Issues in internet marketing (Marketing intelligence & planning ; V.21, No. 2 ) 9780861768257, 9781845445584, 0861768256

Citation preview

ISSN 0263-4503

Marketing Intelligence & Planning Volume 21 Number 2 2003

Issues in Internet marketing Guest Editor: Kenneth R. Deans

Contents

74 Access this journal online 75 Abstracts & keywords

105 Web site characteristics and business performance: some evidence from international business-to-business organizations

77 Guest editorial 79 Internet based marketing research: a serious alternative to traditional research methods?

Despina A. Karayanni and George A. Baltas

115 E-relationships – emergence and the small firm Thomas O’Toole

Alan Wilson and Nial Laskey

85 A comparison of online and postal data collection methods in marketing research

123 E-trust: the influence of perceived interactivity on e-retailing users Bill Merrilees and Marie-Louise Fry

Heath McDonald and Stewart Adam

96 Luxury branding on the Internet: lost opportunity or impossibility? Francesca Dall’Olmo Riley and Caroline Lacroix

Access this journal electronically The current and past volumes of this journal are available at

www.emeraldinsight.com/ft You can access over 100 additional Emerald journals, each with a comprehensive searchable archive of articles (many dating back to 1989), a detailed classification system and links to referenced material. See page following contents for full details of what your access includes.

www.emeraldinsight.com/mip.htm As a subscriber to this journal, you can benefit from instant, electronic access to the content of this title. Your access includes a variety of features that increase the value of your journal subscription.

Automatic permission to make up to 25 copies of individual articles This facility can be used for teaching purposes, training courses, meetings or seminars. This only applies to articles of which Emerald owns copyright. For further details visit www.emeraldinsight.com/copyright

How to access this journal electronically To benefit from electronic access to this journal you first need to register on the Internet. Registration is simple and full instructions are available online at www.emeraldinsight.com/register Once completed, your institution will have instant access to the journal content from www.emeraldinsight.com or via the journal homepage www.emeraldinsight.com/mip.htm Our liberal institution-wide licence allows everyone within your institution to access your journal electronically, thus making your subscription more cost-effective. Our Web site has been designed to provide you with a comprehensive, simple system that needs only minimum administration. Access is available via IP authentication or username and password.

Key features of Emerald electronic journals Online Publishing and Archiving You can gain access to past volumes as well as new material from this journal on the Internet via Emerald Fulltext. You can browse or search the database for relevant articles. Key Reading This feature provides abstracts of articles chosen by the journal editor on the basis of their subject-matter. These are selected to provide readers with current awareness of interesting articles from other publications within the same field. Reference Linking Direct links are provided from the journal article references to abstracts of the most influential articles cited (where possible this is to the full text of the article). E-mail an Article This facility allows you to e-mail relevant and interesting articles in PDF format to another PC for later use, reference or printing purposes.

Additional Complementary Services Available Your access includes a variety of features that add to the functionality and value of your journal subscription:

E-mail Services Emerald’s range of free e-mail alerting services is designed to deliver personal notification of news and features in a number of different interest areas. Emerald WIRE (World Independent Reviews) Emerald WIRE is a fully searchable, subject-specific database brought to you by Emerald Management Reviews. It provides article reviews from the world’s top management journals. The database is updated monthly and gives users details of how to obtain the full text of original articles. Research Register Located at www.emeraldinsight.com/researchregister, the Emerald Research Register is an Internet research forum where you and your readers can identify information on research activity world-wide. This feature is invaluable if you are seeking information or wish to disseminate information about your own research. Support Resources Comprehensive librarian and user toolkits have been created to help you get the most from your journal subscription. For further information about what is available visit www.emeraldinsight.com/usertoolkit

Choice of Access Electronic access to this journal is available via a number of channels, enabling end users and libraries to reach the content through their preferred delivery system. The Emerald Fulltext Web site – www.emeraldinsight.com/ft – is the recommended means of electronic access, as it provides fully searchable and high value-added access to the complete content of the journal. Subscribers can also access and search the article content of this journal through the following journal delivery services: EBSCOhost EJS journals.ebsco.com Huber E-Journals e-journals.hanshuber.com/english/index.htm Minerva www.minerva.at OCLC Firstsearch Electronic Collections Online www.uk.oclc.org/oclc/menu/eco.htm RoweCom’s Information Quest www.informationquest.com SilverPlatter www.silverplatter.com SwetsBlackwell’s SwetsWise www.swetswise.com

Emerald Customer Support Services For customer service and technical help, contact: E-mail [email protected] Tel +44 (0) 1274 785278 Fax +44 (0) 1274 785204

Abstracts & keywords

Internet based marketing research: a serious alternative to traditional research methods?

A comparison of online and postal data collection methods in marketing research

Alan Wilson and Nial Laskey

Heath McDonald and Stewart Adam

Keywords Internet, Market research, Online computing, Questionnaires

Keywords Internet, Marketing research, Data collection, Online retrieval, Postal services, Surveys

This paper examines how online market research is utilised within UK market research agencies and what opportunities or problems this new research tool is giving the market research industry. The research involved sending 120 postal questionnaires to UK based market research agencies that claimed to do Internet based research. The main findings from the study show that the dramatic growth predicted by industry commentators is not evident among practising market researchers. Internet research is tending to be used for specialist types of research study such as Web site evaluation, business to business and employee research, where the audience is more likely to be online, or are part of an easily accessible database list supplied by clients. The research also found continuing concern about sample frames, attrition of panel members and response rates. Many in the industry are still undecided on the likely problems and opportunities associated with Internet research. There needs to be far more research and development if Internet based research is to become a serious alternative to traditional research methods.

The widespread acceptance of the use of online techniques in market research necessitates appreciation of the relative advantages and disadvantages of these techniques over more traditional research methods. This paper reports on a study which directly compares online and postal data collection methods using the same survey instrument on two samples drawn from the same population of football club subscribers. The results confirm that the online and postal respondents are demographically different. Online data collection is shown to be less expensive per respondent and that data collection is faster, however, an overall lower response level is achieved relative to the postal data collection method. Of greater importance, though, are the findings that respondents seem to answer questions differently online than they do via postal methods. The conclusion here is that online data collection should not be treated as a direct substitute for postal data collection in every instance.

Luxury branding on the Internet: lost opportunity or impossibility? Francesca Dall’Olmo Riley and Caroline Lacroix Keywords Brand names, Brands, Internet, Brand image, France, United Kingdom The paper addresses the issue of the use of the Internet for the management of luxury brands. We investigate luxury managers’ attitudes towards the use of the Internet to promote their brands and also assess consumers’ opinions and behaviour concerning luxury brands Web sites. Finally, we present the results of a content analysis of a number of luxury brands sites. Any discrepancies between expectations and reality are investigated. Results indicate that reality falls short of the expectations of both managers and customers and that the interactive potential of the Internet is not exploited. Furthermore, there is evidence that the Internet works best as a communication than as a customer acquisition channel for luxury brands.

Marketing Intelligence & Planning 21/2 [2003] Abstracts & keywords # MCB UP Limited [ISSN 0263-4503]

[ 75 ]

Abstracts & keywords Marketing Intelligence & Planning 21/2 [2003] 75-76

Web site characteristics and business performance: some evidence from international business-to-business organizations Despina A. Karayanni and George A. Baltas Keywords Web sites, Business-to-business marketing, Marketing communications, Internet, Business strategy, Interaction This paper addresses the effectiveness of Web site capabilities. It is intended to empirically determine the importance of site properties for Internet sales performance. More specifically, we consider the relationship of characteristics such as navigability, interactivity, multimedia design and marketing communications content with sales attributed to the Internet. Advanced econometric modeling of cross-section data on business-to-business firms demonstrates that interactive responsiveness of the site, enabling customized interactive programs and club membership, fast-downloading multimedia, such as frames and animation, corporate information stressing corporate positioning and active information submission on behalf of the target audience may influence sales performance. Important implications for optimization of business-tobusiness Internet strategies are also considered.

