an event log and automatically constructs a process model. The primary goal of this technique is to create transparency and the acquisition of current “AsIs” process knowledge, it highlights routing probabilities, determines the most frequent common path in the process and discovers common and uncommon behaviour of cases by displaying different process flow variants (cf. Ailenei et al. 2012: 5). Process discovery can potentially offer a greater level of flexibility when mining complex and infrequent process patterns (cf. Turner et al. 2012: 505). For example, using the α-algorithm (The associated data can be viewed in #9 in the electronic supplementary material). α-Algorithm is a typical example of a process discovery, which clarifies the basic idea behind many process mining algorithms. There are many ways of representing a process model using notation/ languages like BPMN, EPK, YAWL or Petri-Nets. The α algorithm discovers a Petri net (The associated data can be viewed in #9 of the electronic supplementary material) by identifying process patterns in the event log (cf. Aalst 2016: 194) Figure 2.11 illustrates first four discovered variants of process model based on a larger event log sample for a P2P process, produced automatically through a process mining software. The dotted arrows represent the start and end of the described case where each case represents a single Purchase Order (PO) line-item number. The first variant A shows 107,688 cases that follow the most common process path. Variant B represents the second most common variant where 38,069 cases out of 145,777 were generated without creation of requisition item, variant C shows 37,270 cases of “change price” creating deviation and delay in the

Figure 2.11 An example of process discovery output based on event log of a P2P process. (Source: Celonis Training)

2.3 Process Mining Techniques 23

24

2

Process Mining

process, and variant D answers compliance related questions where 18,938 cases start with invoices scanning activity, a typical maverick buying example.

2.3.2

Conformance Checking

The inputs for this technique are the event logs and the a-priori model. The aim of this technique is to define an a-priori model and compare it with reality from event logs, to detect and diagnose for any discrepancies and commonalities between an a-priori model and the event log (cf. Netjes et al. 2007: 18). Conformance checking helps to discover the number of cases that do not conform to the a-priori model. It explains where the processes deviate themselves and show process flow violations (as discussed for Variant B, C and D in Figure 2.11), it supports in measuring the severity of their occurrences, find process loops if any. And measure the overall level of compliance performance (cf. G”unther et al. 2007: 111); (cf. Ailenei et al. 2012: 5). Different quality dimensions are used to evaluate how well a process model describes the observed behavior. Conformance checking is particularly important in order to check the quality of reproduced process model in comparison to the event log. (The associated data can be viewed in #11 in the electronic supplementary material) (Figure 2.12).

Figure 2.12 Four quality factors for assessing process model quality. (Source: Van der Aalst (2016:189))

2.3 Process Mining Techniques

2.3.3

25

Enhancement / Process Re-engineering

Enhancement techniques also require a-priori model just like conformance technique. However, the difference lies in the purpose. Enhancement technique aims at changing or extending the a-priori model and has two goals: 1. Repair the process model to better reflect reality: It concerns to the improvement of the existing process model by directly using information about the actual process recorded in the event log. e.g., repairing the process model in case it was wrong, extending it in case a bottleneck was identified, or to interpret safe decisions. 2. Extension: Adding additional perspective Aalst (cf. 2016: 34) claims that when process models are further supported with supplementary resource (metaattributes), additional perspectives can be found within an analysis by cross relating it with the log, that provides more information to support the process mining analysis like “WHO?” was involved—organizational perspective, “What?” happened Case perspective and “When” did it happen? time perspective. This area of “Process Enhancement” is called an extension. The additional perspective are presented in Table 2.2 below. If a detailed explanation of each perspective is required then it can be found in (The associated data can be viewed in #11 in the electronic supplementary material) Table 2.2 Process Mining Perspective

Source: Own illustration, based on Aalst (2016: 17 f., 27, 290–293, 294 f.)

26

2

Process Mining

To discover different redesigns and control strategies, the additional perspectives defined above can be combined into a single unified process model for next simulations and “what-if” analysis. Figure 2.13 illustrates a fully integrated model covering the organizational, time and case perspective.

Figure 2.13 Connecting event log and model (with extension). (Source: (Aalst 2016: 298))

However, the three process mining techniques described above performs analysis only on “post-mortem” event data i.e., cases that have been already executed “offline analysis” or historical event data. Therefore, taking actions before things go wrong is not possible. The Process Mining Manifesto (Aalst et al. 2012: 176) has claimed, process mining techniques to be not just limited to historical data analysis but it is also capable in dealing with currently running or incomplete cases, known as “online analysis”. This is possible due to advancements in information technology since process event data is continuously getting updated as soon as process activities are executed and therefore running cases are added

2.3 Process Mining Techniques

27

to the event logs in real-time. Hence, process mining supports both offline as well as online analysis, this technique is known as “operational support” (cf. Aalst 2016: 301). Figure 2.14 shows process mining techniques in terms of its inputs and outputs.

Figure 2.14 Process mining techniques explained in terms of input and output. (Source: Own illustration based on Aalst (cf. 2012: 190))

2.3.4

Operational Support

This technique supports business operations by finding tricky running cases, anticipating the future, and recommending corrective actions. This technique is based on the process model derived from event logs on real-time basis and is heavily supported by advanced technologies like Artificial Intelligence (AI) and

28

2

Process Mining

Machine Learning (ML) algorithms. Mainly, three activities are performed under operational support: 1. Detect: An automatic alert is generated to business users as soon as process model deviates from the a-priori model at run-time. The IT systems (automation) and business users can take immediate actions based on the type of alert received. For example, in a P2P process when supplier invoice is scanned before the goods are received, an alert is sent to responsible buyer to investigate the missing goods (Figure 2.15).

Figure 2.15 Detecting violations at run-time. (Source: Aalst (cf. 2016: 307f.))

2. Predict: It is supported by offline analysis where current status of a running case is compared to similar cases from the past to predict what may happen next. For example, predicting throughput time for a case based on similar cases from the past (Figure 2.16).

Figure 2.16 Partial trace of a running case and some predictive models, providing a prediction. (Source: Aalst (2016: 311f.))

2.3 Process Mining Techniques

29

3. Recommend: It operates similar to the “predict” activity but differs in its output. The output for “predict” activity is the prediction of what may happen next whereas the “recommend” activity gives output by proposing what countermeasures should be taken based on the prediction, in order to eliminate upcoming problems and ensure correct and efficient handling of cases. For example, minimize total costs by sending alerts to business users to execute supplier payment on-time to utilize the cash discounts, improve throughput time by investigating why an activity is running longer than expected, save excessive resource usage etc (Figure 2.17).

Figure 2.17 Model based on historic data, providing recommendations for running cases. (Source: Aalst (2016: 316))

By the above literature study conducted on process mining, it is clear that creating an “event logs” is the most important pre-requisite for executing process mining techniques. Creating an event log isn’t a simple task and involves manual efforts by data scientists. Since Process mining is still an emerging technology it is constrained to manual errors in real-life (cf. Aalst et al. 2012: 179). The output of process mining is highly dependent on the quality of event logs generated. Hence, in order to ensure relevant and appropriate extraction of raw event data from the data source systems, the IEEE task force has listed six important guiding principles for data scientists. (The associated data can be viewed in #13 of the electronic supplementary material). Furthermore, existing frameworks which aim to structure process mining techniques, such as L*-methodology (cf. Aalst 2016: 329–396) and the PM2-methodology (The associated data can be viewed in #14 of the electronic supplementary material), have given limited attention for data quality assessment. Numerous authors in the literature, to name a few like (cf. Fox et al. 2018: 2–9); (cf. Wynn et al. 2019: 10–15); (cf. Cruchten 2019: 1–6) emphasize that along with proper extraction of raw event data it is also necessary to have good quality event data. The literature study on data quality in process

30

2

Process Mining

mining has given more attention to data quality of “event logs” and only a few authors like Andrew et al. (cf. 2019: 25) highlights the importance of data quality assessment at an early stages of process mining project i.e., before the creation of event logs. Several data quality and technical challenges are known and listed throughout the process mining literature. This thesis has conducted an in-depth study of these challenges and has ensured its incorporation with the integrated holistic concept for process mining assessment illustrated in Chapter 5. These challenges are elaborated in (The associated data can be viewed in #15 and #20 of the electronic supplementary material)This master thesis has chosen indirect procurement as an objective function to explore the feasibility of applying process mining. The next chapter displays indirect procurement as a function, its corelationship with direct procurement and its importance within organization.

3

Indirect Procurement

3.1

Introduction to Procurement

Organizations today are dwelling in a demanding and turbulent environment under intense pressure of tough competition. Hence it becomes increasingly necessary for companies to secure their competitive edge by procuring efficiently. This demands effective internal procurement processes. Today, a procurement function has transformed from being transaction-oriented perspective to a strategic oriented sourcing enterprise (The Future of Procurement 2017: 1). It is no longer viewed only as tactical, clerical, or administrative function, but as an active catalyst which enables transformation and assist in achieving strategic organizational objectives through competitive advantage. ‘Strategic purchasing’ may be defined as: “the process of planning, implementing, evaluating, and controlling strategic and operative purchasing decisions for directing all activities of the purchasing function towards opportunities consistent with the firm’s capabilities to achieve its long-term goals” – (Carr et al. 1997: 200; Ellram et al. 1994)

Whereas procurement is defined as: “the whole process and responsibility of acquiring goods, works and services for the production and operation of an organization from external suppliers.” – (Lysons et al. 2012: 6; Iloranta et al. 2012: 49–53). Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-3-658-41453-5_3. © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5_3

31

32

3

Indirect Procurement

The most important function of procurement is to enhance organization’s profitability by reducing costs. Apart from the cost prices, the quality and on-time supply of the procured materials are key factors in successful procurement. (cf. Van Weele 2010: 4 f.). Henderson (cf. 1975: 44) had predicted the increasing importance of procurement in corporate management in early sixties. Today, procurement has transformed from being a passive, administrative and reactive process to a strategic, proactive, and multi-faceted function which directly contributes value to organizational goals (cf. Rendon 2005: 18). The procurement function also embraces other supply management functions like materials management, logistics and physical distribution and this holistic approach has been labeled by many industries as “supply a management” (cf. Rendon 2005: 5). Conventionally, performance of buyers was judged on their ability to keep production line in operation with least possible price for bought goods and services leading to virtuous savings. But with the introduction of new “strategic sourcing” concept in supply management, companies now focus more on the value-adding output aspects such as quality, Total Cost of Ownership (TCO), time to market, assurance to continuity of supply, use of advanced technologies etc. Today’s supply management function promotes breaking of barriers and working closely within multi-functional teams in order to best utilize and channelize available discrete competences. It also utilizes the advent of digital means through electronic procurement and business data analytics in order to estimate the performance most optimally while adopting strategic sourcing approaches (cf. Rendon 2005: 6; cf. Burt et al. 2003)

3.2

Direct and Indirect Procurement

A company‘s purchasing activities can be classified as direct procurement or indirect procurement.

3.2.1

Direct Procurement

Direct procurement can be defined as: revenue generating expenditure, or expenditure which can be related directly to the product or services being sold to the customer. – (cf. Cox et al. 2005: 41, 39–51; cf. Xideas & Moschuris 1998: 32, 974–992).

3.2 Direct and Indirect Procurement

33

Direct procurement has been more thoroughly studied in the literature and is far better understood in industry, for a multitude of reasons: (cf. Israel et al. (2020): 1) 1. “Most procurement research has been done in the manufacturing industry, which typically consists of a large proportion of direct spend within total spend. 2. Established cost-based accounting procedures and current IT technologies allow for easier tracking and analysis of costs that directly impact end products. 3. The purchasing of direct items is usually centralized and supported by the organizational structure of the company, as direct procurement has more often been perceived as strategic to financial performance”. For these and other reasons, in comparison to indirect procurement, direct procurement is more properly defined and better understood by both academia and industry. And considering its strategic relevance, tremendous efforts have been put to streamline the inflow of direct goods to manufacturing lines and in turn increase the procurement efficiency (cf. Kim et al. 2003: 53). Direct procurement can be scheduled in timely manner to meet the demand using EDI (Electronic data interchange) applications and automated replenishment systems through JIT (Just in time) set-up (Shrinivasan et al. 1994: 1291 f.). Buying process in direct procurement is started via planned demand which is forecasted through a relevant BOM (Bill Of Materials). BOM contains list of items required to create a product and the instruction on how it can be done is further demonstrated in appropriate MRP (Material requirement planning) or ERP. In contrary indirect procurement demands are unplanned and cannot be regulated via defined bill of materials (Hamblin 2020: 1). The next sub-section describes fundamental characteristics and functions of an indirect procurement in detail.

3.2.2

Indirect Procurement

Indirect procurement is defined as: “non-revenue generating expenses or expenses which do not directly relate to the products or services being sold to the customer” – (cf. Cox et al., 2005: 41; cf. Xideas et al. 1998: 32).