E-relationships – emergence and the small firm Thomas O’Toole Keywords Electronic resources, Small firms, Relational databases, Competitive advantage, Information technology The e-business model has become an integral feature of business practice. Its existence has created an extra electronic layer to business relationships. This has facilitated close partnerships to add unique and complex electronic components to their relationships that are not easily copied by other firms. The emergence of e-relationships is detailed and their value creating potential explored. The e-relationship concept is then applied to the small firm. The strategic and implementation barriers that inhibit small firms from capitalising on such relationships are examined. E-relationships are found to offer competitive advantage to few small firms. Critical to an e-relationship’s success is the relationship orientation of the partners.

[ 76 ]

E-trust: the influence of perceived interactivity on e-retailing users Bill Merrilees and Marie-Louise Fry Keywords Electronic commerce, Interaction, Internet, Retailing, Consumer behaviour, World Wide Web E-retailing has received considerable attention in recent years, especially with the help of consumer behaviour concepts and tools. This particular paper focuses on e-trust and its antecedents. Special attention is given to the proposition that the more a site is perceived to be interactive, then the greater the trust that users attach to that site. Previous research has demonstrated this relationship with respect to a particular site. The current paper re-examines this relationship with respect to another site. More importantly, the paper tests whether the nexus between interactivity and e-trust applies to both pure ‘‘click’’ and ‘‘brick and click’’ e-retailers.

Guest editorial

About the Guest Editor Originally a scientist, Dr Kenneth R. Deans moved to the dark side of the force in the 1970s and is now a Senior Lecturer in the Department of Marketing within the internationally ranked School of Business at the University of Otago, Dunedin, New Zealand. He lectures undergraduate, postgraduate, MBA and executives in principles of marketing, marketing strategy, Internet marketing and services marketing. His research interests are Internet marketing, services and marketing education. His work, in collaboration with overseas researchers has led to more than 50 publications in journals, books and refereed conference proceedings as well as numerous seminars to businesses and organisations. Over the years Ken has worked for and with a number of SMEs as well as blue-chip companies such as Scottish & Newcastle and Pilkington. Since moving to Otago from Strathclyde he has been heavily involved in the department’s Marketing Performance Centre, researching service sector competitiveness. In his spare time he enjoys photography, travel, mountain biking, running, swimming and tramping, as well as the obligatory post-activity beer, food and wine. Visit him at: http:// marketing.otago.ac.nz/ marketing/staff/deansk.html

Marketing Intelligence & Planning 21/2 [2003] 77-78 # MCB UP Limited [ISSN 0263-4503]

Issues in Internet marketing The Internet, and its incorporation in what many of us refer to as ‘‘traditional’’ marketing management, presents marketing mangers and academics with as many challenges as it does opportunities. Personally, I believe we could have avoided much of the hype and cut to the reality of what was really happening and evolving much sooner. In many ways the Internet simply presents other ways of doing what we have always done. However, many viewed the Internet as a panacea for falling sales or loss of market share. So the euphoric gold rush to embrace the technology, the countless millions of dollars, pounds, Francs and Yen that were thrown at it and the cataclysmic crash that ensued could all have been avoided. It appeared that we had learnt little from historical accounts of business growth, shakeout and eventual ‘‘winners’’. In our collective naı¨vety we believed there were going to be clear winners in each market, industry, sector or product class. Inevitably, then, the bomb or, as I prefer to call it, the ‘‘World Wide Waste of money dot con’’ was going to happen. It was simply a matter of time and, unfortunately, money! On a more optimistic note, my own research has highlighted a gradual shift in firms’ approach to Internet-based marketing. Increasingly, they are adopting a more strategic perspective with fewer firms viewing it as a tactical ‘‘because we can’’ or ‘‘because they have’’ issue. Working with Stewart Adam I have also come to rationalise Web presence down to three broad categories: 1 communication; 2 transaction/logistics; and 3 relationship enhancement. Not all are appropriate for all firms, nor is progression sequential. It is quite acceptable for a firm to do nothing more than communicate and for others to start by communicating and then progress to a relationship building focus, totally missing out transactional and logistics issues. Transactions and logistics may not be of interest or appropriate for many firms. In the light of research into customer relationship management and, more recently, customer satisfaction management, the Internet facilitates that shift in focus from revenue generation to integrated business support.

The challenge to marketers is to clearly identify and articulate Internet objectives as integral constituents of a successful strategy and its implementation. Too often objectives are non-specific and there are no identifiable or quantifiable measures of success. Establishing such metrics is an emerging area of research and one that is being investigated by two of my own doctoral candidates. So when Michael Thomas and I first talked about this special edition, ‘‘Issues in Internet marketing’’, I felt both thrilled and challenged. From the comfortable perspective of 20:20 hindsight I can confidently say the experience was first class. I would like to acknowledge the considerable help and advice I was given by my friend and colleague Philip Osborne. Also, all the contributors were a delight to work with and almost every submission I received was worthy of inclusion at best or modification for future consideration. That level of quality made my job harder as I sought a spread of articles that would provide both geographic spread and topic variation. As I was on sabbatical while some of the editing and administration was undertaken I was able to consider most of the accepted papers in the light of my research and experiences in a number of countries across Australasia, China, Europe and the USA. It struck me that the biggest challenges to the less technologically advanced countries were; getting the infrastructure right, educating the business community and learning from the successes and failures from ‘‘more Internet advanced’’ economies while avoiding the temptation to seek a blueprint. The papers presented here are all from established researchers and contributors to our understanding of marketing theory, Webbased and otherwise. I hope you enjoy reading them and learn from their insightful research and commentary. I think there is something in each of the papers for all marketing academics and practitioners. Alan Wilson and Nial Laskey, both from the University of Strathclyde, investigate the appropriateness and usefulness of Internet based marketing research. They report on the reality of its use compared to the dramatic growth predicted. They also offer some commentary on common issues such as ‘‘sample frames, attrition of panel members and response rates’’. A sample of 120 UKbased market research agencies yielded a 53

[ 77 ]

Guest editorial Marketing Intelligence & Planning 21/2 [2003] 77-78

[ 78 ]

percent response rate and some generalisable findings and conclusions. Heath McDonald and Stewart Adam (Deakin University) have examined the online research debate from a postal:online data collection perspective. Many firms seeking to collect valuable and timely customer data are considering the Internet as an alternative to traditional postal collection methods, the ongoing e-mail:snail-mail debate. The authors provide a succinct and useful overview of previous research before presenting their own work. The third paper by Francesca Dall’Olmo Riley and Caroline Lacroix from Kingston University Business School considers the appropriateness or otherwise of the Internet as tool to assist in the management of luxury goods. They have approached it from both a managerial and customer perspective. Additionally, they have included a technique that I endorse and is gaining favour among researchers, that of Web site content analysis. Their findings endorse my comments made earlier regarding Web presence. Despina Karayanni (University of Patras ) and George Baltas (Athens University of Economics and Business) have written an interesting B2B piece that considers the

contribution of Web site characteristics to sales performance. There has been much criticism of firms which simply transfer paper materials to the Web in the mistaken belief that the characteristics of the two media are sufficiently similar to permit such transfer, brochureware. This paper includes sections on interactivity, navigability, multimedia and communication content in the light of their contribution to Internet attributed sales. Waterford Institute of Technology’s Tomas O’Toole has written a paper that belongs in my third category – relationship enhancement. He considers the opportunity offered by the Internet to small firms seeking to establish closer and more meaningful business relationships that create value. Barriers to implementation are also examined. The final paper in this edition, by Bill Merrilees and Marie-Louise Fry, provides an exploration of trust in the online environment. Its inclusion is timely as they examine interactivity, a hot topic, and its effect on trust in both pure ‘‘click’’ and ‘‘brick and click’’ e-retailers. Enjoy them all. Kenneth R. Deans Guest Editor

Internet based marketing research: a serious alternative to traditional research methods?

Alan Wilson Department of Marketing, Strathclyde University, Glasgow, UK Nial Laskey Department of Marketing, Strathclyde University, Glasgow, UK

Keywords Internet, Market research, Online computing, Questionnaires

Abstract This paper examines how online market research is utilised within UK market research agencies and what opportunities or problems this new research tool is giving the market research industry. The research involved sending 120 postal questionnaires to UK based market research agencies that claimed to do Internet based research. The main findings from the study show that the dramatic growth predicted by industry commentators is not evident among practising market researchers. Internet research is tending to be used for specialist types of research study such as Web site evaluation, business to business and employee research, where the audience is more likely to be online, or are part of an easily accessible database list supplied by clients. The research also found continuing concern about sample frames, attrition of panel members and response rates. Many in the industry are still undecided on the likely problems and opportunities associated with Internet research. There needs to be far more research and development if Internet based research is to become a serious alternative to traditional research methods.

Introduction The Internet has grown exponentially over the last decade from its humble beginnings as a tool created by the US military in the Cold War, to its present incarnation as a seemingly omnipresent entity that, in the Western world at least, has revolutionised a lot of business and consumer behaviour (Harter, 1999). One consequence of this so-called ‘‘Internet revolution’’ is the impact that it may have in providing an alternative research tool for market research agencies. The Chairman of Harris Black International was quoted as saying, ‘‘all research is going to migrate to the Internet’’ (McDaniel and Gates, 2001) and ESOMAR also predicted that 30 per cent of research world-wide will be conducted online by the year 2010 (Savage, 2001). Although these predictions exist, there is very little research, particularly outside the USA, to determine the current usage of Internet based research and the experiences of those involved in providing such methodology. This paper will endeavour to redress the paucity of non-US research in the area of Internet market research by examining how online market research is utilised within UK market research agencies and what opportunities or problems the market research industry is experiencing with this new research tool.

The cost of sending e-mails or setting up a Web questionnaire is much cheaper than mail equivalents. Weible and Wallace (1998) estimated that the variable cost of sending a questionnaire (via e-mail) or survey invitations (for a Web survey) to 150 additional e-mail addresses was the equivalent of adding one extra contact to a mailing sample. Watt (1997) states that Internet interviewing becomes significantly cheaper than postal surveys particularly when there are over 500 respondents. Below this number, the set up costs will tend to make the traditional postal method more cost effective.

3. Sampling difficulties

If a researcher is using an Internet survey much of the administration burden of The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0263-4503.htm

The literature suggests a number of positive and negative attributes of the new research medium. These are as follows:

1. Ease of use

# MCB UP Limited [ISSN 0263-4503] [DOI 10.1108/02634500310465380]

2. Cost

In the UK the number of people that have Internet access has increased from approximately 960,000 in June 1997, to 19.98 million in November 2000 that represents an estimated 33.58 per cent of Britain’s population (NUA Web site, 2001). Despite these impressive growth rates the penetration of the Internet has not, as yet, developed to the extent that it represents the population as a whole. Taylor (2000) describes the Internet population as being younger and possessing a higher standard of

Perceived advantages and disadvantages of Internet based research

Marketing Intelligence & Planning 21/2 [2003] 79-84

sending/receiving questionnaires and data entry is significantly reduced. Once a questionnaire is completed the data are already in electronic form and can be downloaded instantaneously into a database for complex analysis (Iyer, 1996). In effect, once the last questionnaire for a survey is complete, researchers have all the data stored in a database at their fingertips instantaneously. Reminders to complete the questionnaire can be permanently placed on Web sites or can be sent out by e-mail at no additional cost and with minimal effort.

[ 79 ]

Alan Wilson and Nial Laskey Internet based marketing research: a serious alternative to traditional research methods? Marketing Intelligence & Planning 21/2 [2003] 79-84

education than the public at large. This may not be as big a problem as it seems, for example in the USA, Internet research tends to focus on doing research aimed at the middle to upper income segment (Mehta and Sivadas, 1995; Oppermann, 1995). There is, however, a major difficulty relating to the lack of a centralised database of e-mail addresses (Litvin and Kar, 2001). This is in stark contrast to mail surveys where addresses are commonly found in a wide variety of directories and databases. Even where sample frames are available, they frequently become outdated as Internet users change their e-mail providers (Dommeyer and Moriarty, 2000). In a study undertaken by Oppermann (1995) it was discovered that of the 500 e-mail addresses that were selected from the members directory of Association of American Geographers (AAG), 25 per cent were out of date. The advantage to researchers using email (whether to deliver a questionnaire or as an invitation to a Web site survey) is that the error messages are sent back to the sender of the e-mail allowing the researcher (as in the case of Oppermann, 1995) to select another set of individuals from the list in order to attempt to obtain the sample size that the researcher originally envisaged. Without a centralised e-mail database that covers the whole online population, researchers must use alternative methods to get their samples. Mehta and Sivadas (1995) selected their respondents by initiating a programme that collected the e-mail addresses from individuals that posted discussion articles on the 20 most popular discussion groups within an Internet community known as Usenet, which is basically a world-wide system of discussion groups (Kumar et al., 1999; Jackson and DeComormier, 1999). Although this method was efficient at collecting a large database of e-mail addresses it still may fail to develop a sample that is representative of the population as a whole, it may include the segment that contribute to discussion groups.

4. Response rates Another attribute of Internet surveys surrounds the responses from the sample. Most studies agree that when comparing email surveys with mail surveys, response rates can be rapid, in some extremes overnight (Mehta and Sivads, 1995; Tse et al., 1994; Bachmann et al., 2000; Taylor, 2000). As with normal postal surveys, responses to Internet surveys tend to vary according to the study. However, Table I indicates how various studies have been constructed and how the response rates have differed. It

[ 80 ]

Table I Summary of research studies using e-mail showing response rates Author Parker (1992) Mehta and Sivadas (1995) Tse et al. (1994) Tse (1998) Weible and Wallace (1998)

Methods used Response rate in study (%) Mail E-mail Mail E-mail Mail E-mail Mail E-mail Mail E-mail Web form

38 68 57 54 27 6 52 7 36 30 33

Note: Adapted from Sheehan and McMillan (1999) generally appears from the summary in Table I that, with the exception of Parker’s (1992) study, response rates from e-mail surveys tend to be lower than those of traditional postal surveys. A concern brought up by Bachmann et al. (2000) is that, as the population becomes more accustomed to e-mail they may become more reluctant to respond to surveys. Essentially as the ‘‘novelty’’ of the Internet wears off there is the potential that online research may suffer as Internet users become more apathetic towards such studies. Burkeman (2001) has indicated that the volume of e-mails has increased substantially to the point that the phrase ‘‘e-mail burnout’’ has been created. He states that the volume of e-mail that an individual can receive is so great that many delete up to 60 per cent of their messages based on the subject line alone. Due to these poor return rates (even if they tend to be much faster) there has been some investigation into using Internet survey response as an option in a mixed mode methodology. The reason for this probably came from studies such as Parker (1992) where a small proportion (28 per cent) of the e-mail respondents decided to print out hard copies of the questionnaire and send them back by more conventional mailing systems.

5. Response quality If the consensus of the literature reviewed to date is that the responses from Internet surveys (mainly e-mail questionnaires) are lower than postal surveys then there also tends to be agreement within the literature that online surveys tend to have response quality equal if not superior to the other forms of survey. In the studies by Tse et al. (1994; Tse, 1998) the quality of answers was

Alan Wilson and Nial Laskey Internet based marketing research: a serious alternative to traditional research methods? Marketing Intelligence & Planning 21/2 [2003] 79-84

comparable to those from the postal survey. In fact, Bachmann et al. (2000) indicated that participants of online surveys were much more willing to respond to open-ended questions. Mehta and Sivadas (1995) also discovered that responses to open-ended questions in their Internet survey were significantly longer than those obtained through similar postal surveys. Taking account of these issues (ease of use; cost; sampling difficulties; response rates and response quality) there is a need to understand how Internet based marketing research is currently being used by the market research industry and the extent to which the research tool is creating opportunities or problems.

Research design The research involved sending 120 postal questionnaires to all of the member agencies of the UK’s Market Research Society (the leading professional body for marketing research in the UK) that claimed to provide Internet surveys to clients. These claims were made in the agencies’ entries within the members’ directory. A total of 64 usable responses (53 per cent response rate) was obtained, of which 53 (83 per cent) undertook Internet based research for clients and 11 (17 per cent) indicated that although they planned to offer such research had not as yet done so.

Findings The majority of the respondents had only started undertaking Internet surveys during the last 24 months. However, just over half of the respondents (55 per cent) considered Internet based research to be very important to their organisation. During the previous 12 months, 57 per cent of the respondents had undertaken more than five online studies.