34

3

Indirect Procurement

The scope of Indirect procurement encompasses virtually every aspect of an organization ranging from IT equipment, engineering services, freights and logistics, MRO (maintenance, repair, and operations) related to tools and machines, facility management services, consumables, consulting services, marketing utilities, legal and human resources, temporary labor, and travel. Indirect procurement has been less studied by the academia and industry (cf. de Boer et al., 2003: 911 f.; cf. Gunasekaran et al. 2008: 113–159). One reason also being confusing terminologies used as “Indirect procurement” or “Indirect Spend” is referred to any number of goods and services. Many studies refer indirect spending by various names thus making any conclusion challenging. Segev et al. (cf. 2000: 107–128) defined indirect spend most briefly as: 1) Capital equipment 2) Services 3) Maintenance, Repair and Operating supplies “MRO” ideally refers to a subcategory of Indirect spend (cf. Gunasekaran et al. 2009: 122, 161; cf. Yu et al. 2015) or is used synonymously with entirety of indirect spend (cf. Cox et al. 2005; cf. Yates 1998: 80). Other terms for indirect spending include “non-product related” spending (cf. Xideas et al. 1998), “service purchasing” (cf. Thomas et al. 2015: 314–332), “operating resources” (cf. Segev et al. 2000: 107–128), as well as being referred to implicitly when referring to the outsourcing of certain industries. Since 2011 indirect spend has been growing at an estimated 7 % per year globally (cf. Boulaye et al. 2019: 1). Therefore, untapped cost savings potential of indirect spending is being recognized now by firms and scholars in spite of its smaller contribution to total spending. Although direct spending still accounts for maximum value of company‘s purchases, the indirect spend accounts for majority number of purchases a company makes (cf. Barry et al. 1996: 35–44; cf. Cox et al. 2005). Indirect spend contributes about 80 % of purchases with 20 % of value. (cf. Van Weele 2005). Hence, signifying low value high frequency purchases which require huge time commitment from organization (cf. Bechtel et al. 1997: 18–33). This signifies a need for optimized indirect purchasing strategy which will enable considerable amount of time and resources savings for a company. These savings are more of a soft/intangible nature in the form of reduced process times, efficient resource planning and reduced time to market (cf. Porter 1999: 127, 55–60). Nakata et al. (cf. 2008) highlights that indirect purchasing also contains low frequency purchases like Information technology which is of high importance.

3.4 Excellence in Indirect Category Management

35

In the digital era, IT procurement is of vital importance to firms success and possessing higher IT capability implies improved market and financial performance. Concluding, Indirect thus encompasses high and low frequency purchases of varying importance. Bad organization, reporting and management makes indirect procurement to be perceived as a supporting function (cf. Iloranta et al. 2012: 62). Spend related to indirect procurement is mostly concealed under other expenses and spread between different departments. Hence it is challenging to distinguish this spend. In many companies indirect goods/services are bought by different departments based on their choices without an professional procurement competence (cf. Iloranta & Pajunen-Muhonen 2012: 21–22, 64) This dispersed, overlapping and unregulated buying leads to enormous operational transactional cost (cf. Van Weele 2010: 86). Such an environment is also vulnerable to procurement of goods/services out of agreed supplier contracts. This leads to buying in a non-complaint ways against company´s procurement policies and is referred as “Maverick buying” (cf. Chaffey 2011: 366). Aforementioned facts make Indirect procurement in most companies as a sporadic and imperceptible section which is difficult to comprehend. Hence better control of indirect goods/services can be achieved through transparency, thus leading to enhanced cost savings. Table 3.1 in the next sub section illustrates detailed differentiation between a direct and indirect procurement functions.

3.3

Direct v/s Indirect procurement

The portion of direct procurement in comparison to indirect procurement varies depending on the type of industry. Manufacturing industries depend more on direct procurement and indirect procurement is a key driver for service-oriented companies.

3.4

Excellence in Indirect Category Management

In order to overcome the challenges of monitoring and accurately reflecting the performance of fragmented indirect categories located in diverse locations, companies need a clear direction which combines technology along with traditional approaches of category management. This will enable companies to address the key issues of their indirect process, capabilities and assist in making the most efficient use of available data. A study by McKinsey & Company

36

3

Indirect Procurement

Table 3.1 Differentiation between direct and indirect procurement

Source: (cf. Loi 2013: 1; cf. Ahmad & Muddassir 2021:1)

(cf. Boulaye 2019: 1) shown in Figure 3.1 reflects focus on six elements through which indirect procurement function can unleash their untapped potential. In order to maximize value realization through implementation of cutting-edge technologies like intelligent spend engines, automated procure-to-pay and advanced analytics solutions, process mining can play an important role by cleaning the existing systems, processes and preparing the foundation to enable best use of these advanced innovations within companies. A detailed explanation of each element and its benefit can be found in (The associated data can be viewed in #12 of the electronic supplementary material)

3.4 Excellence in Indirect Category Management

37

Figure 3.1 Indirect spend optimization through value-capture best practices. (Source: (cf. Boulaye 2019: 1))

4

Business Case

4.1

Digitalization

“Change is the law of life and those who look only to the past or present are certain to miss the future.”—words by John F. Kennedy aptly describe the urge for companies to adapt and be ready for upcoming digital revolution. Even Jeff Bezos the founder of Amazon mentioned that “In today’s era of volatility, there is no other way but to re-invent. The only sustainable advantage you can have over others is agility, that’s it. Because nothing else is sustainable, everything else you create somebody else will replicate”. Digital transformation has become a buzzword in the business world in recent years. It is also believed to be the next industrial revolution in terms of size and impact and consequently calls for a major organizational adjustments. Companies which manage to transform survive, and the ones which fail to pursue this change perish consequently. Hence it has become vital to consider digital transformation as a business way of life or discipline. Digitalization can be defined as: “the use of digital technologies to change a business model and provide new revenue and value-producing opportunities; it is the process of moving to a digital business.” – Gartner (2021: 1)

In today’s dynamic world, digitalization is no longer only a strategic prospect. It has become a matter of survival. Research shows that 70% of companies do not correctly utilize the full potential of their digital transformation (cf. Deloitte 2021: 1) and a survey done by McKinsey (cf. 2018: 1) found that only 30% of companies succeed in their initiatives for digital transformation. The survey © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5_4

39

40

4

Business Case

also discovered that companies with successful digital transformation used multiple sophisticated technologies such as AI, Internet of things (IoT) and advance neural machine learning techniques. Figure 4.1 shows a broad spectrum of technologies that can be used for digital transformation. The bar columns marked in dark blue implicitly display technologies which have enabled successful digital transformation. Digitalization implies a broader perspective consisting of changing the way an organization is managed. It consist of deploying digital means to improve existing operations and retain competitive edge within the market. The market has changed drastically in recent years and digital revolution has swamped many aspects of our lives. Specifically current pandemic augmented the need for digitalization, and it has become more vital. “Disrupt or be disrupted”: advice by former Cisco’s CEO—Mr. John Chambers, very suitably highlights the necessity for companies to urgently adapt and be part of current digital evolution. Hence digitalization is no more a matter of choice. Literature research in regards of a business case for digitalization/digital transformation on platforms like ResearchGate, Academia.edu, Google scholars, EBSCOhost have not shown satisfactory results. Considering the higher failure rate, it becomes increasingly necessary to ensure a successful digital transformation. A research by McKinsey & Company (cf. 2018: 1) proposes golden rules which can enable successful digitalization scheme (Figure 4.2). Competitiveness, innovation, and customer orientation are key to any business. In order to achieve this a successful enterprise-wide digital transformation is necessary (cf. IBM 2015: 2). Therefore, many businesses have made digital transformation as their organizational strategy. But adapting new technological transformation which can best fit company objectives and strategies could be very challenging. In order to secure and accommodate project funding, executive management needs strong evidence that specific technology will deliver required output. They need to trust that the new technology will reap more benefits (tangible and intangible) in relation to its investment. These digital technologies used for business function transformation could range from a few thousand to millions and could affect the company with regards to loss of time, costs, and resources, if wrong technologies are selected. In order to ensure selection of appropriate digital technologies which meet business objectives and brings true value to the company, an evaluation through a robust and actionable business case becomes mandatory. An actionable business case will not only provide required confidence to the executive management or decision-making board to establish a new initiative or

Figure 4.1 Digital technologies, tools, and methods currently used by organizations. (Source: Deloitte (2021: 1))

4.1 Digitalization 41

42

4

Business Case

Figure 4.2 Golden rules for successful digitalization. (Source: McKinsey & Company (2018: 1))

project but also act as a roadmap that demonstrates clear commitments and profitable execution throughout the transformation (cf. IBM 2015: 8). Hence in order to identify and adopt best technologies, it is crucial to understand the most appropriate way to formulate a logical business case. A good business case supports to evaluate whether a project or initiative is advantageous, feasible, achievable and a valuable investment (cf. Microtool 2021: 1). In the next sub-section this thesis reflects literature review study on generic business case and its vital components.

4.2

Introduction to Business Case

A standard definition of what actually represents a business case is still lacking in the literature. Because literature knowledge concentrating exclusively on business case is scarce. Maes et al. (cf. 2013: 47) highlights that within the little literature available most researchers have only covered a limited aspects of a business case without much further elaboration in their article and few have incorporated

4.2 Introduction to Business Case

43

the business case concept directly in their main research scope. As a result, the knowledge on a business case is scattered throughout literature and lacks a clear standard definition. IBM defines business case as: “a well-structured, formal document, that tells the story of an initiative from the beginning—what situation triggered the initiative, to the end, what benefit, value, or return is expected. An actionable business case is typically used to craft a unified vision, secure project funding, and obtain commitment.” – (IBM Global Business Services 2015)

A business case not only examines prediction of estimated benefits, costs, time, resources, and risks, but a running project is also justified with it because the business case is regularly checked and updated from the time of project proposal until the project is successfully completed (cf. Al-Mudimigh et al. 2001: 216–226). Therefore, it is one of the most important document used for decision making throughout a project life cycle (from initiation, realization until evaluation). Such a preparation helps in increasing the value for a project and reduces risks in every stage of the project life cycle. A well thought, defined, and documented business case also provides decision making stakeholders required confidence and level of certainty about the successful implementation of the project. In the next sub-section, contents of a general/conventional business case are illustrated in order to get a solid understanding of which factors constitute the decision-making while introducing new innovations or investment opportunities within an organization.

4.2.1

A general Business Case Template

A general business case template consisting of various elements shown in Figure 4.3

44

4

Business Case

Figure 4.3 General/conventional business case template. (Source: Own illustration based on IBM (cf. 2015: 8); Sawant (cf. 20201: 8–9); Marshall (cf. 2018: 21–25))

4.2.2

Advantages of a Conventional Business Case

• Making a business case facilitates collection of basic information and clarify assignment of responsibility. (cf. Smith et al. 2010: 65–81) • It can be used as a communication instrument to convince relevant stakeholders.

4.2 Introduction to Business Case

45

• It enables comparison and prioritization of investment options. (cf. LeFave et al. 2008: 171–179, Smith et al. 2010: 65–81) • It can be used as an instrument to evaluate results and display its impact. (cf. Raymond et al. 1995: 3–16) • As several alternatives are assessed and evaluated for implementation, the decision-making covers a wider scope. • As the possible effects of projects are highlighted the decision conviction is enhanced. • As all monetary and non-monetary facts are summarized, this leads to increased transparency within entire project. • Transparency within the projects can further enable traceability. • As potential risks and corresponding mitigation actions are focused from beginning at a very early stage, risk management is considerably improved. • As all the reasons in relation to decision of business are well documented, commitment towards all actions to reach planned objectives is robust.

4.2.3

Limitations of Conventional Business Case for it Investments

• Conventional business case focuses more on financial aspect while ignoring the qualitative benefits (intangible) which in fact provides better view of potential business value for an IT investment. • Technical feasibility of targeted IT investment to the organizations current “As-Is” IT infrastructure is not covered. • In order to set a baseline and plan “To-be” aspiration, visualization of current process breakdowns, identification of overall organizational risks and areas of improvement is crucial to be carried through “As-Is” analysis. Traditional business case does not guide for an “As-Is” analysis to be conducted (cf. Cousins 2021: 1) • Researchers (cf. Bruch et al. 2002: 62–69) highlighted that conventional business cases miss the link between investment objectives and organizational goals. • Apart from financial benefits, there is no recognition of other benefits and consequently no measures are defined for these benefits. (cf. Ward et al. 2008: 3) • Ideally, benefits considered need to be proved in size and magnitude through evidence and in order to ensure commitment, benefit owner should be identified. Identifying a benefit owner ensures commitment to the project and

46

4

Business Case

displays its vitality by adding certain weightage to the project. Traditional business case does not make such a detailed analysis (cf. Ward et al. 2008: 3) • The way Business case is ought to be realized is not highlighted. And benefits are not linked directly to the changes which need to be carried out in order to achieve those benefits (cf. Ward et al. 2008: 3) • Traditional business case also lacks in identifying change owners for business which does not safeguard the completion of changes, thus resulting in failed delivery of benefits. Identifying a change owner helps to build commitment and shows not only the investment profit but also determines how it can be achieved (cf. Ward et al. 2008: 8). Owing to the above limitations of a conventional business case it is evident that IT innovations and investments require a new revised business case template which fits the evaluation criteria’s for robust decision making. Hence in the next subsection, literature-based business case framework for IT investment is discussed to conduct a new IT initiative.

4.3

Business Case Framework for it Investments

Companies can and should make a business case in order to decide on a range of digital technologies, depicted in Figure 4.1 while ensuring that these technologies also fit the organizational and strategic needs of a company. Authors in literature, Maes et al. (cf. 2013: 55), Ward (cf. 2008: 4–5) claim that the aforementioned conventional business case template is insufficient and does not completely support to evaluate the best fit and implement right IT technologies for a company. Because the contents of a conventional business case focuses more on financial aspect. According to Smithson and Hirschheim (1998: 158–174), strategic alignment between the organizational objectives and investment initiatives cannot be achieved if only quantifiable aspects are considered. (Urbach et all 2010: 184–206) also mentioned that business cases considering only monetary aspects are dubious. (Ward et al. 2008: 7) noticed that along with quantitative benefits qualitative advantages provided a holistic view of potential value which could be achieved from project. But these aspects are often overlooked considering the sensitive and political nature along with their ability to hamper approval of the project. (Farbey et.al. 1999: 189–207). These limitations posed by conventional business case signify a need for additional benefit factors to be considered while formulating a business case for IT investments.

4.3 Business Case Framework for it Investments

47

Literature research in relation to Business case for IT investment has shown very few good outcomes. Two research papers have been found with virtuous inputs. One paper by Ward et al. (cf. 2008: 1–15) is focused on building a better business case for IT Investments. Figure 4.4 below displays Ward´s perception about a possible business case mainly focusing on 5 aspects of Drivers, Objectives, Benefits, Costs and Risks.

Figure 4.4 Business case focusing on 5 aspects by Ward et al. (Source: Ward et al. (cf. 2008: 1–15))

Second paper by (cf. Maes et. Al. 2013: 47–57) focuses on preparing a complete framework for business case in relation to IT investment. Figure 4.5 shows 6 key elements as a part of business case framework.

48

4

Business Case

Figure 4.5 Business case framework by Maes et al. (Source: Own illustration based on (cf. Maes et. Al. 2013: 48))

Business case framework signifies a holistic view encompassing 6 dimensions required to prepare a robust business case. Some on the dimensions can be linked to three stages of investment life cycle: before, during and after implementation (cf. Hitt et al. 2002: 71–78). This thesis displays a global view of a logical business case framework without dwelling in detailed with implementation phases. Figure 4.6 shows overview of elements associated to each dimensions within a Business case Framework.

4.3 Business Case Framework for it Investments

49

Applicaon Area Defines the area which will be affected / benefied from intended investment. It mainly divides into 2 subdivisions : Technological Orientaon : e.g., Data warehousing system, E-business, Global data synchronizaon network, Soware repositories, Soware product line adopon, Group decision support systems etc. Organizaonal Orientaon : e.g., Business processes, Effecve global business teams, Strategic vision for IT, Supply chain integraon, Global shared service centers, Execung strategic change, IT, and service-oriented architecture etc.

Business Case Content It is an organized overview of investment elements which characterize vision, objecves, scope, drivers, roadmap, responsibility and accountability, risk, associate impacts, and governance. a) Investment descripon b) Objecves c) Requirements d) Impact e) Risks f) Assumpon’s, consideraons, and scenarios g) Governance A detailed list of business case content as a template for IT investment is illustrated in Figure 4.7 Business Case Process (5 Step approach) The Business case process is a step-by-step guide which closely defines the business case development. It requires to be organized as per 3 stages of Investment life cycle: Before implementaon (Iniaon), During implementaon (Realizaon) and Aer implementaon (Evaluaon): Step 1. Idenficaon of internal and external business drivers Step 2. Idenficaon of ancipated benefits with their corresponding measures and owners Step 3. Development of a framework to categorize and link benefits and changes (Organize benefits as Do New Things/ Do Things beer / Stop Doing Things) Step 4. Idenficaon of organizaonal changes along with respecve change owners Step 5. Determining clear value of each benefit based on evidence Step 6. Idenficaon of all associated costs and resulng risks which could affect ancipated results Business Case Goals It covers diverse objecves to portray the intenon of business case and show different tangible and intangible benefits investment can bring.

Figure 4.6 Integrated business case framework. (Source: (cf. Maes et. Al. 2013: 47–57); Ward et al. (cf. 2008: 1–15))

50

4

Business Case

Relevant Stakeholders It displays the list of stakeholders from business as well as IT who are responsible for development, approval, funding, and consulng. The aim is to idenfy the people who can affect and be accountable for realizing the acons necessary for project. Reviewing the combined list of all stakeholders (IT and Business) enables genuine assessment of investment feasibility and find real potenal to add value.

Value of each benefit Idenfied benefits need to be organized based on increasing level of explicitness and knowledge about value of benefit. Every benefit is allocated to Observable row inially. Evidence needs to be provided in order to move the benefit in rows above. Classificaon of benefits in 4 rows:

Risks and relevant migaon Summarizes project risks What risks are involved? What are the consequences of a risk happening? What opportunies may emerge? What plans are in place to deal with the risks?

Types of risks: Internal Risk: a. Financial risk analysis: managing funds for the project, allocang costs within defined business budget, Payback period for the investment, b. Technical risk analysis: Risk of failing due to technical shoralls

Potenal risks to be organized as per different investment life cycle (Before, During and Aer Implementaon) along with Environmental and business case limitaons

Figure 4.6 (continued)

4.4 Revised Business Case Template for it Investment

4.4

51

Revised Business Case Template for it Investment

An selective amalgamation of both reference papers discussed previously has been made considering their best practices. This outcome along with best suitable contents supported by the conventional business case has been used to develop a business case template for IT investment, Figure 4.7 Key take away from literature-based business case for IT investments ✓ In comparison to Conventional business case, business case for IT investment demands a detailed analysis. ✓ Unlike conventional business case, IT business case focuses on both tangible as well as intangible benefits. In fact, there are more intangible benefits resulting from IT investment. And it is increasingly important to concentrate on intangible advantages for IT solutions, as these intangible benefits bring more strategic gains in long term. ✓ IT investment business case also emphasizes on clear identification and segregation benefits, the way it should be measured and its owners (the one who gains from the mentioned benefit and is ready to work within a team to make the benefit realize through personal efforts or through resources and influence) ✓ IT investment business case also goes further to structure and classify the benefits as per the change which are required to realize it under categories of Do new things, Do things better and Stop doing things. ✓ One peculiar characteristics of IT investment business case is identification of Change owners. A change owner is the one who is responsible to recognize the changes necessary to achieve a particular benefit. This helps in structuring certain assurance for the benefit agreed. Additionally, it also provides confidence not only on the investment productivity, but also on how it will be realized. ✓ Categorized Benefits and then classified as per degree of explicitness in 4 rows from bottom row to top as Observable, Measurable, Quantifiable and Financial. Initially the benefit is allocated as Observable and later moved to upper rows depending on the provided evidence. Each next row represents higher explicitness and improved understanding about the worth of benefit. Such a categorization assist is following each benefit closely from initiation to its realization.

52

4

Business Case

Figure 4.7 Business case template for IT investment. (Source: Maes et al. (Cf. 2013: 47– 57); Ward et al. (cf. 2008: 1–15); IBM (cf. 2015: 8); Sawant (cf. 20201: 8–9); Marshall (cf. 2018: 21–25))

4.5 Business Case for Process Mining

4.5

53

Business Case for Process Mining

Process mining today is performed via digital tools therefore, it is considered as an IT investment. A business case content similar to the one displayed in Figure 4.7 can be applied for decision making to implement Process Mining. Additionally, four important “before implementation” aspects (“As-Is” analysis, Technical Feasibility, Data Quality and Financial Feasibility) should be evaluated in order to ensure process mining readiness within targeted business function. The outcome of obtained assessment should be taken into consideration as a part of the business case, that aids to the decision making of successfully evaluating if the targeted organizations current business functions are ready for process mining implementation. Reasons for the assessment of four factors for Process Mining are explained further: 1. “As-Is” analysis Harvard business review in their report highlight that only 30% of digital transformations succeed. The report also revealed that out of $1,3 trillion spent on digital transformation, about $900 billion were wasted (cf. Tabrizi 2019: 1); (cf. ZoBell 2018: 1). One of the main reasons for such failure was that the existing organizational practices and processes were blemished i.e., the state of current process and wrong practices followed in the company did not support successful implementation of targeted digital technology. Considering this failure rate, it signifies the importance to identify and comprehend the “As-Is” situation of processes within an organization before investing in digital technologies. It has following gains: 1. It enables to get an exhaustive overview of how inefficiencies are deep rooted within the function and its associated processes. It supports in evaluating and identifying existing performance issues, bad practices, and inefficiencies. 2. Making an “As-Is” analysis not only strengthens control over baseline for improvement, but also provides a clear picture about existing and potential issues. Unless one doesn’t have a comprehensive overview of existing situation, it is difficult to judge what exactly needs to be improved 3. Additionally, it helps in aligning the focus on core challenges for the process mining implementation. 4. If the concept of GIGO (Garbage In –Garbage Out) is followed, striving for improvements without making an As-Is overview, will lead to results without any constructive output.

54

4

Business Case

Technical feasibility Technical feasibility should be conducted to evaluate if the implementation of the target Process Mining tool is possible with the current technologies used in the organization. It anticipates how much technical risk exists in the implementation and evaluates the degree of compatibility of all involved IT applications and Infrastructure within targeted function for process mining (cf. Campbell 2007: 6). Prior assessment of technical feasibility leads to understanding of practicability problems in early stages of the implementation process, enhancing the implementation success rate by evaluating multiple parameters and ensuring there is no loss of productivity. A technical feasibility study should be relatively quick, economical and should be conducted considering the current and only required IT environment in sufficient detail to develop a “Problem Definition Statement” (cf. Chatty 1998: 341). The result of technical feasibility should intuit the decision makers to either go ahead with further analysis or stop the initiative. Data Quality assessment Coined by IBM programmer George Fuechsel, the principle of “Garbage in, Garbage Out” (GIGO) succinctly summarizes that computer simply process the data that they are fed (cf. Gärtner, Homann 2013: 134). The strategic decision making is based on the results interpreted through the data. Therefore, the term “GIGO” is applicable for process mining since the results of process mining techniques are data oriented (cf. Cruchten 2019: 1). Like any other data-driven approaches the data quality issues are pervasive to Process Mining which can lead to counter-intuitive or misleading decisions. Several authors in the literature Verhulst (cf. 2016: 40–56); Bose et al. (cf. 2013: 3–13); Rozinat (cf. 2016: 1); Aalst (cf. 2016: 144–151) and the Process Mining Manifesto (cf. 2012: 180) have emphasized that to enable effective Process Mining techniques it is crucially important to have a robust foundation of good quality process event data. In order to achieve this, data scientists need to overcome certain challenges and ensure no data quality compromises are made that could affect the quality of process mining results. Hence the assessment of data quality is an important criterion and should be conducted before implementing Process Mining, in order to ensure relevant data is available with sufficient detail and in proper format as required for process mining. In literature there are several known data quality challenges that may severely affect quality of process mining results (The associated data can be viewed in Appendix 15 and Appendix 16 in the electronic supplementary material). The integrated model generated in the next chapter will ensure that the data quality assessment will consider all known data quality issues that are applicable at the raw process event data level.

4.5 Business Case for Process Mining

55

Financial benefit check: Evaluation of a meaningful financial case is given high importance because it determines the financial feasibility of the project and supports decision makers in determining if the investment should be made or be scrapped. Financial aspect within business case for process mining is crucial owing to following reasons: 1) To provide decision makers a complete overview of involved costs to implement process mining within a particular function for certain processes. 2) To enable decision makers, estimate a logical Return on Investment (ROI) for incurred costs. 3) To plan and ensure appropriate allocation of funding for the project within yearly budget. 4) To assist decision makers in anticipating all associated financial risks for process mining implementation. In order to mitigate the financial risk, it is also crucially important to evaluate the quantitative and qualitative (intangible) output brought by an IT investment. The scope of financial benefits for IT tools and application start from the beginning of the implementation phase and last until the final day when the application is no more in use. Hence the duration in which companies usually reap profits within IT investments is comparatively long (cf. Feng et al. 2010: 552). In the next chapter the above mentioned four “before implementation” aspects are explained in detailed as an integrated holistic concept for assessment of process mining readiness within target organization. The goal of the integrated holistic concept is to provide a blue print/guidance to organizations intended to evaluate if Process Mining makes sense to their existing indirect procurement process.

5

Integrated Holistic Concept to Assess Process Mining in Indirect Procurement (Holistic Model)

Process mining brings the best of both worlds being a bridge between data science and process science. Effective value realization through Process mining requires an overall portrait consisting of fundamentals associated with business function, existing systems, their interlinks, received data along with its superiority, and definitely the benefits which can be achieved through intended change. Without making the first step of knowing the depth of water it is not advisable to swim, similarly, initiating a process mining without knowing what really needs to be improved is naive. Research trend also shows business case moving more towards becoming an insignia for holistic approach which includes not only financial data but strategic value also. Figure 5.1 below explains in detailed the four steps in the form of an integrated holistic concept to ensure and prepare the foundation for implementation of value realizing Process mining initiative.

© The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5_5

57

58

5

Integrated Holistic Concept to Assess Process …

Figure 5.1 Overview of before implementation analysis as integrated holistic concept for process mining. (Source: Own illustration)

5.1

“As-Is” Indirect Procurement Process Analysis

In this sub-section the description of the procedure to conduct an “As-Is” analysis is discussed to identify pain points and improvement opportunities in the indirect procurement process before process mining is implemented. This step intends to gather and understand basic company information and the way indirect procurement business function is currently running in the organization. Collection of following information can be helpful in order to understand the overall current situation and interpret opportunities for improvements: 1. Key company facts and figures 2. Indirect procurement process work instruction guideline—The guideline should be considered as a starting point to build foundational understanding of indirect procurement process within the company. It also highlights the way procurement is managed centrally and locally. 3. Organizational structure of indirect procurement—This helps in understanding where the indirect procurement business of the organization is situated and operated from.

5.1 “As-Is” Indirect Procurement Process Analysis

59

4. Material group structure analysis of indirect procurement—This analysis supports in comprehending allocation of different commodities and their associated goods or services. This assist the strategic buyers in merging different goods/services under standardized groups and decide on related strategy. 5. Evaluate indirect global spend overview—It is necessary to understand the distribution of spend across companies worldwide locations. This assists in comprehending the spend emerging from various locations or to find out locations with highest spend. 6. Evaluate ABC-Analysis (suppliers/sub-commodity)—(The associated data can be viewed in Appendix 22 in the electronic supplementary material) 7. Perform indirect total spend analysis: a) Spend trend analysis—Studying a trend of spend across a number of years provides a better understanding of the way spend of the company is evolving. b) Evaluate indirect total spend relation with direct spend—It helps in knowing the distribution and comprehending the efficiency of indirect procurement function. As per industry experts an indirect spend usually accounts for 13.5% to 22% of revenue, depending on the industry. The ratio is calculated by dividing indirect spend by the revenue. (cf. Sourcing innovation 2021: 1) c) Identify addressable and non-addressable spend—It helps in comprehending the correlation of savings with actual spend under control. d) Identify PO and non-PO volume—PO volume represents cost of all the items that went through a systematic chronology of PO creation until the payment is cleared to the suppliers. This gives an estimation of total PO volume that can be analyzed end-to-end after implementing process mining tool and highlights improvement opportunities for the non-PO volume. 8. Evaluate savings performance analysis—For a procurement function “savings” are the most important KPI which signifies the real performance. 9. Identify standard and non-standard indirect process flow variants—Some indirect commodities like the travel management or commodities with framework contracts under indirect procurement, may not follow a standard process flow chronology, and may have an entirely different nature. For example, for travel management the company employees may make an upfront payment

60

5

Integrated Holistic Concept to Assess Process …

for the travel and hotel booking via company credit card. The aim is to identify such different process variants which do not follow a systematic process activity chronology that starts with purchase order creation and end with payment made to the supplier. This is required in order to define the total volume that can be analyzed end-to-end for process mining analysis. 10. Overview of ERP applications in use—Since the raw process event data usually originates from various ERP applications. Identifying all involved ERP system within indirect procurement process is mandatory along with its nature, role, and technical capabilities. This is necessary in order to define the scope of ERP applications that needs to be connected to get a holistic overview of process data. 11. Identify existing known issues and other observations—Subject matter experts and direct business users within the company are the best source of knowledge about existing issues and problems. Interviewing them helps identify and gather constraints affecting operational efficiency. Outsider view with a different perspective should also be considered in order to encompass full view of inefficiencies including the ones un-identified by direct business users.