Reasons for offering Internet surveys The majority of respondents that provided Internet surveys agreed that the reasons for moving into this area were principally: 1 to have a balanced portfolio of research methodologies; and 2 to develop new research areas (see Table II). Responding to client demand was not viewed as such a significant motivational force, neither was the desire to become an online survey specialist.

Types of market research As Table III shows, the majority of research being undertaken is quantitative in nature

and is delivered through e-mail, pre-recruited panels or through electronic mailing lists/discussion groups. Of the respondents, 47 per cent combined Internet research with other methodologies in a mixed mode approach with techniques such as telephone or postal surveys. Only 23 per cent of the respondents had attempted qualitative research on the Web by using on-line discussion groups. In terms of the purpose of the research, Internet based market research is frequently linked to Web-based activities such as Web site evaluation (see Table IV) or to research areas where respondents have easy access to the Web such as is the case with employee surveys and those in business to business research.

Types of client Internet based research tends to be conducted for organisations that have a significant Internet presence such as financial services, Internet based businesses, IT, the media and telecommunications. This may reflect the fact that Web site evaluations are one of the main types of research (Table V).

Attitudes towards Internet surveys Table VI sets out respondents’ agreement with a variety of attitude statements. As expected, the most important factor for determining whether Internet survey methodology is used is the extent to which the audience is online. The majority of respondents agreed that Internet research required less administration and saw no significant problems with security. However, there were concerns about the problems associated with sampling, drop out rates for panels and response rates. Although 42 per cent of respondents thought that Internet surveys are used for important decision making research, it is interesting to note that more than one third of respondents felt that they were not. This negative viewpoint is supported by the 25-30 per cent of respondents who stated that cost or speed rather than quality determined when online surveys are used. This suggests that there are still doubts about Internet based research and these doubts are also evident by the large proportion of respondents who were undecided and failed to agree or disagree with many of the statements.

Future prospects Only 7 per cent of respondents expected the number of Internet based studies to increase significantly over the next year. The remainder expected no growth or only a

[ 81 ]

Alan Wilson and Nial Laskey Internet based marketing research: a serious alternative to traditional research methods? Marketing Intelligence & Planning 21/2 [2003] 79-84

moderate increase. So there does not seem to be any evidence of the dramatic growth that many industry commentators were predicting. Some commentators had suggested that potential clients would start carrying out their own online surveys, bypassing agencies due to the cost advantages and the availability of technology. However, only 10 per cent of the agencies agreed that they would be doing less Internet research in the future as a result of clients doing their own online research (see Table VII).

Table II Reasons for provision of Internet based research (base = 53)

Statements To have a balanced portfolio of research methodologies To develop new research areas Reacting to competitors moving into this area Clients were pushing us towards this research To become an online survey specialist

Strongly agree/ Neither agree/ agree disagree (%) (%)

Strongly disagree/ disagree (%)

Conclusions 89 79

5 15

6 6

51

20

29

39

27

34

36

24

40

Table III Research methods used (base = 53) Research methods used

%

E-mail questionnaires Pre-recruited panel Web sites List based Web surveys Online focus groups

62 51 42 23

Table IV Research applications (base = 53) Type of research

%

Web site evaluation Customer satisfaction Employee research Business-to-business

66 51 47 43

Table V Main sectors undertaking Internet research (base = 53)

[ 82 ]

In relation to other commentators’ predictions, the majority of market research firms from the study did not agree (64 per cent) that most of their quantitative research would be done through Internet surveys in the future. Almost all of the respondents agreed (91 per cent) with the statement that their firm would only use their online surveys as one part of a portfolio of research offerings and emphasised that they had no intention of becoming specialised in the use of Internet surveys as a data collection tool. However, 65 per cent of agencies that responded to this statement agreed they would be looking to use their Internet surveys for a wider range of market research study (for example – employee surveys) in the future.

Leading sectors

%

Financial services Internet business (dotcoms/e-commerce) Information technology Media (TV/publishers) Telecommunications Government

47 43 36 34 32 25

The main findings from the study show that Internet surveys are generally offered by marketing research agencies in order to provide a balanced portfolio of research services to their clients. The Internet can enable researchers to access certain markets and audiences such as teenagers and business people that may prove difficult through traditional research methods. However, the dramatic growth in Internet surveys predicted by industry commentators is not evident among practising market researchers. The use of Internet research is predominately limited to specialist types of research study such as Web site evaluation, business to business and employee research, where the audience is more likely to be online, or are part of an easily accessible database list supplied by clients. As a result the sectoral concentration of activity relates to those sectors that make most use of the Internet such as financial services, Internet businesses and IT. In terms of operational issues, there are continuing concerns about sample frames, attrition of panel members and response rates, although it is also noticeable that many attitudinal responses in the study were neutral suggesting that much of the industry has still to reach a verdict on the problems and opportunities posed by Internet research. Generally, there is not an issue with the ability of Internet surveys to collect data, the doubts relate more to the representativeness of any data collected. The nature of Internet sample frames and the characteristics of the people willing to take time to respond to Internet surveys or participate in Internet panels is likely to

Alan Wilson and Nial Laskey Internet based marketing research: a serious alternative to traditional research methods? Marketing Intelligence & Planning 21/2 [2003] 79-84

Table VI Attitudes towards Internet surveys (base = 68) Strongly agree/ Neither agree/ disagree agree (%) (%)

Statements The most important factor for Internet survey use is the extent to which the audience is online Internet surveys help to reduce administration Security is a big issue with online surveys Sampling problems are more prominent with Internet panels Drop out rates are a particularly large problem with Internet surveys You receive excellent response rates from Internet surveys Cost rather than quality determine when it is better to use online surveys Speed rather than quality determine when it is better to use online surveys Internet surveys are not used for highly important decision making research

Strongly disagree/ disagree (%)

82 55 25

12 23 31

6 22 44

61

25

14

55

22

23

21

38

41

25

26

49

30

26

44

36

22

42

Table VII Future propects (base = 68) Strongly agree/ Neither agree/ agree disagree (%) (%)

Statements We will use Internet surveys to conduct most of our quantitative research Internet surveys will only be part of our portfolio, we will not become a specialist We will look to use Internet surveys for other types of research studies We will be doing less Internet studies because companies will be doing online surveys for themselves

mean that respondents may not be representative of typical computer users, never mind the general public. This may explain why the majority of market research companies do not see Internet based research as being their main approach to quantitative research in the future, although they do agree that Internet based research will continue to make up part of their portfolio of services. As to future academic research, there is certainly a need to track the changing attitudes towards Internet surveys over time as more and more people go online. Changes in technology, such as interactive digital broadcasting and Web-enabled mobile telephones, combined with a more widespread use of the Internet for shopping, banking and education, are likely to alter

Strongly disagree/ disagree (%)

14

22

64

91

2

7

65

26

9

10

42

48

the usage and appropriateness of online research. Research is also required to determine whether attitudes towards Internet surveys are the same world-wide. Attitudes towards other methodologies vary from country to country; do similar variations occur with Internet surveys? Finally, if Internet based research is to grow, there also needs to be significantly more research undertaken to address the perceived weaknesses (i.e. sampling, response rates and panel drop out rates) of the methodology. Many practitioners are still undecided about Internet methods and need further academic research and guidance if Web-based research is to become a serious alternative to traditional research methods.

[ 83 ]

Alan Wilson and Nial Laskey Internet based marketing research: a serious alternative to traditional research methods? Marketing Intelligence & Planning 21/2 [2003] 79-84

Managerial implications For the majority of organisations, this research has shown that Internet based research should be seen as an additional supporting methodology rather than as an alternative to traditional research approaches. In other words, the Internet is best used alongside other methodologies to pick up additional respondent groups such as business people and teenagers who may be more difficult to access through personal, telephone or postal surveys. However, for some organisations requiring specialist areas of research such as employee surveys and Web site evaluation, an Internet survey may be the most cost effective and appropriate way of carrying out the research. Wherever Internet surveys are used, a word of caution, clients and researchers should pay particular attention to the extent which the survey respondents are representative of the specific population of interest. Although, this issue may become less of a problem in the future, sample frames and response rates will continue to be critical in evaluating the worth of Internet surveys in the short to medium term.

References Bachmann, D.P., Elfrink, J. and Vazzana, G. (2000), ‘‘E-mail and snail mail face off in rematch’’, Marketing Research, Winter 1999/ Spring 2000, Vol. 11 No. 4, p. 10. Burkeman, O. (2001), ‘‘Postmodern’’, Guardian Newspaper G2 Section, 20 June, pp. 2-3. Dommeyer, C.J. and Moriarty, E. (2000), ‘‘Comparing two forms of an e-mail survey: embedded vs attached’’, Journal of the Market Research Society, Winter, Vol. 42 No. 1, pp. 39-50. Harter, B. (1999), ‘‘On-line data collection’’, Wireless Review, September, Vol. 16 No. 18, pp. 54-62. Iyer, R. (1996), ‘‘The Internet: a new opportunity for marketing research firms’’, Quirk’s Marketing Research Review, May. Jackson, A. and DeCormier, R. (1999), ‘‘E-mail survey response rates: targeting increases response’’, Marketing Intelligence & Planning, Vol. 17 No. 3, pp. 135-9.

[ 84 ]

Kumar, V., Aaker, D.A. and Day, G.S. (1999), Essentials of Marketing Research, John Wiley & Sons, New York, NY. Litvin, S.W. and Kar, G.H. (2001), ‘‘E-surveying for tourism research: legitimate tool or a researcher’s fantasy?’’, Journal of Travel Research, February, Vol. 39 No. 3, p. 308. McDaniel, C. and Gates, R. (2001), Marketing Research Essentials, 3rd ed., South-Western College Publishing, Cincinatti, OH. Mehta, R. and Sivades, E. (1995), ‘‘Comparing response rates and response content in mail versus electronic mail surveys’’, Journal of the Market Research Society, October, Vol. 37 No. 4, pp. 429-40. NUA Web site, West European Internet Figures, available at: www.nua.net/surveys/ how.many.online/europe/html (accessed 19 July 2001). Oppermann, M. (1995), ‘‘E-mail surveys – potentials and pitfalls’’, Marketing Research, Vol. 7 No. 3, p. 28. Parker, L. (1992), ‘‘Collecting data the e-mail way’’, Training and Development, July, pp. 135-9. Savage, M. (2001), ‘‘Road to nowhere?’’, Research, June, No. 421, pp. 20-1. Sheehan, K.B. and McMillan, S.J. (1999), ‘‘Response variation in e-mail surveys: an exploration’’, Journal of Advertising Research, July, Vol. 39 No. 4, p. 45. Taylor, H. (2000), ‘‘Does Internet research work?’’, International Journal of Market Research, Vol. 42 No. 1, pp. 51-63. Tse, A., Ching, R., Ding, Y.B., Fong, R., Yeung, E. and Au, A. (1994), ‘‘A comparison of the effectiveness of mail and facsimile as survey media on response rate, speed and quality’’, Journal of the Market Research Society, October, Vol. 36 No. 2, pp. 349-56. Tse, A.C.B. (1998), ‘‘Comparing the response rate, response speed and response quality of two methods of sending questionnaires: e-mail vs mail’’, Journal of the Market Research Society, October, Vol. 40 No. 4, p. 353. Watt, J. (1997), ‘‘Using the Internet for quantitative research’’, Quirk’s Marketing Research Review, June. Weible, R. and Wallace, J. (1998), ‘‘Cyber research: the impact of the Internet on data collection’’, Marketing Research, Vol. 10 No. 3, pp. 19-24.

A comparison of online and postal data collection methods in marketing research

Heath McDonald Deakin University, Melbourne, Australia Stewart Adam Deakin University, Melbourne, Australia

Keywords Internet, Marketing research, Data collection, Online retrieval, Postal services, Surveys

Abstract The widespread acceptance of the use of online techniques in market research necessitates appreciation of the relative advantages and disadvantages of these techniques over more traditional research methods. This paper reports on a study which directly compares online and postal data collection methods using the same survey instrument on two samples drawn from the same population of football club subscribers. The results confirm that the online and postal respondents are demographically different. Online data collection is shown to be less expensive per respondent and that data collection is faster, however, an overall lower response level is achieved relative to the postal data collection method. Of greater importance, though, are the findings that respondents seem to answer questions differently online than they do via postal methods. The conclusion here is that online data collection should not be treated as a direct substitute for postal data collection in every instance.

Marketing Intelligence & Planning 21/2 [2003] 85-95 # MCB UP Limited [ISSN 0263-4503] [DOI 10.1108/02634500310465399]

Introduction Use of the Internet (Net) as a medium, and the World Wide Web (Web) as an evolving technology has made it less costly and allowed marketers to get information – both of low quality and high quality – more quickly and easily than ever before. Couper (2000, p. 466) notes, however, that the Net offers ‘‘a wide array of approaches representing varying levels of quality and cost’’. This is but one of a number of commentaries concerning the quality and cost of Internet marketing research by a long list of researchers such as Dommeyer and Moriarty (2000), Weible and Wallace (1998) and Comely (1996). A number of problematic issues exist in the undertaking of marketing research in general, and in online marketing research in particular, including declining response levels and the resulting representativeness of those responding. Past research into online research has often produced conflicting results on such crucial issues as response levels and data quality. Some of these conflicting results can be attributed to the variety in research designs used, however, the broader issues remain unresolved. This paper continues the discussion on the quality issues in online marketing research by comparing the outcomes of a survey involving both postal and online delivery of the same questionnaire (in terms of items) delivered to two sample frames from the same population of football club members. The focus of the paper is on the response patterns and the demographic profiles of the two groups of respondents using the two data collection methods employed. The main objective of the research presented here is to clarify the exact nature of the advantages and disadvantages online market research offers The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

in comparison to postal delivery. It is hoped the results will help guide researchers who are contemplating the use of one, or both, of these data collection methods.

Online technologies in marketing research Most marketing scientists and practitioners have used one or all of the possible research approaches – observational research, survey research and experimental research – at one point in time. Moreover, they have in all likelihood used one or more of the four survey contact methods – mail, telephone, personal interviews and electronic networks. With the convergence of telecommunications, media and TCP/IP (Transmission Control Protocol/Internet Protocol) technology, collectively called TMT, we might extend this list to include interactive television, cellphones, and even touch-screen responses to 3D simulations in store situations. Recognising the ever expanding list of options open to market researchers seeking to collect data, it becomes increasingly important to understand the unique characteristics of each method. A detailed understanding of these unique characteristics ensures that the problems inherent in using these research methods, such as non-response bias, can be controlled or overcome. To date, much of what we know of these emerging methods of data collection has mostly been learnt from small-scale academic studies. This study replicates and extends relevant past work, using the same research instrument delivered via online and postal methods to two separate samples randomly drawn from the same population of subscription based customers. Online data The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0263-4503.htm

[ 85 ]

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

collection, the most widely used of the emerging data collection methods previously listed, is compared directly with postal data collection.

Antecedent studies and issues Traditional marketing research is suffering from falling participation rates, rising costs, respondent fears concerning misuse of personal information and managerial issues resulting from the time taken to conduct postal surveys (Jarvis, 2002). In contrast, a number of claimed advantages are put forward for using online data collection methods in survey research, in particular: lower costs; faster turnaround; higher response levels; lower respondent error; broader stimuli potential through the inclusion of colour, graphics and sound; flexibility in the form of adaptive questioning; and even greater enjoyment (Forrest, 1999; Kehoe and Pitkow, 1996). Most of these advantages relate to what has been called the efficiency of online as a data collection method (Weible and Wallace, 1998).