5.2

Technical Feasibility

The scope for technical feasibility evaluation of this model is to ensure that process mining tool can be efficiently connected to target company’s ERP system and all the indirect procurement process event data relevant to process mining project can be successfully extracted from the data source tables of the ERP systems with ease. Therefore, the “Problem Definition Statement” for technical feasibility of process mining comprises of three core questions shown in Figure 5.2 below, and the determinants of these questions dictate the technical requirements for the technical feasibility evaluation.

5.2 Technical Feasibility

61

Figure 5.2 Overview of technical feasibility assessment. (Source: Own illustration)

1. What needs to be connected? Today’s organizations use varieties of IT applications and have diverse business unit locations, due to which event data may originate from multiple data source of locations. Therefore, understanding where and how the raw process event data

62

5

Integrated Holistic Concept to Assess Process …

is stored is necessary for data scientists to define an appropriate data extraction strategy. 2. How to connect process mining tool to the target source systems? There are a series of software vendors available in the market that provide process mining solutions. Every vendor may differ based on deployment strategy and the mechanism they provide to support the extraction of raw process event data. Most process mining vendors provide multiple forms of deployment strategy and mechanism for extraction of raw event data (cf. Aalst 2016: 330). Since this model does not focus on any specific vendor and discusses concepts on general basis, evaluating this aspect of data extraction mechanism is only meaningful for real-life projects where a process mining vendor is already chosen since the extraction mechanism should be supported by the technology of process mining vendor. A detailed description of various event log data extraction mechanisms supported by process mining technology in general are described below. Deployment strategies: A deployment strategy may be decided in advance before a process mining vendor is selected. Since it is decided based on factors like, data protection and data compliance practices of an organization, followed by costs and efforts. Table 5.1 shows different deployment strategies used for process mining. Table 5.1 Various Process mining application deployment strategies

Source: Own illustration based on van der Aalst (cf. 2016: 330)

5.2 Technical Feasibility

63

The most widely used deployment strategies by organizations are either an “On premise” or the “Cloud” system. Therefore, in order to have a detailed understanding the differentiation between these deployment strategies is illustrated in Table 5.2 based on selected important parameters. Table 5.2 Differentiation between on-premise and cloud deployment strategy

Source: Own illustration based on Mina et al. (cf. 2019: 12f.)

• Extraction mechanisms to get event data: How to extract raw event data from target companies ERP systems depends on the selection of the extraction method. The event data extraction mechanism is nothing, but the ETL process. Process mining technology provides multiple options for the way data can be extracted and are defined in Table 5.3 below:

64

5

Integrated Holistic Concept to Assess Process …

Table 5.3 Different mechanisms used for raw process event data extraction

Source: Own illustration based on Aalst (cf. 2016: 330)

3. Other factors that enable successful implementation of process mining tool Undoubtedly, a prerequisite for applying process mining techniques is to have an appropriate “event log” whose structure is suitable for process mining (cf. Aalst 2015: 3). Event logs cannot be generated automatically. It has to be transformed from its unstructured raw event data format into a structured event log format. This task is performed by data scientists through simple computational programming. Hence, creation of event logs is usually perceived as a pure technical task (cf. Howe et al. 2017: 1473f.). As learnt in the process mining literature, event logs are the starting point for any process mining analysis and without proper event logs process mining is not possible. Consequently, evaluating components of an event log can be defined using “Guiding Principles” proposed by Process Mining Manifesto as extraction and transformation of raw event data into high quality event log is necessary. The decisions taken during this technical feasibility preparatory phase not only has the ability to impact the success rate of process mining implementation but also decides the components for data quality assessment, which is one of the before implementation criteria of the integrated model. In the next section, the technical feasibility model for implementing process mining in indirect procurement is outlined and supported with a step-by-step procedure.

5.2 Technical Feasibility

5.2.1

65

Stepwise Guidance for Technical Feasibility Assessment

The technical feasibility model defined in this thesis is inspired from the approach presented by Jans et al. (cf. 2019: 9–12). The authors focused on two objectives: Objective #1: To develop a stepwise procedural guidance for novice data scientists to create high quality event logs that can lead to the project task goal. Objective #2: Simultaneously evaluate if following a stepwise procedure defined in objective # 1 has positive impacts on the quality of event log created against a random non-stepwise procedure. The proposed stepwise framework by Jans et al. (cf. 2019: 9–12) is defined by collecting best practices implied in various industries (utility services, internet and telephony provider, and chemicals) and different types of projects (operational excellence and financial auditing). Their study results concluded that, when event logs were created by following their stepwise procedure, high-quality event logs were generated, while the other methods which did not follow any stepwise procedural guidance showed complexities in the quality of event log. Therefore, the authors claim that in order to obtain high-quality event logs data scientists should follow a systematic stepwise procedure guidance. The technical feasibility evaluation model described below for the integrated model is inspired by the stepwise procedure for creating high-quality event logs proposed by Jans et al. (cf. 2019: 9–12) and has been slightly adapted by thesis author to make it compatible for the technical evaluation which is the target goal of the integrated model. The only difference is that Jans et al. (cf. 2019: 9–12) have used this approach for practical creation of an event log and this thesis uses their approach/framework to evaluate if high-quality can be generated. This should briefly answer the second question—Which factors enable the successful implementation of an event log? illustrated in Figure 5.2. The stepwise procedure for technical feasibility evaluation outlined below covers four types of technical factors that will define the overall technical feasibility results: 1. 2. 3. 4.

Process activity analysis Data model analysis Data structure analysis Change table analysis

66

5

Integrated Holistic Concept to Assess Process …

Stepwise procedure for technical feasibility is as follows: Step 1: Define an “A-priori model”. Business processes although are divided in different departments based on the individual activities they perform, often these processes are closely interlinked to each other. Hence, in order to understand involvement of such tangled core business processes, an “A-priori model” should be first defined. Doing this not only gives a good understanding of involved core business processes but also helps to identify key milestone activities defined for indirect P2P process of the target company. This makes it easy for a data scientist to ensure all key activities are systematically recorded in the system and their data can be transformed into an event log with no complexities. It also guides data scientists to narrow down their focus on selecting and gathering information about processes, IT applications and data source tables for further analysis and target right stakeholders for collecting required information. Typically, any indirect P2P process comprises of three processes: purchasing, logistics and accounts payable. Figure 5.3 shows a typical “A-priori model” of a typical indirect P2P process using BPMN notations.

Figure 5.3 Typical A-priori model for a Purchase-to-pay process. (Source: Own illustration)

Step 2: Identify key process cornerstones. The key cornerstones are the process activities that take place within different processes identified in step 1. The aim of this step is to gather all kinds of daily process activities (manual and digital) executed by business users that belong to the identified business functions from the “A-priori model”. This is usually done by interrogating direct process business users of the respective business function. The aim is to understand all the relevant process activities that fall within the scope of the organization’s indirect P2P process.

5.2 Technical Feasibility

67

Step 3: Distinguish between manual and ERP system performed activities. The aim of this step is to classify all identified process activities from Step 2 into manual and ERP system (digitally) performed activities. Davenport et al. (cf. 2019: 1) in their business review for process mining claim that organizations benefit best value when most process activities have been digitized (i.e., supported by an IT system). Hence, more ERP systems performed activities imply more detailed Process Mining activities. If parts of a process are executed manually or in analog form e.g., when documents have to be printed out and signed, then process mining cannot determine meaningful data about the entire process. In such cases a lot of relevant data for analysis and evaluation can go missing and no meaningful picture of the complete real process can be obtained (cf. Becker & Buchkremer 2018). Therefore, process activities that are not supported by ERP systems cannot be further used in this technical feasibility model until they are digitized. This marks as a potential feedback for the target company to prioritize the digitization of identified manual activities. Hence, the next step of this model will focus only on identified ERP system activities for further evaluation. Step 4: Identify ERP systems in use The aim of this step is to identify all involved ERP systems where day-to-day process activities are executed. Further each process activity should be linked to its respective ERP system, in which they are performed. This answers to the first question “What needs to be connected”. From a data scientist perspective this step is crucial, as it narrow downs the focus on only those set of event data for extraction that are of strategic relevance for the process mining project. Step 5: Identify if ERP system performed activities from step 3 are recorded in the data source tables of the ERP system If the ERP system performed activities are not systematically recorded in data source tables, that means there is no process data available for those activities in order to apply process mining techniques. In this case, a feedback to the case company is necessary to further investigate the necessity to record such activities in the future, to bring more transparency into the business. Step 6: Identify the process instance document. Although ERP systems support daily business process operations, they are ideally positioned along business document types and not along business processes. Different business processes follow different document types. For example, purchase requisition and purchase order document types belongs to the procurement process while invoice document type to the accounts payable process. Therefore, recorded

68

5

Integrated Holistic Concept to Assess Process …

activities from step 5 should be linked with their relevant business document types in order to match it with cases (cf. Jans et al. 2019: 10). The information related to business document type collected in this step creates a baseline to generate a data model which is the aim of Step 8. Identifying involved document types is necessary because: 1. It is an initial step to further evaluate if the raw process event data under investigation fulfills criteria for an appropriate CaseID granularity level (Technical Challenge 5: (The associated data can be viewed in Appendix 13 in the electronic supplementary material)) required for process mining which is the “line-item level”. This is identified in Step 9. 2. It is also mandatory to evaluate if all involved document types are linked to each other in the data source tables with the help of primary and secondary keys (The associated data can be viewed in Appendix 18 in the electronic supplementary material). Process Mining follow a single object ID, which is represented as a unique “CaseID” in the event log. When document types are linked together for example, an invoice always belongs to a purchase order, and the connection between these two document types needs to be identified, only then a complete end-to-end P2P process activities for each CaseID can be tracked. e.g., Assume invoice table represents each row at invoice line-item number level, then this table should link back its each row with its corresponding PO line-item number. Step 7: Identify key tables. The raw process event data is not collectively available, due to the fact that it is scattered and stored in various data source tables and locations of a company. An SAP system itself uses thousands of tables to store and organize data. The data sources store purchase orders, material movements, invoices, and other activities as entries in various tables that are related to a specific document type in use. Therefore, each document type identified in step 6 will contain its own tables and all activities related to a document type are recorded in their respective tables. Understanding the table names and the type of data it stores is necessary to form a data model structure of the target company (The associated data can be viewed in Appendix 20 in the electronic supplementary material). This step also supports in overcoming process mining challenges like: Challenge 1: Correlation, Challenge 4: Scoping and C1: Finding, Merging and Cleaning Event Data (The associated data can be viewed in Appendix 15 in the electronic supplementary material) as it guides to filter only those tables that are required within the project scope.

5.2 Technical Feasibility

69

Step 8: Create a data model and describe table relationships. In order to transform the process event data from the above identified data source tables into appropriate event log, a logical connection (relationship) between the data source tables should be established. This is to ensure reliable extraction of data and is usually done by creating a data model (relational schema). Data models are built based on business needs. It defines the logical structure of a database, how data is connected to each other and how they are processed and stored inside the system (The associated data can be viewed in Appendix 20 in the electronic supplementary material). Further, different mapping cardinalities are used to define the relationship between data source tables with the help of primary and foreign keys. The aim of this step is to check if: 1. There exists any granularity issues within data source tables (Challenge 5: Granularity, (The associated data can be viewed in Appendix 15 in the electronic supplementary material)) 2. Evaluate if Object ID (“purchase line-order item number” for an indirect P2P process) makes the data extraction possible by linking all other relevant tables that belong to different document types. i.e., to ensure convergence and divergence issues do not exist (The associated data can be viewed in Appendix 19 in the electronic supplementary material). In order to evaluate this, following procedure is recommended: Find the center table and identify the CaseID level: Usually the table that contains purchase order line-item table is the center table for any P2P procurement process. Distinguish tables: Classify all identified tables into transactional data tables and Master data tables. Create a data model and define relationship between tables: Establish relationship (using mapping cardinalities) of identified data source tables from Step 7 with the center table and their header tables. Note: • For Process Mining, the center table will always hold a one-to-one relationship with all other line-item tables and a one-to-many relationship with the change log tables. • All header tables will always hold a one-to-many relationship with their line-item tables.

70

5

Integrated Holistic Concept to Assess Process …

• Moreover, when many-to-one, many-to-many or no relationships are detected between the data source tables, it determines abnormalities between the table connections and impacts the ability of an event log to accurately represent the reality. This situation is known as convergence and divergence of cases. Step 9: Evaluate if identified tables in step 7 record timestamps. Further, all system performed activities should be checked if timestamps are recorded when the activities took place in their respective identified tables. Timestamps are one of the three mandatory fields for an event log without which process mining techniques cannot be applied. Activities with missing timestamps can be potential feedback to the IT department in order to develop timestamp tracking functionality for future. Step 10: Identify target companies data structure. The way data is stored and accessed in an organization may differ significantly and therefore, data scientists should have a good understanding of all locations where indirect process raw event data is generated and eventually stored. The aim is to ensure that data extraction process is executed with ease and for that data scientists need to develop strategies for data extraction in the best possible way. Extraction of discrete data can be time consuming and demands additional efforts and resources. Hence, organizations particularly aiming to implement process mining must anticipate and may prepare collection of data relevant to Process Mining into a central “data warehouse” as a preferred initial step. This can be optional since process mining is also capable to connect with multiple data source systems. Step 11: Evaluate technical feasibility of change logs. Change activities are considered as a challenge since occurrences of any changes on a purchase order increases number of activities in the original process model (cf. Günther et al. (2006): 10). Thus, causing deviations and delays in business operations. In order to realize the effects of change activities on business processes certain technical criterias should be met, and are as described below: 1. The change log tables should record timestamps when the change was executed. 2. The change log tables should store “new values” and “old values”. 3. The center table and other line-item tables will hold a one-to-many relationship with the change log table. 4. Each row of the change log table should explicitly state what was changed. For example: change of price, change of quantity etc. In the SAP change log table

5.2 Technical Feasibility

71

environment shown in Figure 5.4, it is represented in the column FNAME where NETPR indicates change of price and WERKS represent change of plant. 5. The Change Data Capture (CDC) mechanism required for process mining can be expressed using Scenario 1 and Scenario 2 illustrated in Figure 5.4 and Figure 5.5 respectively, using a standard SAP change table example. Note: CDC is the “process of observing all data changes written to a database and extracting them in a form in which they can be replicated to derived data systems” (Dhanushka 2021:1)

Figure 5.4 Scenario 1 of change activity. (Source: Celonis Training)

Figure 5.5 Scenario 2 of change activities. (Source: Celonis Training)

72

5

5.3

Integrated Holistic Concept to Assess Process …

Data Quality Assessment

Data quality is widely used to represent a set of “characteristics” data, such as its accuracy, completeness, consistency, and timeliness (cf. Fu et al. 2017: 3792f.). These characteristics determine different dimensions that data quality can be represented upon. Wang (1998: 58) states “to increase productivity, organizations must manage information as they manage products” and also defines DQ in terms of its “fitnessfor-purpose”. Wynn et al. (cf. 2019: 4), Doyle (cf. 2013: 5), and Ladly (cf. 2016:1) correlate and apply “fit-for-purpose” concept of quality to the data being used in general and within process mining. Taleb et al. (cf. 2016: 1) highlights DQ to be: i. domain related, ii. defined through a set of attributes and iii. relies on measurement and assessment methods. According to Wand et al. (cf. 1996: 86f.), in order to measure and assess if the data is “fit-for-purpose”, data quality requires a multi-dimensional concept where: i. “Each dimension represents some measurable quality property, and ii. no single dimension can adequately assess overall data quality”. “A Data Quality Dimension (DQD) is a recognized term used by data management professionals to describe a feature of data that can be measured or assessed against defined standards in order to determine the quality of data.” DAMA (2013: 3) (International Data Analytics Management Association)

Based on the above specified conception, to define a systematic DQD assessment criterion for raw process event data, following objectives need to be achieved: Objective #1: To identify the most suitable “Big Data” Data Quality Dimensions (DQD), since raw event data represents the forms and features of “Big Data”. Objective #2: Define selected DQD through a framework and define measurement criteria for the assessment of each DQD. Objective #3: Describe how these DQD’s can be applied on raw event data in order to evaluate the data quality.

5.3 Data Quality Assessment

5.3.1

73

DQD for Process Mining

In literature study related to DQD, there are various types of DQD that have been defined and used on “big data” but there is no existence of key data quality dimensions universally agreed amongst DQ experts. DAMA (2013) in total presents 60 different types of DQD for assessing big data quality (Van Nederpelt & Black 2020: 43–112). Existence of such varieties of DQD has always led to a confusion among many researchers within data quality community. Hence, to overcome this confusion DAMA (cf. 2013: 1–13) has proposed six primary DQD for assessment of “big data” so that there can be a common understanding between professionals. This initiative by DAMA was strongly supported by many researchers and other data management professional organizations like EUROPEAN COMMISSION Directorate-General for Informatics (DIGIT) and are the most recognized and widely used DQD for “big data” (DAMA 2013: 1); (cf. DIGIT 2019: 4 f.). Exploring further on the application of DQD for Process Mining, literature search reflected very few results showing two authors that have applied DQD proposed by DAMA on process mining and are as follows: 1. Andrews et al. (cf. 2019: 13f.) in their study successfully applied three DQD out of the six primary DQD proposed by DAMA and are Completeness, Precision (Validity) and Uniqueness for the data quality assessment of raw process event data. 2. Verhulst (cf. 2016: 40–56) also successfully utilized all the six primary DQD proposed by DAMA along with additional DQD to anticipate and understand the superiority of data for Process Mining on event logs. Table 5.4 gives a holistic overview of the “Big-Data” DQD suggested by different scholars for process mining and to understand the rationale behind selection of DQD for this thesis. DQD’s “Believability/Credibility”, “Relevancy”, “Security/Confidentiality”, “Complexity”, “Coherence”, “Representation/format” firstly are additional DQD that do not belong to the six general DQD’s and are not considered further since they are not suitable for early assessment. They have only been proven to be more suitable only for event logs by Verhulst (cf. 2016: 40–56) and not on raw event data. (The associated data can be viewed in Appendix 23 in the electronic supplementary material). Hence, the original six primary DQD defined by DAMA (cf. 2013: 1–13) remain as the basis for data quality assessment of raw process event data, illustrated in Figure 5.6. This completes the objective #1 and to proceed further with

74

5

Integrated Holistic Concept to Assess Process …

Table 5.4 Overview of data quality dimension for process mining

Source: Own illustration based on DAMA (2013: 1–13 ), Andrews et al (2019: 13 f.) and Verhulst (2016: 40–56)

objective #2 the selected six DQD are defined along with its measurement criteria with the help of a framework. During data quality assessment it is crucial to have some sort of judgment indication for the quality of a dimension. To know what kind of scoring system is being used, and the values the score can take, the six primary DQD should be defined using a data quality assessment framework, that defines each DQD through various aspects. The framework used in this thesis is illustrated in Table 5.5, adopted from the DQD framework defined by Doyle et al. (cf. 2013: 8–15) The next sub-section focuses on describing the six primary DQD using the framework from Table 5.5 that supports in being specific about measuring each DQD (Tables 5.6, 5.7, 5.8, 5.9, 5.10 and 5.11).

5.3 Data Quality Assessment

75

Figure 5.6 Data quality dimensions for raw event data assessment applicable to this thesis. (Source: Own illustrations based on DAMA (cf. 2013: 1–13)) Table 5.5 Data Quality Framework used in this thesis

Source: Own illustration adopted from Data Quality Framework as defined by Doyle et al. (cf. 2013: 8–15)

76

5.3.2

5

Integrated Holistic Concept to Assess Process …

Defining the Six Primary DQD

5.3.2.1 Completeness Table 5.6 DQD—Completeness

Source: Ramasamy (cf. 2020: 3), DAMA (cf. 2013: 8) and Verhulst (cf. 2016: 42)

5.3.2.2 Uniqueness Table 5.7 DQD—Uniqueness

Source: Verhulst (cf. 2016: 45) and DAMA (cf. 2013: 9)

5.3 Data Quality Assessment

5.3.2.3 Timeliness Table 5.8 DQD—Timeliness

Source: Verhulst (cf. 2016: 46) and DAMA (cf. 2013: 10)

5.3.2.4 Validity Table 5.9 DQD—Validity

Source: Verhulst (cf. 2016: 47) and DAMA (cf. 2013: 11)

77

78

5

Integrated Holistic Concept to Assess Process …

5.3.2.5 Accuracy/Correctness Table 5.10 DQD—Accuracy/Correctness

Source: DAMA (cf. 2013: 12)

5.3.2.6 Consistency Table 5.11 DQD—Consistency

Source: DAMA (cf. 2013: 13) and Verhulst (cf. 2016: 49)

5.3 Data Quality Assessment

79

Applying data quality assessment on a complete set of raw event data can be a challenge as the entire data set represents characteristics of “Big Data” containing voluminous and rapidly growing data. Therefore, to overcome this challenge and achieve Objective #3 this model suggests to assess the data quality of raw process event data by separating the evaluation on three different levels. The next sub-section defines the data quality assessment levels and how the DQD can be applied for each level.

5.3.3

Data Quality Assessment Levels

Owing to the problem that the raw process event data possesses characteristics and properties of “Big Data”, handling large amount of data with thousands of data set rows even as a sample dataset cannot be an easy task for a data quality assessment. There was also no literature support available to guide how to manage large set of data for evaluation of big data DQD. Therefore, to ease the data quality assessment, the author of this thesis suggests in executing the data quality assessment based on factors that contribute as elements for the event log. Because eventually the final aim is to check if high quality event logs can be generated. The quality of the event log defines the quality of the process mining results. Figure 5.7 illustrates the scope of three proposed data quality assessment levels with respect to the event log elements.

Figure 5.7 Three levels for data quality assessment. (Source: Own illustration)

5.3.3.1 Activity Level This assessment level aims to evaluate if all the system recorded activities identified in the technical feasibility model are recorded timely. Although system

80

5

Integrated Holistic Concept to Assess Process …

performed activities might show timestamps in the data source tables but this does not justify that the timestamp is accurate or recorded timely as soon as the process activity was executed, the nature of timestamp recording can be different. For example, there can be a substantial gap between the date and time when goods were received in the warehouse to the date and time its entry was updated in the system. Hence, only the existence of a valid timestamp is not sufficient for process, its “timeliness” is more important, as such problems may lead to faulty case durations, impacting the process model itself. (cf. Rozinat 2016: 1). Table 5.12 illustrates an example template for assessing the timeliness of the activities. The DQD is rated “True” if the timestamp represents the actual date and time when the activity was executed or else it is rated “False” if time gap is observed. The IT experts are usually aware of delayed timestamp recording in the data source tables. Table 5.12 Template for DQ assessment at activity level

Source: Own illustration

5.3.3.2 Timestamps Level Timestamps are one of the mandatory elements of an event log and the “As-Is” process discovery is dependent on the quality of timestamps recorded. Therefore, if the timestamps are incorrect or not detailed enough, it is difficult to create the correct order of current “As-Is” (cf. Rozinat 2016: 1). Hence, the aim at this level is to check if process activities recorded in ERP systems have systematic timestamps in appropriate format as required for Process Mining. Table 5.13 illustrates method to assess data quality at timestamp level. Two time-stamp criteria’s have been considered: Transactional table timestamps and change log table timestamps. These criteria’s are assessed with only three DQD: Completeness

5.3 Data Quality Assessment

81

(no empty fields), Validity (timestamp format) and Consistency (similar purchase order activity shows equal timestamps in all the data source systems it gets stored.) Table 5.13 Template for DQ assessment at various timestamp level

Source: Own illustration

The other remaining four DQD (Accuracy, Uniqueness, Timeliness and Consistency) are not suitable at this level since the criteria’s cannot be judged for timestamps. Moreover, the aim is to only assess if timestamps in the given data set are not missing and are recorded in right format as required for process mining.

5.3.3.3 Meta-attribute Level Meta-attributes of every data-source tables provide additional business information about each activity and are useful for root cause analysis in process mining. These meta-attributes are mostly helpful for applying the fourth process mining technique “operational support” and relates to the “organizational case perspective”. Figure 5.8 illustrates different meta-attributes (business aspects) required for process mining in a P2P process. Table 5.14 provides a description of standard indirect procurement metaattributes that should be recorded by the target company’s data source tables and are required for data quality assessment at this level. From a data quality perspective, the existence of these meta-attributes within the data source is vital.

82

5

Integrated Holistic Concept to Assess Process …

Figure 5.8 Questions answered with the help of meta-attributes. (Source: Sievo (cf. 2021: 1))

DQD “Completeness”, “Uniqueness”, “Accuracy” and “Consistency” can be applied at meta-attribute level. “Completeness” checks if fields in the columns do not have null values, “Accuracy” checks if data represents to “real-life”, “Consistency” checks if meta-attribute data is similar across multiple data source systems of the organization and “Uniqueness” checks if there are any data redundancies in the master data tables. The values of the meta-attribute columns mostly originate from the master data tables. For example, a vendor name may be recorded multiple times and therefore would be represented in the vendor master data table differently. For instance, vendor “ABC corporation” may be also stored as “ABC ltd”, creating two profiles of the vendor, this will lower the results for DQD “Uniqueness” and may lead to false vendor performance analysis within process mining (Table 5.15).

5.3 Data Quality Assessment Table 5.14 List of standard indirect procurement process meta-attributes

Source: Own illustration based on SAP documentation Table 5.15 Template for Data quality assessment at meta-attributes level

Source: Own illustration

83

84

5

Integrated Holistic Concept to Assess Process …

5.4

Financial Case

5.4.1

Financial Measurement Method

In literature there are several measures discussed for calculating the investment benefits related to monetary value mainly, the Internal Rate of Return (IRR), Net Present Value (NPV), payback period, Return on Investment (ROI), Return on Assets (ROA) etc. Botchkarev et al. (cf. 2011: 247) claim that most of these financial measures are all Return on Investment (ROI) related measures and can be collectively known as “ROI analysis”. ROI as an individual measure has been seen as one of the most popular and widely used measure when it comes to evaluating the monetary benefits related to any investment of IT tools and applications in literature as well as in real life. ROI measurement basically compares the cost of doing the project to the financial benefits gained by the project. To evaluate ROI for information systems, Botchkarev et al. (cf. 2011: 246) proposed the following formula:

The output of the above ROI formula can either be a numeric value or can be expressed in percentage. For instance, if gain on investment (total return) is e200 and the cost is e100 then the ROI would have a 1, or 100% when expressed as a percentage. In this proposed financial model, other ROI-related measures are not included, and the focus is only on the ROI formula stated above for primary scope of work. The next scope of discussion is to understand how the ROI calculation metrics (Gain on investment and cost of investment) should be decided for process mining projects. Estimating hidden costs is more challenging than calculating upfront costs. Hence in order to examine a real financial impact of implementing a new technology within business requires consideration of a Total cost of ownership (TCO) aspect. This provides a wider understanding of investment and resulting value realization over lifespan. TCO which is often underrated should consist of factors like organizational cultural change costs, management and support, communications, end-user expenses, opportunity cost of downtime and cost arising from productivity losses (Gartner).