‘ ... a number of claimed advantages are put forward for using online data collection methods in survey research, in particular: lower costs; faster turnaround; higher response levels; lower respondent error; broader stimuli potential through the inclusion of colour, graphics and sound; flexibility in the form of adaptive questioning; and even greater enjoyment ... ’

There is some concern though, that given time, as novelty wears off, online data collection will suffer from the same disadvantages as the traditional methods. For instance, we can already see that the issue of the fear of misuse of respondent information does not disappear, but rather is exacerbated in the online environment (Cho and LaRose, 1999), and that this has contributed to falling confidence in online marketing and research (Australian Privacy Commissioner, 2001). It is instructive to examine each of the claimed benefits, albeit briefly in the context of survey research. In doing so, we acknowledge that there are other forms of online research that present opportunities, such as moderated e-mail groups (Eke and Comley, 1999), however, this paper is restricted to online surveys given the increasing prevalance of this particular methodology. The costs of online surveys are frequently claimed to be lower, provided it is not necessary to continually develop specialised software so that one can treat such

[ 86 ]

development costs as sunk costs (Weible and Wallace, 1998; Zadeh et al., 2000). The need for printing, stuffing, two-way postage, data entry, handling and tracking is removed in the case of online surveys, making them less costly. Turnaround times are frequently reported to be much faster for online surveys. Adam and Deans (2000) conducted a probabilistic e-mail and HTML form survey with 2,976 online businesses and gained 40 per cent of their overall 500 responses within seven days. In the case of the mixed mode e-mail invitation and HTML form study we later report on, 40 per cent of responses (334) were received in the first 24 hours, and over 85 per cent of responses were received within seven days. E-mail surveys, in particular, offer higher response speed than postal surveys (Comely, 1996; Cho and LaRose, 1999; Dommeyer and Moriarty, 2000; Schaefer and Dillman, 1998), however, the choice of a ‘‘technologically sophisticated population’’ (Weible and Wallace, 1998, p. 22) may bring about higher response speed than might be achieved when surveying the general population. Higher response levels are often claimed for online marketing research involving either e-mail or HTML forms or a combination of the two, over postal surveys (Bachmann et al., 1996, 2000; Jackson and DeCormier, 1999). Table I suggests that the situation is not clear-cut, however, and it is difficult to state with statistical certainty that there has been a decline in response levels for both postal and online surveys over time. It has been previously suggested that early novelty seeking behaviour in regards to Internet based surveys may have artificially increased online response levels initially (Zadeh et al., 2000). The evidence would suggest that any ‘‘novelty’’ effect is now wearing off. Another benefit of online surveys is a claimed lower respondent error as determined by such measures as the completeness of response and quality of the response to open text box questions (Weible and Wallace, 1998). One issue in this regard is that HTML form based questionnaires may become too complex for human use simply because the technology permits more intricate rank and rate matrices to be presented than hard copy surveys allow (Zadeh et al., 2000). Analysis of response quality, a vital issue, has been limited to date and thus claims of the superiority of online collection seem premature.

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

Table I Response levels for surveys employing online and traditional data collection methods (percentages) Author(s) and year

Online survey

Traditional survey

AFC (present study) Couper et al. (2001)

21w 41w

Adam and Deans (2000)

17w

Bachmann et al. (2000) Weible and Wallace (1998) Schaefer and Dillman (1998) Tse (1998)

19e 30e 34h 54e 7e

46p 36p 31f 58p 52p

Noh (1998)b Zelwetro (1998)b Besser (1997)c

31e 38e 20e

36p 16p

Couper et al. (1997)c Smith (1997)c

Hertz et al. (1996)a Parks and Floyd (1996)a Mehta and Sivadas (1995)c Opperman (1995)c

43e 8e 13e 27e 53e 14e 17o 69e 33e 40e 49e

Mehta and Sivadas (1995)c Kittleson (1995)c Tse (1995)c

40e 28e 6e

Schult and Totten (1994)c Parker (1992)c Walsh et al. (1992)c Kiesler and Sproull (1986)b Sproull (1986)c

19e 68e 76e 67e 73e

Williams et al. (1997)c Bachmann et al. (1996)c Comely (1996)

46p

71p

75p 66p 15p 96t 64p 26p 33p 64p 77p 27p

38p 75p 87t

Population AFC members Random sample of 1,602 University of Michigan students Australian and NZ online business directors and managers Business school deans and chairs MIS faculty, mainly North American Washington State University faculty Chinese University of Hong Kong teaching and admin staff Public e-mail directoryb nab Former members of the Rural Sociological Society Employees of US government statistics agencies Members of Web consultants association Iowa State University students Business school deans in the USA Purchased list of UK Internet magazine subscribers Health care workers using CDC Wonder USENET newsgroups Users of electronic bulletin board American Association of Geographers Users of an electronic bulletin board International Directory for Health Educators Chinese University of Hong Kong teaching and admin staff Marketing and MIS faculty at US universities AT&T employees Subscribers to online user group Fortune 500 company department Fortune 500 company department

Notes: e = e-mail; h = HTML form; w = e-mail plus HTML form; o = postal plus HTML form; p = post; t = telephone; f = fax Sources: a Cho and LaRose (1999); b Dommeyer and Moriarty (2000); c Schaefer and Dillman (1998) As a final advantage, there is the opportunity to use a broader range of stimuli in the case of HTML forms (Web pages), and with those whose e-mail clients are known to be set to read HTML (Rae and Brennan, 1998). However, little research appears to have been published on this matter to date, although experimentation is underway (Couper et al., 2001). There is also the ability to use filtering processes to increase the flexibility of online surveys by employing adaptive questioning. This approach appears to be used only in a small number of commercial surveys, and little has been reported thus far.

A comparative subscriber study Many of the previous studies in this field have examined the claimed advantages of online surveys. However, there are a number of issues that may affect the validity of such studies (Couper, 2000). Such issues include the use of small samples, low response levels, non-probability sampling and other matters such as the use of reminders and in some cases the application of incentives, which do not allow for ease of direct comparisons. The need for a comparative study that avoided these shortcomings was clear.

[ 87 ]

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

The study reported in this paper includes most of the areas of interest within the one investigation, and involves methodology that enables a sound comparison of online and postal data collection techniques.

Hypotheses A number of hypotheses were developed from the studies discussed earlier and outlined in Table I, and are grouped under the areas of interest in this section of the paper. These hypotheses concern not only managerial issues such as the speed and efficiency of the contrasted data collection methods but also concern research quality issues such as the incidence of missing data and the levels of variation amongst the responses received.

Response level and speed H1a: Online data collection manifests a higher response level than the postal data collection method. H1b: Online data collection manifests a higher level of undelivered questionnaires than the postal data collection method. H1c: Online data collection is less expensive than is the case with the postal data collection method. H1d: Online data collection manifests a higher response speed than the postal data collection method.

Data quality H2a: Online data collection manifests less item-missing non-demographic data than the postal data collection method. H2b: Online data collection manifests a lower omission of personal demographic data than the postal data collection method. H2c: Online data collection manifests a lesser use of end and mid points in the case of scale items than the postal data collection method.

Comparability of resulting data and respondents H3a: Online data collection manifests different mean responses to scale items than the postal data collection method. H3b: Online data collection attracts respondents with different demographic profiles than the postal data collection method.

[ 88 ]

Methodology The study we report involved use of a 70-item, seven page (postal version) questionnaire entitled ‘‘Club name (withheld) member satisfaction survey’’. As the title suggests, the questionnaire dealt with the satisfaction of members of a professional Australian Rules Football (AFL) club, which we hereafter refer to as AFC to maintain the club’s anonymity. In the main, Likert scale items were used, with an 11-point (0-10) ‘‘poor’’ to ‘‘excellent’’ scale. In the year of the study (2001), the AFC claimed over 21,000 paid-up members. The AFC also maintains an ‘‘opt-in’’ e-mail list for those members who wish to interact with the club via this medium. There were 3,900 members on this e-mail list at the time of the study. The club sent an e-mail to the members on this list inviting them to complete an online questionnaire (HTML form), and advising them of a username and password to gain access to the questionnaire. On completing the questionnaire, they were presented with a ‘‘success’’ Web page. The online responses were automatically parsed to a flat file database. Additionally, 1,026 members were randomly selected, and sent a questionnaire through the post to their home address, together with a reply-paid return envelope. The attempt was made to have the online questionnaire appear as close as possible to the printed version, with minimal use of colour (for instructions and the like), except for the arguably preferred use of radio buttons (Couper et al., 2001) presented in a ‘‘scrollable’’ HTML form (Dillman 2000) that disclosed all questions to respondents. The combined responses were then assessed for their representativeness. Both online and postal survey responses were compared with the overall profile of the full list of AFC members. No meaningful differences were found between those who returned the survey and the full membership list, suggesting the sampling process was not biased toward a particular demographic or membership type.

Findings The hypotheses set out in the preceding section are directly tested and reported next.