5.4 Financial Case

5.4.2

85

Cost of Investment

According to Botchkarev et al. (cf. 2011: 251) the cost contribution involved in calculating the total cost of investment for IT investments can be divided in three groups, mainly the IT Infrastructure, labor, and training costs. Table 5.16 describes these three cost groups by listing its various cost components. As this master thesis focuses on pre-implementation phase, only upfront foreseeable costs suggested by Botchkarev et al. (cf. 2011: 251) have been considered for cost of investment. Calculating TCO which could be subjective to respective organization requires clarity and can be estimated when companies wisely choose to improve the improvement opportunities realized through “As-Is”, technical feasibility and data quality assessment. For example, cost of organizational changes like integration of a “data warehouse”, are dependent on the decision taken by the top management of the company. Therefore, this thesis has considered costs that are directly related to the Process Mining tool. Table 5.16 Typical ROI cost components for IT investment

Source: Own illustration, adapted for process mining costs, based on Botchkarev et al. (cf. 2011: 251)

86

5

Integrated Holistic Concept to Assess Process …

Based on the cost components defined in Table 5.16 a template for calculating the total cost of investment, specific to implementation of process mining is created (shown in Figure 5.9). The template is divided in two scenarios based on the deployment strategy the target company chooses to implement. Scenarios 1: Covers costs involved for cloud subscription –Software as a Service (SaaS). Scenario 2: Covers costs involved in for on-premises solutions

5.4.3

Gain on Investment

Owing to the uncertain nature and dependency of financial benefits on multiple factors which are not straightforward and easy to be defined in the pre-study or before implementation phase, the evaluation of financial benefits that can be achieved through implementation of process mining tool poses a big challenge. Literature based research has also not reflected much on calculating accurate hard monetary savings resulting from process mining initiatives. Hence, this thesis refers to the financial saving benefit calculation method provided by industry expert consultants for process mining. The metrics defined for the savings calculation originates from best practices and industry benchmark values defined by market leader “Celonis”, through their customer experiences. As the Celonis proposed savings calculation formulas are specific to P2P process for before implementation it supports the pre-study phase of this model. The calculations are based on metrics like the total PO item count/year, total PO spend per year, Full-Time Employee (FTE) costs per minute etc., that can be easily available from the target company and Celonis defined industry benchmarking values. The total savings benefit calculation shown in Figure 5.10 is influenced by four factors—“Employee productivity”, “Purchasing spend”, “Inventory Holding cost” and “Working Capital”. Each of these factors are the representation of the total benefits achieved by its sub-factors. Figure 5.10 illustrates a clear view on the calculation method proposed by Celonis GmbH, that forms the overall financial benefit calculation for the financial case of this thesis.

5.4 Financial Case

87 Process Mining Enablement Cost Esmaon Scenario 1: Process Mining Cloud Subscripon Fees (SaaS) Type Indicave Pricing

1 Process Mining Subscripon fees/month

Monthly fees €

-

Subscripon fees 2 Year 1 subscripon fees 3 Year 2 subscripon fees 4 Year 3 subscripon fees

€ € €

Implementaon costs 5 Implementaon service cost : Year 1 6 Implementaon service cost : Year 2 7 Implementaon service cost : Year 3

€ € €

-

Expert support internal/external (analyse and realize savings) 8 Expert support: Year 1 9 Expert support: Year 2 10 Expert support: Year 3

€ € €

-

Total projected costs for 3-years subscripon

-

€

-

Scenario 2: Process Mining On Premise Costs Type

Annual Fees - €

Indicave Pricing 11 Year 1 licencing fees 12 Year 2 licencing fees 13 Year 3 licencing fees

€ € €

-

14 Year 1 maintenance costs 15 Year 2 maintenance costs 16 Year 3 maintenance costs

€ € €

-

€ €

-

Years # 3

17 Total years 18 Total Licencing fees 19 Total Maintenance fees Projected cost for 3-Years On Premise Costs

€

-

20 Cost of employee training per training group

€

-

Trainee # 21 Year 1 Training Session 22 Year 2 Training Session 23 Year 3 Training Session Projected cost for 3-Years employee training

Amount € € €

-

€

-

Implementaon costs by service provider 24 Implementaon service cost : Year 1 25 Implementaon service cost : Year 2 26 Implementaon service cost : Year 3

€ € €

-

Expert support internal/external (analyse and realize savings) 27 Expert support: Year 1 28 Expert support: Year 2 29 Expert support: Year 3

€ € €

-

Projected cost for one me implementaon and addional resource requirements

€

-

Total projected costs: Year 1 Total projected costs: Year 2 Total projected costs: Year 3 Total Projected cost for 3-years subscripon

€ € € €

-

Projected cost for Scenario: 2 Total projected costs: Year 1 Total projected costs: Year 2 Total projected costs: Year 3 Total Projected cost for 3-years subscripon

€ € € €

-

Projected cost for Scenario: 1

Figure 5.9 Template for cost of investment calculation for process mining implementation. (Source: Own illustration)

88

5

Integrated Holistic Concept to Assess Process …

Since the aim of this financial case is to calculate ROI for three consecutive years, the total potential savings obtained from each sub-factors shown in Figure 5.10 will require certain percentage assumptions that will split the total

Figure 5.10 Gain on investment calculation method. (Source: Own illustration, savings calculation method proposed by Celonis)

5.4 Financial Case

89

Figure 5.10 (continued)

potential savings benefits meaningfully. Table 5.17 illustrates the assumption values considered for this thesis. These assumption values shown in percentages originates from the opinion of process mining industry experts. Table 5.17 Savings benefit assumptions for the first three years Business Impact Manual rework Automaon Electronic orders Employee Producvity Free-text requisions Order bundling Three-way matches Material prices Purchasing Spend Contract usage Early deliveries Inventory Holding Cost Late deliveries Early deliveries Working Capital Late deliveries

Year 1

Year 2

Year 3

Assumpon of realizaon Assumpon of realizaon Assumpon of realizaon 40% 70% 100% 10%

30%

100%

40%

90%

100%

40%

100%

100%

50%

100%

100%

30%

80%

100%

50%

90%

100%

50%

100%

100%

20%

50%

100%

20%

50%

100%

40%

60%

100%

40%

60%

100%

Source: Assumption list provided by industry experts

90

5

Integrated Holistic Concept to Assess Process …

Limitations: • The scope of calculating financial benefits for process mining may be broad. However, due to lack of literature knowledge on how to calculate a complete financial gain on investment for process mining, only methods proposed by Celonis have been utilized. Savings that are currently visible for before implementation phase are only considered. If there are any additional savings perspectives that could be realized at the time of during and after implementation phase, then such financial benefits are not considered due to lack of visibility in the literature.

5.4.4

Intangible Benefits

Although cost is a crucial factor while investing in any business tools, few authors and researchers have shown different view on calculating the ROI for investments related to IT. Remenyi et al. (cf. 2007) highlights that when determining benefits from digital information, in most cases it creates intangible benefits and are hard to measure, particularly in financial terms. Hall (2003: 1) points his view as “The best, most innovative IT improvements have no ROI. There was no decent ROI on installing the first Wang word processor in the 1970s or the first PC to run VisiCalc in the 1980s or the first Linux server for corporate Web sites in the 1990s. If we would have let the ROI spectre rule the day, this decade would have never seen an analogue to the technological achievements of past decades.” A similar view was shared by Gartner’s report for IT research and advisory firm suggesting that, “cost should be a secondary consideration to the achievement of business benefits” (Hoppe 2019: 1). Chareonsuk et al. (cf. 2010: 1095f.) are on the same track and supports the idea that earning intangible benefits today are becoming more crucial for any organization as a supporting mechanism for generating direct revenue within the value creating process. The authors claim this fact based on the evidence on how modern organizations today have integrated intangibles throughout their value chain to aid value creation through timely and accurate decision making. The examples stated above convey the message that focusing more on the functionalities and intangible benefits will be beneficial in the long run than focusing just on the primary costs for any IT or digital technology investment on short term. The indirect strategic advantages achieved via digital technologies are beneficial for the entire organization from a long-term perspective.

5.4 Financial Case

91

This model suggests that having extreme views of either aspects only monetary or intangible benefits are not advantageous. Companies should strike a balance between the monetary benefits and intangible values depending on the company strategy. For example, if implementing process mining software for a P2P process within indirect procurement can enable timely payment for the suppliers, intangible benefits like improved supplier relationship, robust partnership with suppliers, virtuous on-time-delivery and quality of supplies can be outweighed over direct return on investment. Therefore, the financial model proposed in the integrated holistic concept emphasizes to consider both qualitative (intangible) and quantitative (tangible) factors for evaluating benefits.

6

Final Remarks

6.1

Conclusion

Digitalization today has become the need of hour and is no more a matter of choice. In order to secure a viable competitive edge within market and fortify customer loyalty it has become crucially important for companies to make digitalization as core part of their corporate strategy. Furthermore, companies truly need to live this devised strategy by implementing new digital means. Statistics by HBR display a higher failure rate to carry out digital transformation mainly due to mindset and flawed existing business processes. Mindset change to allow acceptance of digitalization as the new evolution can be achieved at outset by recognizing the concerns of employees and preparing them in the light of upskilling for the imminent transformation. Flawed business practices can be improved by analyzing and addressing them with a perspective of removing wastes. Process mining can potentially play a vital role in improving “As-Is” situation by removing wastes on a real time basis and preparing a baseline for digital transformation of a company. Business processes being the pulse of any organization portray the firm’s ability to successfully serve their customers. Hence these processes should be tidy to allow efficient and effective operational performance. Process mining being a promising technique could be seen as a potential solution to enable elimination of complexities, inefficiencies, and wastes. And consequently, can be used as a stepping stone for the superior digital transformation via cutting edge technologies like AI, ML, IoT, and data analytics etc. Process mining being an intelligent technology for enhancement of a company´s execution capacity strikes a perfect © The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5_6

93

94

6

Final Remarks

balance by bridging and providing the best of data mining and BPM. Hence it poses as a technology for present and future with immense potential to empower organizations through enormous opportunities in Procure to Pay, Order to Cash, Resource planning and processes driving HR, Finance and Customer services, production etc. Companies are recognizing this immense potential and including process mining as a new IT innovation to achieve real time improvements. In order to decide or choose right IT innovations mostly companies use business case approach in order to decide for new initiatives or innovations improvements. Pre-implementation analysis is extremely crucial to secure successful execution of new technology which can reap not only good ROI, but more than that significant intangible benefits. This thesis found that business case should not only be considered as a tool to get budget allocation or project approval, but as a means of pursuing a project successfully until its completion. Likewise, this thesis reasoned that conventional business case template is insufficient to cover all aspects necessary for deciding IT Investment. Literature review discovered that for virtuous value realization, IT investments should additionally focus on detailed synopsis of benefits, benefit owners, resulting change and respective change owners. This thesis argues that it is particularly vital to carry out a pre-implementation analysis to safeguard successful execution and noteworthy return on investment. Conventional business case was found to emphasize more on financial benefits in comparison to IT investment business case which gives more importance to intangible advantages. This thesis claims that it is necessary to strike an appropriate balance between tangible and intangible benefits considering the long-term perspective for any IT investment. Implementation of process mining has remarkable intangible benefits which outweigh even the financial advantages. Integrated holistic concept should be used along with business case to secure successful implementation of process mining within Indirect procurement. This thesis addressed the creation of a business case template for IT investment in order to judge and make precise decisions. In particular a sequential blueprint considering holistic view required to explore the feasibility of implementing Process mining within Indirect Procurement was developed. “As-Is” analysis, Technical feasibility check, data quality assessment, and financial feasibility check were found as the most important pre-implementation stages. Along with the business case template for IT investment, aforementioned four aspects should be considered as mandatory steps to ensure flawless process mining execution.

6.1 Conclusion

95

Creation of business case and guideline prepared for holistic assessment of process mining within indirect procurement was achieved through systematic literature review regarding process mining, indirect procurement, digitalization, and business case. Validation of developed model was ensured through its application to case company’s indirect procurement function using empirical findings. In general, IT investments demand considerable number of resources and time. If done wrongly, it leads to monetary losses, loss of management confidence in initiative and loss of employee morale. It could also adversely affect time to market, consequently affecting company´s competitive edge. This developed integrated holistic concept will allow the aspirants of process mining implementation to follow a step-by-step guide for impeccable implementation within indirect procurement as a starting point for the initiative. Companies thriving in digital economy today need a guideline which enables timely decision making and tap the unexplored potential of process mining. The integrated holistic concept developed as a part of this thesis will enable company functions to get better overview of return on their investment, long term intangible benefits and help them anticipate the related risk and consequent mitigation at an early stage within indirect procurement. Process mining is a boon to secure sustainable growth and retain competitive advantage in modern times. Optimum usage of enormous available data, continuous improvement as a part of DNA and digital transformation are key factors for any company to thrive in digital economy and be an integral part of current digital revolution. Many companies tend to concentrate more on their “To-be” or desired state. And in this attempt, they ignore/forget to work on their “As-Is” state. This way doesn’t prove fruitful and leads to failure in achieving any improvements. Hence literature emphasizes that it is very important to focus first on improving “As-Is” state to achieve sustainable enhancements. Author of this thesis is convinced that process mining technology is a boon to eliminate wastes and achieve desired improvements. Hence it can be used as a catalyzer to prepare baseline for successful digital transformation. Not only does it assist in reducing the time to realize value, but it also acts as a sustainable source of continuous improvement to increase productivity. Additionally, process mining contributes to lot of intangible benefits like reduced throughput times, optimized process flows, improved transparency, enhanced compliance and above all enriched customer experience and consequent virtuous relationship. With such remarkable benefits, shouldn’t companies make use of the enormous potential of Process Mining?