Speed and efficiency Response details for the two data collection methods are shown in Table II. The postal survey response level is similar to the response achieved by Bachmann et al. (2000). Moreover, the online response level is similar to the levels achieved in the Adam and Deans (2000) and Bachmann et al. (2000) studies. The questionnaire was more easily

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

completed than the Adam and Deans (2000) online questionnaire; taking around eight minutes on average (from dump files and pretesting), compared with 20 minutes noted by Adam and Deans. The response levels might have been expected to be higher, given that this is a subscriber survey dealing with what is an emotive matter for AFL fans. Dillman (1978) observed that interest in a topic can lead to higher response levels. A number of respondents commented on how they appreciated the opportunity to have ‘‘their say’’ in the running of the club, which they had expressed strong attachment to. The postal data collection method resulted in a response level more than twice that of the online data collection method. Although findings on this matter reported in antecedent studies were mixed, there is a clear rejection of H1a in this study. This finding is in line with Dommeyer and Moriarty’s (1999) argument that online data collection methods do not result in higher response levels.

‘ ... findings support past research, which has typically found e-mail non-delivery rates of 35 per cent) . . . and as high as 67 per cent... ’

We employed Webalizer software on the host server to analyse the server log files and are able to report that for the response level of 826, the questionnaire page was visited 1,298 times. Due to the weakness of ‘‘hits’’ as a measure, it is not possible to report whether or not the additional 473 questionnaire hits represent individuals who examined the questionnaire and then left the site, or a few people returning many times, or many people returning a few times to the questionnaire page. It is, however, known that they did not submit the questionnaire. We acknowledge that there are commercial survey programs that do provide this information, however, they were not employed here. The level of non-delivered online questionnaires was found to be unexpectedly high, particularly given the fact that these addresses had been collected in the ten

Table II Response levels for online and postal data collection methods Sent Undelivered (invalid address) Completed Response level (%)

Online

Postal

3,900 700 (18%) 826 21

1,026 6 (< 0.5%) 471 46

months prior to this study, and were volunteered by participants. These findings support past research, which has typically found e-mail non-delivery rates of 35 per cent (Comely 1996), and as high as 67 per cent in one early study (Weible and Wallace, 1998). The mailing list in the present study was, however, more accurate in this regard and we accept H1b, acknowledging the superior reliability of mailing address lists, while also acknowledging that it is not possible to accurately measure the true undelivered rate in the case of postal surveys. A cost comparison is presented in Table III. In this study, costs were minimised by using an existing Linux server running Apache Web server software configured to run a PERL script that parsed HTML (hypertext markup language) form responses to a pipe (|) delimited flat text file. Thus, no fixed costs relating to the host Web server nor Net access costs were involved, as these were treated as sunk costs. The costs involved in setting up for online data collection include the costs of HTML coding the questionnaire, and setting up a secure area using the Apache Web hosting software. Postal distribution costs are much higher given they include envelope stuffing and two-way postage costs. By comparison, the online costs relate solely to computer administrator maintenance costs. Collation costs are negligible in the case of online data collection, and data entry costs are nonexistent. In line with the findings of earlier studies (Mehta and Sivadas, 1995; Weible and Wallace, 1998), we conclude that the cost of postal data collection is higher than where online methods are used, supporting H1c. In particular, there are substantially higher costs incurred upon the return of postal surveys, in terms of both collation and data entry. Correctly designed online surveys employing the Net avoid these costs, making them a far cheaper option per usable response returned. Since mainly semi-fixed costs are involved in online data collection, the unit costs shown would be much lower in a survey involving a larger number of respondents. Antecedent studies indicate that online data collection provides a higher response speed than postal surveys. The findings in this study support the earlier studies in that 75 per cent of the online questionnaires were completed within four days following the e-mail invitation to participate, and 40 per cent were received within the first 24 hours. Questionnaires delivered by post, whilst starting more slowly, were mostly returned in the second week of collection. Over

[ 89 ]

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

40 per cent of postal responses were received on one day in the second week, as shown in Table IV. The early response speed in online data collection is notable. However, when examining response speed it is usual to present average response times when examining postal surveys, and to include weekends and other days when post is not delivered. On this basis we find that H1d is supported, since postal data collection took an average 10.8 days compared to 3.9 days in the case of online data collection (t = 58.9, p = 0.00).

Data quality Comparisons concerning the quality of data obtained using online and postal data collection methods are often reported at a basic level. Typically, judgements on data quality are restricted to analysis of itemmissing data, the number of responses to open text boxes, and item-missing responses for personal demographic questions. It is assumed here that respondents are less likely to provide personal data in online surveys because of a perceived lack of security. In the present study, item-missing data have been analysed in a more complete manner, and analysis of response variance was undertaken to increase the richness of the comparison. Scale items in this study included a ‘‘don’t know’’ or ‘‘refused’ option. Where respondents who completed the HTML form questionnaire did not select a response of their own, the default response was ‘‘don’t know’’. Those who completed the postal survey questionnaire could either answer ‘‘don’t know’’ or leave the question unanswered. A comparison was made of the combined item-missing and ‘‘don’t know’’ responses in the postal survey and the number of ‘‘don’t know’’ responses in the case of the online survey. On the basis of an analysis that excludes responses to demographic questions, H2a is rejected. On average there was a 23 per cent item-missing data/don’t know in the case of the online survey and 19 per cent in the case of the postal survey; which is not a statistically significant

Table III Unit cost comparison for online and postal data collection methods Set-up (e.g. Linus server administrator setup v. printing) Distribution (includes reply costs) Collation (e.g. extraction of data file) Data entry Total per usable response [ 90 ]

Online ($)

Postal ($)

0.91 0.20 0.05 0.0 1.16

0.82 1.40 0.76 1.86 4.84

Table IV Cumulative response for online and postal data collection Days from invitation 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Cutt-off date

Online cumulative response (%)

Postala cumulative response (%)

39.1 52.0 60.2 75.5 81.9 84.5 86.7 88.8 89.9 90.9 92.0 93.1 94.1 95.3 96.2 96.7 97.0 98.3 99.6 100.0

0.0 0.0 0.0 2.3 2.7 9.0 9.0 9.86 10.2 52.6 76.4 85.2 95.1 95.8 97.5 97.5 98.4 98.6 98.6 100.0

Note: a Weekends and non-postal delivery days not included difference. Item-missing data levels are high in both cases, as respondents were not required to comment on aspects of the AFC they may not have directly experienced (e.g. the club Web site and/or family days). Prevention of question skipping in an online questionnaire changes this outcome (Couper et al., 2001). The analysis of responses to personal demographic questions is presented in Table V. Numerically, there was more item-missing data in the postal surveys in four out of the five demographic questions. However, the level of item-missing data is significantly different in only two of the five cases, viz occupation and life cycle stage. These results

Table V Analysis of item-missing data: personal demographic items Demographic variable

Item-missing data (%) Online Postal

Age Income Occupation Life stage Years as member

1.9 16.5 3.4 1.2 6.1

1.6 19.1 6.4 6.2 7.6

Note: Figures shown in bold are significantly different (p < 0.05)

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

suggest that we should accept H2b, in that there is a higher incidence of item-missing data on personal demographic questions in the case of the postal survey. This finding supports Basi (1999), who suggested that those who complete online questionnaires complete more questions than those completing postal surveys. In order to assess the differences in response quality in more depth, analysis of variation in the responses was undertaken. Because of the ease with which questions can be answered online, it was felt that there might be less discrimination on scale items by online participants. This would present itself as a lower variation in responses, with individuals not using extreme points of the scales. An analysis of both the way individuals answered the questionnaires and of the way the sample as a whole answered specific questions indicates that, in general, those responding online are less likely to use the full scale and are using end and mid-points less than mail respondents. Analyses of responses to two questions that illustrate this general trend are presented in Figures 1 and 2, for illustrative purposes.

Figure 1 On-field performance scale item response pattern for online and postal data collection methods

Figure 2 AFC communication performance scale item response pattern for online and postal data collection methods

The scale item responses highlighted in Figures 1 and 2 are significantly different between the online and postal surveys. These findings suggest that H2c should be accepted, and highlight a possible weakness of online data collection.