96

6.2

6

Final Remarks

Limitations

➢ Current literature in process mining provides very less and insufficient direction on how to avoid implementation crisis. Guidance for evaluating the technical feasibility of the target companies source system and data quality assessment at early stages of process mining project is very less discussed in the literature. Hence, the technical feasibility and data quality assessment guidance developed in this thesis is solely based on the authors technical experience and process mining knowledge gained through Celonis process mining implementation training. ➢ The viability of the developed integrated holistic concept to check readiness for process mining will be guaranteed after its application in various organizations. ➢ Due to lack of literature and academic support to calculate financial benefit (savings) for process mining, this thesis has used industry benchmarking and best practices calculation metrics defined by process mining vendor “Celonis”. ➢ Costs, efforts, human resource and time are required in order to apply this business case model.

6.3

Future Work

➢ The life cycle of a business case consists of three stages: Initialization, Realization and Evaluation. This master thesis focused on developing an integrated holistic concept for initialization (before implementation) phase. Further research work is required to explore the realization and evaluation phases for process mining projects. ➢ The developed integrated holistic concept can further be extended for developing guidance in relation to other business processes like: Order-To-Cash, Production etc. ➢ Prepared business case template for IT investment does not have TCO aspect in detailed because it was limited to pre-implementation phase of process mining. TCO is a vast concept and subjective to organizations. Preparing a model that considers and calculates the factors influencing TCO for cloud and on-premises solutions could be potential future task.

References

Aalst, van der, W. M. P., Adriansyah, A., Alves De Medeiros, A. K., Arcieri, F., Baier, T., Blickle, T., Bose, J. C., Brand, van den, P. C. W., Brandtjen, R., Buijs, J. C. A. M., Burattin, A., Carmona, J., Castellanos, M., Claes, J., Cook, J., Costantini, N., Curbera, F., Damiani, E., Leoni, de, M., ... Wynn, M. T. (2012). Process mining manifesto. In F. Daniel, K. Barkaoui, & S. Dustdar (Eds.), Business Process Management Workshops (BPM 2011 International Workshops, Clermont-Ferrand Ailenei I., Rozinat A., Eckert A., van der Aalst W.M.P. (2012) Definition and Validation of Process Mining Use Cases. In: Daniel F., Barkaoui K., Dustdar S. (eds) Business Process Management Workshops. BPM 2011. Lecture Notes in Business Information Processing, vol 99. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-28108-2_7 Al-Mudimigh, A., Zairi, M., Al-Mashari, M. and others. (2001) “ERP software implementation: an integrative framework”, European Journal of Information Systems, Vol. 10, No. 4 Apromore 2021. Estimate your return on Investment for process mining capabilities. Online available : Azevedo, A., & Santos, M. (2008). KDD, SEMMA and CRISP-DM: a parallel overview. IADIS European Conf. Data Mining. Pg 183–185 Barry, J., Cavinato, J. L., Green, A., & Young, R. R. (1996). A Development Model of Effective MRO Procurement. International Journal of Purchasing and Material Management, 35–44. https://doi.org/10.1111/j.1745-493X.1996.tb00284.x Bechtel, C., & Patterson, J. L. (1997). MRO Partnerships: A Case Study. International Journal of Purchasing and Materials Management, 33, 18–23. https://doi.org/10.1111/j.1745493X.1997.tb00027.x Behnam Tabrizi, Ed Lam, Kirk Girard, Vernon Irvin. 2019. Digital Transformation is not about technology. Harvard Business Review. Change Management. Online Available

Bernd Gärtner, Michael Homann. (2013) INFORMATIK fur Mathematiker und Physiker Eine Einführung in C++ Skript zur Vorlesung 252–0847–00 Herbstsemester 2013, ETH Zürich1 pg: 134. notes.pdf (ethz.ch) Becker, M., & Buchkremer, R. (2019). A practical process mining approach for compliance management. Journal of Financial Regulation and Compliance.nther © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Fachmedien Wiesbaden GmbH, part of Springer Nature 2023 V. Pawar, Holistic Assessment of Process Mining in Indirect Procurement, BestMasters, https://doi.org/10.1007/978-3-658-41453-5

97

98

References

Black Andrew, Nederpelt Van Peter. (2020). Dictionary of dimensions of data quality (3DQ). Dictionary of 60 Standardized Definitions . Online Available < http://www.dama-nl.org/ wp-content/uploads/2020/11/3DQ-Dictionary-of-Dimensions-of-Data-Quality-version1.2-d.d.-14-Nov-2020.pdf> Borek Alexander, Ajith Kumar Parklikad, Jela Webb, Philip Woodall. (2014). Total Information Risk Management: Maximizing the Value of Data and Information Assets Paperback Bose R. P. J. C., R. S. Mans and W. M. P. van der Aalst. 2013. „Wanna improve process mining results?,“ IEEE Symposium on Computational Intelligence and Data Mining (CIDM), 2013, pp. 127–134, doi: https://doi.org/10.1109/CIDM.2013.6597227. Botchkarev, A., & Andru, P. (2011). A Return on Investment as a Metric for Evaluating Information Systems: Taxonomy and Application. Interdisciplinary Journal of Information, Knowledge, and Management, 6, 245–269. Boulaye Pierre de la, Mauro Erriquez, Manuel Gener Bago, Patricio Ibanez, Raul Santos, and Alfredo Vaghi. 2019. Revolutionizing indirect procurement for the 2020s. Mckinsey & Company. Online available: Bruch, H. and Ghoshal, S. (2002) “Beware the Busy Manager”, Harvard Business Review, Vol. 80, No. 2, Burt, D. N., D. W. Dobler and S. Starling (2003). World Class Supply Management: The Key to Supply Chain Management. Boston, McGraw-Hill/Irwin. Campbell J. (2007) Why a feasibility study? University of Toronto. Microsoft PowerPoint— L3-part2-feasbility.ppt (toronto.edu) Carr, A. S. and L. R. Smeltzer (1997). „An Empirically Based Operational Definition of Strategic Purchasing.“ European Journal of Purchasing and Supply Management 3(4) Celonis customer stories. 2021. Chaffey, D. 2011. E-Business & E-commerce management: Strategy, implementation and practice. 5 th edition. Essex: Pearson Education Limited Chareonsuk, C. and Chansa-ngavej, C. (2010). Intangible asset management framework: an empirical evidence. Industrial Management & Data Systems, vol. 110, no. 7, pp. 1094– 1112. Chatty Stéphane, Dewan Prasun. (1998) Engineering for Human-Computer Interaction. I FIP TC2/TC13 WG2.7/WG13.4 Seventh Working Conference on Engineering for HumanComputer Interaction September 14–18, 1998, Heraklion, Crete, Greece Cousins. (2021). The Impact of Digitalisation on Data Professionals. Online Available < https://oguk.org.uk/wp-content/uploads/2021/07/OGUK-The-Impact-of-Digitalisationon-Data-Professionals.pdf> Cox, A., Chicksand, D., Ireland, P., & Davies, T. (2005). Sourcing Indirect Spend: A Survey of Current Internal and External Strategies for Non-Revenue Generating Goods and Services. Journal of Supply Chain Management, 41, 39–51 https://doi.org/https://doi.org/10. 1111/j.1055-6001.2005.04102004.x Cruchten, R.V. (2019). Data Quality in Process Mining: A Rule-based Approach. D. A. Piessens. (2011) “Event log extraction from SAP ECC 6.0,” M.Sc. thesis, TU/e, Eindhoven, Netherlands. Dakic, D[usanka]; Stefanovic, D[arko]; Cosic, I[lija]; Lolic, T[eodora] & Medojevic, M[ilovan] (2018). Business Process Mining Application: A Literature Review, Proceedings of the 29th DAAAM International Symposium, pp.0866–0875, B. Katalinic (Ed.),

References

99

Published by DAAAM International, ISBN 978- 3–902734–20–4, ISSN 1726–9679, Vienna, Austria DOI: https://doi.org/10.2507/29th.daaam.proceedings.125 Davenport H. Thomas, Spanyi Andrew April 2019. What Process Mining Is, and Why Companies Should Do It. Harward business review. What Process Mining Is, and Why Companies Should Do It (hbr.org) de Boer, L., Holmen, E., & Sitar, C. P. (2003). Purchasing as an Organizational Design Problem: The Case of Non-Product Related Items and Services. Management Decision, 41(9), 911–922. Deloitte. (2021) The digital business case done right. Online Available https://www2. deloitte.com/de/de/pages/operations/articles/digital-business-case.html Retrieved on 30.09.2021 Dhanushka Dunith. 2021 A Gentle Introduction to Event-driven Change Data Capture. Online : A Gentle Introduction to Event-driven Ckamberge Data Capture | by Dunith Dhanushka | Event-driven Utopia | Medium DIGIT (Directorate-General for Informatics). (2019) EUROPEAN COMMISSION. SEMIC Study on data quality management.pdf (europa.eu) Dilmegani Cem. (2020). Process Mining Statistics: Market Size, Adoption Online available: < https://research.aimultiple.com/process-mining-stats/> Retrieved on: 01.09.2021 Dumas, M.; La Rosa, M.; Mendling, J.; Reijers, H. A. (2013): Fundamentals of Business Process Management. Wiesbaden: Springer Ellram, L. M. (1994). Total Cost Modelling in Purchasing. Tempe AZ, Center for Advanced Purchasing Studies. Everest Group. Amardeep Modi, Harpreet Kaur Makan, Santhosh Kumar. Process Mining – Technology Vendor Compendium 2021. Service Provider Compendium. [online] Available at: Everest Group. Process Mining Products PEAK Matrix ® assessment. Online Available < https://www.celonis.com/blog/celonis-a-leader-in-process-mining-assessment/> Farbey, B., Land, F. and Targett, D. (1999) “Moving IS evaluation forward: learning themes and research issues”, The Journal of Strategic Information Systems, Vol. 8, No. 2 Fayyad Usama, Piatetsky-Shapiro Gregory, Smyth Padhraic. 1996. From Data Mining to Knowledge Discovery in Databases. Feng, Jingchun, Fujie Zhang, and Lei Li. 2010. “Identification and Calculation Method of the Financial Benefits of IT Projects for Better Financial Evaluation.” Intelligent Information Management 2:552–58. doi: https://doi.org/10.4236/iim.2010.29063. Fortune Business Insights. 2020. Process Mining software market. Online Available https:/ /www.fortunebusinessinsights.com/process-mining-software-market-104792 Retrieved on: 01.10.2021 Foster Provost, Tom Fawcett. 2013. Data Science and Its Relationship to Big Data and DataDriven Decision Making. Data Science for Business. DOI:https://doi.org/10.1089/big. 2013.1508 Fox, F, Aggarwal, VR orcid.org/0000–0003–0838–9682, Whelton, H et al. (1 more author) (2018) A Data Quality Framework for Process Mining of Electronic Health Record Data. In: 2018 IEEE 6th International Conference on Healthcare Informatics (ICHI). IEEE ICHI 2018, 04–07 Jun 2018, New York, NY, USA. IEEE , pp. 12–21. ISBN 978–1–5386– 5377–7

100

References

Frawley, W. J., Piatetsky-Shapiro, G., & Matheus, C. J. (1992). Knowledge Discovery in Databases: An Overview. AI Magazine, 13(3), 57. https://doi.org/https://doi.org/10.1609/ aimag.v13i3.1011 Fu, Q., & Easton, J. M. (2017). Understanding data quality: Ensuring data quality by design in the rail industry. In 2017 IEEE International Conference on Big Data (Big Data). https:/ /ieeexplore.ieee.org/abstract/document/8258380 Gartner. 2021. Gartner Glossary. Digitaliztaion. Online Available < https://www.gartner. com/en/information-technology/glossary/digitalization> Retrieved on 30.09.2021 Ghosh, Rajarshi, Strategies in the VUCA World (July 2, 2020). Online available < delivery.php (ssrn.com)> Graafmans, T., Turetken, O., Poppelaars, H. (2020) Process Mining for Six Sigma. Bus Inf Syst Eng 63, 277–300 (2021). https://doi.org/10.1007/s12599-020-00649-w Gunasekaran, A., & Ngai, E. W. T. (2008). Adoption of e-Procurement in Hong Kong: An Empirical Research. International Journal of Production Economics Gunasekaran, A., McGaughey, R. E., Ngai, E. W. T., & Rai, B. K. (2009). E-Procurement Adoption in the Southcoast SMEs. International Journal of Production Economics, 122, 161. Günther, C., Rinderle-Ma, S., Reichert, M., & Aalst, W. (2006). Change Mining in Adaptive Process Management Systems. OTM Conferences. Hall Mark. 2003. Forget ROI. Online available https://www.computerworld.com/article/257 9489/forget-roi.html Hamblin Mark. (2020). Direct vs. Indirect Materials in Procurement: What’s More Challenging? Online Available Retrieved on: 02.09.2021 Hammer, M. and Champy, J. (1993) Reengineering the Corporation: A Manifesto for Business Revolution. Harper Collins, New York Han, J., Kamber, M. and Pei, J. (2011) Data Mining: Concepts and Techniques. 3rd Edison, Morgan Kaufmann Publishers, Burlington. Hand, D. J.; Smyth, P.; Mannila, H. (2001): Principles of Data Mining. In: The MIT Press, Vol. 30 (2001), No. 7, pp. 621–622. Henderson Bruce D. (1975). The Coming Revolution in Purchasing. Journal of Purchasing and Materials Management. https://doi.org/10.1111/j.1745-493X.1975.tb00330.x Hind R’bigui; Chiwoon Cho. 2017. The state-of-the-art of business process mining challenges, International Journal of Business Process Integration and Management (IJBPIM), Vol. 8, No. 4, DOI:https://doi.org/10.1504/IJBPIM.2017.088819 Pg 1 Hitt, L., Wu, D. and Zhou, X. (2002) “Investment in enterprise resource planning: Business impact and productivity measures”, Journal of Management Information Systems, Vol. 19, No. 1 Hoppe Geoff. 2019. What Does Business Intelligence Software Cost? Online available Howe, Bill, Michael Franklin, Laura Haas, Tim Kraska, and Jeffrey Ullman. 2017. “Data ScienceEducation: We’re Missing the Boat, Again.” In Data Engineering (ICDE), 2017 IEEE 33rdInternational Conference on, 1473–1474. IEEE. IBM Cloud Education 28 April 2020. Data Modelling. What is Data Modeling? | IBM

References

101

IBM. (2015). Why a business case for change management. Online Available < https://www. ibm.com/downloads/cas/YPAPB0A2 > Retrieved on 30.09.2021 Iloranta Kari, Pajunen-Muhonen Hanna. 2012. Hankintojen johtaminen—ostamisesta toimittajamarkkinoiden hallintaan. 4 th revised edition. ISBN: 978–951–885–388–9 Israel, D. and Curkovic, S. (2020) Indirect Procurement: A Literature Review and Study of Trends. American Journal of Industrial and Business Management, 10, 775–792. doi: https://doi.org/10.4236/ajibm.2020.104052. Online Available < https://www.scirp. org/journal/paperinformation.aspx?paperid=99610> J.C.A.M. Buijs. (2010) Mapping Data Sources to XES in a Generic Way. Master’s thesis. Eindhoven University of Technology. Online available Jans, Mieke; Soffer, Pnina & Jouck, Toon (2019) Building a valuable event log for process mining: an experimental exploration of a guided process. In: Enterprise Information Systems, 13 (5), p. 601–630 Kumar Nand, Soni Rohan. 2017. ABC Analysis in the hospital Sector: a Case study. International Journal of Advanced Production and Industrial Engineering. Special Issue on OPERATIONS & SUPPLY CHAIN MANAGEMENT. Online Available < https://www. researchgate.net/publication/314129134_ABC_Analysis_in_the_Hospitality_Sector_a_ Case_Study> Ladly John. (2016). Ensuring the quality of ’fit for purpose’ data. Online Available < Ensuring the quality of ’fit for purpose’ data | CIO > Retrieved on: 10.06.2021 LeFave, R., Branch, B., Brown, C. and Wixom, B. (2008) “How Sprint Nextel reconfigured IT resources for results”, MIS Quarterly Executive, Vol. 7, No. 4. Loi Kenneth. 2013. Direct vs Indirect procurement: what’s the difference? Online Available < https://blog.procurify.com/2013/09/25/whats-the-difference-between-indirectand-direct-procurement/> M. Doyle, H. Fereday, M. Gibson, U. Landbeck, R. Lee, C. Maynard, G. Palmer, N. Askham, D. Cook and J. Schwarzenbach. (October, 2013) The Six Primary Dimensions for Data Quality Assessment: Defining Data Quality Dimensions. DAMA UK. Maes, K., Grembergen, W.V., & Haes, S.D. (2013). Identifying multiple dimensions of a business case : a systematic literature review. Electronic Journal of Information Systems Evaluation, 16, 302–314. Marshall Carrie. (2018). Business cases that get results. Mckinsey & Company. (2018). Unlocking success in digital transformation. Survey. Online Available Retrieved on: 30.10.2021 McKinsey & Company. (2019). “Revolutionizing Indirect Procurement for the 2020s | McKinsey.”Retrieved (https://www.mckinsey.com/business-functions/operations/our-ins ights/revolutionizing-indirect-procurement-for-the-2020s). Microtool. (2021). Business case. Evaluate the benefits, costs and risks of a project. Online Available: < https://www.microtool.de/en/knowledge-base/what-is-a-businesscase/> Retrieved on: 01.10.2021 Mina Ezzat Mina, Eid Seed El Kady, Mahmoud Eid Ahmed Mohamed, Ahmed A. Mawgoud, Mohamed Hamed N. Taha (2019). Cloud Migration Strategy for Legacy Systems using AWS Platform. DOI: https://doi.org/10.13140/RG.2.2.12301.28649

102

References

N. Girija and S.K. Srivatsa, 2006. A Research Study: Using Data Mining in Knowledge Base Business Strategies. Information Technology Journal, 5: Pg 591–592. Nakata, C., Zhu, Z., & Kraimer, M. L. (2008). The Complex Contribution of Information Technology Capability to Business Performance. Journal of Managerial Issues, 20, 485– 506 Netjes, M., Reijers, H., & Aalst, V. (2007). Supporting the full BPM life-cycle using process mining and intelligent redesign. Peters, R.; Nauroth, M. (2019): Process-Mining: Geschäftsprozesse: smart, schnell und einfach. Berlin Heidelberg New York: Springer-Verlag. Porter, A. M. (1999). Taking Control of ‘Indirect’ Corporate Spending. Purchasing Provost F, Fawcett T. (2013). Data Science and its Relationship to Big Data and Data-Driven Decision Making. Big Data. Mar;1(1):51–9. doi: https://doi.org/10.1089/big.2013.1508. PMID: 27447038. Ramasamy Anandhi, Soumitra Chowghury BIG DATA QUALITY DIMENSIONS: A SYSTEMATIC LITERATURE REVIEW (2020), Journal of Information Systems and Technology Management – Jistem USP Vol. 17, 2020, e202017003, SSN online: 1807–1775 DOI: https://doi.org/10.4301/S1807-1775202017003 Raymond, L., Pare, G. and Bergeron, F. (1995) “Matching information technology and organizational structure: an empirical study with implications for performance”, European Journal of Information Systems, Vol. 4, No. 1. Remenyi D, Money AH, Bannister F. (2007) The effective measurement and management of ICT costs and benefits. 3rd ed. Oxford; Burlington, MA: CIMA Rendon, R.G. (2005). Commodity sourcing strategies: supply management in action. Robert Nisbet, Gary Miner and Ken Yale. Handbook of Statistical Analysis and Data Mining Application, Second Edition (2018) https://doi.org/10.1016/C2012-0-06451-4 Robledo Pedro. 2018. Process Mining plays an essential role in Digital Transformation. Online Available https://medium.com/@pedrorobledobpm/process-mining-juegaun-papel-esencial-en-la-transformaci%C3%B3n-digital-789391589588 Retrieved on: 14.08.2021 Rosemann Michael, de Bruin Toni. 2005. Towards a Business Process Management Maturity Model. Queensland University of Technology, 126 Margaret Street Brisbane Qld 4000, Australia. Online Available < https://www.researchgate.net/publication/27482324_Tow ards_a_Business_Process_Management_Maturity_Model > Rozinat Anne. 2016. Data Quality Problems In Process Mining And What To Do About Them — Part 4: Wrong Timestamp Configuration Sani, M.F. (2020). Preprocessing Event Data in Process Mining. CAiSE. Sani, M.F., Zelst, S.V., & Aalst, W. (2019). Repairing Outlier Behavior in Event Logs using Contextual Behavior. Sawant Pratik. 2021 Business case Online Available: < https://www.academia.edu/364 12725/Business_Case> Schönig, Stefan (2015) Ein Process Mining-Rahmenwerk für agile, personenbezogene Prozesse. Bayreuth , 2015 .—XIX, 207 P ( Doctoral thesis, 2015, University of Bayreuth, Faculty of Mathematics, Physics and Computer Sciences) Segers I.E.A. Deloitte Enterprise Risk Services, Investigating the application of process mining for auditing purposes. Master’s thesis. Eindhoven University of Technology, 2007. Online available—*Event Log Extraction from SAP ECC 6.0 (tue.nl)

References

103

Segev, A., & Gebauer, J. (2000). Emerging Technologies to Support Indirect Procurement: Two Case Studies from the Petroleum Industry. Information Technology and Management Seidel, S., Rosemann, M., Hofstede, A., Bradford, L., Shortland, K. and Court, D. (2006), Business Process Management for the Creative Industries: Facing Challenges Arising fromGlobalization and New Technology, Business Process Management for the Creative Industries, Beijing Shafique, U., & Qaiser, H. (2014). A Comparative Study of Data Mining Process Models (KDD, CRISP-DM and SEMMA). International journal of innovation and scientific research, 12, 217–222. Sievo 2021. Spend Analysis 101 | Comprehensive Guide to procurement spend. A complete guide to spend analysis. Available: Spend Analysis 101 | Comprehensive Guide for Beginners | Sievo Smith, H., McKeen, J., Cranston, C. and Benson, M. (2010) “Investment Spend Optimization: A New Approach to IT Investment at BMO Financial Group”, MIS Quarterly Executive, Vol. 9, No. 2. Smithson, S. and Hirschheim, R. (1998) “Analysing information systems evaluation: another look at an old problem”, European Journal of Information Systems, Vol. 7, No. 3 Sourcing Innovation. 2021. Next generation supply management defined. Online available < http://sourcinginnovation.com/wordpress/2014/05/15/analyzing-indirect-spend-the-keyto-success-is-to/ > Taleb, I., Kassabi, H.T., Serhani, M.A., Dssouli, R., & Bouhaddioui, C. (2016). Big Data Quality: A Quality Dimensions Evaluation. 2016 Intl IEEE Conferences on Ubiquitous Intelligence & Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People, and Smart World Congress (UIC/ATC/ScalCom/CBDCom/IoP/SmartWorld), 759–765. The Future of Procurement. (2017). Online Available < https://supplychaingamechanger. com/time-to-look-into-the-future-of-procurement/> Thomas, A. J., Mason-Jones, R., Davies, A., & John, E. G. (2015). Reducing Turn-Round Variability through the Application of Six Sigma in Aerospace MRO Facilities. Journal of Manufacturing Technology Management, 26. https://doi.org/10.1108/JMTM-05-20130052 Turner Chris, Tiwari Ashutosh, Olaiya Richard, Xu Yuchun. 2012. Process mining from theory to practice. Urbach, N., Smolnik, S. and Riempp, G. (2010) “An empirical investigation of employee portal success”, The Journal of Strategic Information Systems, Vol. 19, No. 3 van der Aalst W. (2016) Epilogue. In: Process Mining. Springer, Berlin, Heidelberg. https:// doi.org/10.1007/978-3-662-49851-4_16 Van der Aalst, W. M. P.; Reijers, H. A.; Song M. (2005): Discovering Social Networks from Event Logs. In: Computer Supported Cooperative Work, Vol. 14 (2005), No. 6, pp. 549– 593. van der Aalst, W., Reijers, H. & Song, M. Discovering Social Networks from Event Logs. Comput Supported Coop Work 14, 549–593 (2005). https://doi.org/10.1007/s10 606-005-9005-9 Van der Aalst, W.M.P. 2011. Process Mining: Overview and Opportunities. ACM Trans. Manag. Inform. Syst. 99, 99, Article 99 (February 2012)

104

References

van Eck M.L., Lu X., Leemans S.J.J., van der Aalst W.M.P. (2015) PM2: A Process Mining Project Methodology. In: Zdravkovic J., Kirikova M., Johannesson P. (eds) Advanced Information Systems Engineering. CAiSE 2015. Lecture Notes in Computer Science, vol 9097. Springer, Cham. https://doi.org/10.1007/978-3-319-19069-3_19 Van Nederpelt and Black Andrew (2020). Dimensions of data quality (DDQ). DAMA NL Foundation van Zelst, S. J., Mannhardt, F., de Leoni, M., & Koschmider, A. (2020). Event abstraction in process mining van Zelst, S.J., Mannhardt, F., de Leoni, M. et al. 2021. Event abstraction in process mining: literature review and taxonomy. Verhulst, R. (2016). Evaluating quality of event data within event logs: an extensible framework. Wand, Y., & Wang, R. Y. (1996). Anchoring data quality dimensions in ontological foundations. Communications of the ACM, 39(11), 86–95. Wang, R. Y. & Strong, D. M., (1996). Beyond Accuracy: What Data Quality Means to Data Consumers. Journal of Management Information Systems, 12(4) Wang, R.Y. (1998). A product perspective on total data quality management. Commun. ACM, 41, 58–65. Wannenwetsch (2006),“ABC-analysis”. Online Available: Ward, J., Daniel, E. and Peppard, J. (2008) “Building better business cases for IT investments”, MIS Quarterly Executive, Vol. 7, No. 1 Weele, A. J. (2010). Purchasing & supply chain management: Analysis, strategy, planning and practice. Andover: Cengage Learning. Wil M. P. Aalst, 2015. „Extracting Event Data from Databases to Unleash Process Mining“. Wynn M.T., Sadiq S. (2019) Responsible Process Mining—A Data Quality Perspective. In: Hildebrandt T., van Dongen B., Röglinger M., Mendling J. (eds) Business Process Management. BPM 2019. Lecture Notes in Computer Science, vol 11675. Springer, Cham. https://doi.org/10.1007/978-3-030-26619-6_2 Xideas, E., & Moschuris, S. (1998). The Influence of Product Type on the Purchasing Structure. European Journal of Marketing. Yates, J. (1998). Corporate Purchasing Cards Are Hitting Their Stride. Management Accounting, Yu Seunghee, Abhay Nath Mishra, Anandasivam Gopal, Sandra Slaughter, Tridas Mukhopadhyay. (2015). E-Procurement Infusion and Operational Process Impacts in MRO Procurement: Complementary or Substitutive Effects?. DOI: https://doi.org/10. 1111/poms.12362 ZoBell Steven. 2018. Why Digital Transformations Fail: Closing The $900 Billion Hole In Enterprise Strategy. Forbes Technology Council. Online Available < https://www.forbes. com/sites/forbestechcouncil/2018/03/13/why-digital-transformations-fail-closing-the900-billion-hole-in-enterprise-strategy/?sh=416ff32f7b8b > ZoBell Steven. 2018. Why Digital Transformations Fail: Closing The $900 Billion Hole In Enterprise Strategy. Forbes Technology Council. Online Available < https://www.forbes. com/sites/forbestechcouncil/2018/03/13/why-digital-transformations-fail-closing-the900-billion-hole-in-enterprise-strategy/?sh=416ff32f7b8b >