Comparability of data from online and postal methods The third group of hypotheses relates to whether or not respondents involved in the two data collection methods differ in the way they responded to the questionnaire, and in terms of their demographic profiles. H3a proposes that the means on each of the questions for the two groups would be the same. Analysis of the responses shows that for the non-demographic questions, the groups are statistically different on 26 of the 65 questions (40 per cent of cases). Even when the online survey was matched against a more demographically equivalent group of postal surveys (n = 182), the groups still differed on 22 of the 65 questions (34 per cent). We reject H3a in that the two groups of respondents do not respond to the questions in the same way using the two data collection methods employed. The two groups of respondents also differ in their demographic profile. Income, age, years as a member, occupation and lifestyle category are significantly different between the two groups of respondents (p < 0.01). There are also marked differences within the demographic measures. The online respondents are clearly younger as Figure 3 illustrates. Only the under-19 age group is statistically the same for both collection methods. The online respondents are more likely to be professionals, and report higher incomes, as shown in Figure 4. In terms of income, all categories were significantly different at the p = 0.05 level. We therefore reject H3b as the two data collection methods clearly involve respondents with different demographic profiles.

Managerial and academic implications As Couper (2000, p. 466) points out, ‘‘survey quality is not an absolute but should be evaluated relative to other features of the design (such as accuracy, cost, timeliness, etc)’’. While most marketing professionals – managers and academics alike – can easily see the efficiencies involved in online data collections methods, the sources of error are not always quite as obvious. Survey errors arise involving such matters as sampling, coverage, non-response, and measurement.

[ 91 ]

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research

Figure 3 Comparison of age groups for online and postal data collection methods

Marketing Intelligence & Planning 21/2 [2003] 85-95

The AFC study reported in this paper illustrates that the responses and demographic profile of the two groups of subscribers are statistically different between the data collection methods employed. This difference could be a coverage error (those without an e-mail address or Web access), and/or a sampling

Figure 4 Comparison of income groups for online and postal data collection methods

[ 92 ]

error that arose because not all members of the postal sampling frame were measured due to undelivered mail. Even in this study, where respondents were drawn from the same membership base, there are clear differences between the two groups. Non-response remains an issue, even where an emotive topic such as football fans’ perceptions of their experiences with the club are involved. As antecedent studies show, the issue is not going away with the advent of a new interactive medium, and may become a greater issue as more people become aware that they are paying for their Internet service and therefore for the privilege of being over-surveyed. When and if TMT convergence extends to the point where almost all televisions are interactive devices, then both these issues may be overcome in consumer research. Until that far off day, the issues remain. In this context, it must be remembered that Net access costs are not uniform across all countries, and that the penetration of the technology can in part be explained by differences in broadband availability and access costs (Park and Jun, 2002). Such differences would only exacerbate the inequities found in this research project. There is a need for further research into measurement error where online data collection methods are employed, and particularly where HTML forms are used. The clear message for managers and

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

practitioners from this research is that online surveys should not be expected to produce equivalent results to the more established methods of data collection.

Conclusion In this paper, we have acknowledged that an exciting new interactive, and multi-faceted medium and related technology has spurred innovation in marketing science and practice, and that this innovation includes the field of marketing research. While there are many aspects of online marketing research that require scientific investigation, we have restricted our commentary to the comparative use of the online and postal survey data methods. The study we report is a probabilistic survey of known subscribers, where the same questionnaire was employed. In some ways, the methodology employed might be considered an ideal approach relative to an end-user study that employs non-probability sampling of large numbers of volunteer Net users (e.g. Kehoe and Pitkow, 1996). We conclude that online data collection does offer some advantages over postal collection from an efficiency viewpoint. Although not uniform, there is also evidence to suggest a lower incidence of item-missing data in online surveys on more sensitive personal demographics questions.

‘ ... While . . . differences in thinking may be a function of the demographic differences, it remains clear that online data collection cannot be used as a direct substitute to the postal data collection method... ’

There are, however, a number of unfavourable differences between online and postal methods which should be taken into consideration by researchers employing online methods. The online survey in this study is characterised by a large number of non-delivered questionnaires, and a response level less than half that of the postal data collection method, thus heightening concerns over representativeness. Moreover, clear differences are evident between the two data collection methods in terms of both demographic data and the responses given. The two groups of respondents appear to think differently about the topics presented to them in this study. While those differences in thinking may be a function of the demographic differences, it remains clear that online data collection cannot be

used as a direct substitute to the postal data collection method with this population. The findings in this study are all the more interesting given the nature of the population sampled. The database used to compile the online sample was current, and populated by involved subscribers. It might be said that use of the particular list in this study is an optimal situation, from a marketing research viewpoint. Nevertheless, even in this subscriber survey situation, the significant differences between the two data collection methods suggests that further research on the application of self-administered online questionnaires is warranted. The study we report shows that researchers need to ensure that the practical benefits of using the technology are not outweighed by any survey errors. Better use of the interactive medium used in this study can be made in the longer term, by reducing the likelihood of the errors mentioned in the paper, and reducing non-response in particular.

References Adam, S. and Deans, K.R. (2000), ‘‘Online business in Australia and New Zealand: crossing a chasm’’, in Treloar, A. and Ellis, A. (Eds), Proceedings of AUSWEB2K, The Sixth Australian World Wide Web Conference, Southern Cross University, Lismore, 12-17 June, Cairns, pp. 19-34, available at: http://ausweb.scu.edu.au/ aw2k/papers/adam/index.html (accessed 17 December 2001). Australian Privacy Commissioner’s Web site (2001), available at: Privacy.gov.au (accessed 20 December 2001). Bachmann, D., Elfrink, J. and Vazzana, G. (1996), ‘‘Tracking the progress of e-mail versus snail-mail’’, Marketing Research, Vol. 8 No. 2, pp. 31-5. Bachmann, D., Elfrink, J. and Vazzana, G. (2000), ‘‘E-mail and snail mail face off in rematch’’, Marketing Research, Vol. 11 No. 4, pp. 10-15. Basi, R.K. (1999), ‘‘WWW response rates to socio-demographic items’’, Journal of the Market Research Society, Vol. 41 No. 4, pp. 397-400. Besser, T.L. (1997), ‘‘Non-renewing member survey’’, unpublished report to the Rural Sociological Society, Toronto, cited in Schaefer, D.R. and Dillman, D.A. (1998), ‘‘Development of a standard e-mail methodology: results of an experiment’’, Public Opinion Quarterly, Vol. 62 No. 3, p. 395. Cho, H. and LaRose, R. (1999), ‘‘Privacy issues in Internet surveys’’, Social Science Computer Review, Vol. 17 No. 4, pp. 421-34. Comely, P. (1996), ‘‘The use of the Internet as a data collection method’’, ESOMAR/EMAC

[ 93 ]

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

[ 94 ]

Symposium, available at: http:// Virtualsurveys.com/papers/email.htm (accessed 20 December 2001). Couper, M., Traugott, M.W. and Lamais, M.J. (2001), ‘‘Web survey design and administration’’, Public Opinion Quarterly, Vol. 65 No. 2, pp. 230-53. Couper, M.P. (2000), ‘‘Web surveys: a review of issues and approaches’’, Public Opinion Quarterly, Vol. 64 No. 4, pp. 464-94. Couper, M.P., Blair, J. and Triplett, T. (1997), ‘‘A comparison of mail and e-mail for a survey of employees in federal statistical agencies’’, Annual Conference of the American Association for Public Opinion Research, Norfolk, VA. Dillman, D.A. (1978), Mail and Telephone Surveys. The Total Design Method, John Wiley & Sons, New York, NY. Dillman, D.A. (2000), Mail and Internet Surveys: The Tailored Design Method, Wiley, New York, NY. Dommeyer, C.J. and Moriarty, E. (2000), ‘‘Comparing two forms of an e-mail survey: embedded vs attached’’, International Journal of Market Research, Vol. 42 No. 1, pp. 39-50. Eke, V. and Comley, P. (1999), ‘‘Moderated e-mail groups: computing magazine case study’’, ESOMAR Net Effects Conference, 23 February, available at: http://virtualsurveys.com/ papers/meg.htm (accessed 20 December 2001). Forrest, E. (1999), Internet Marketing Research, McGraw-Hill, Sydney. Hanson, W. (2000), Principles of Internet Marketing, South-Western College Publishing, Cincinnati, OH. Hertz, K., Zook, E., Chitwood, K.S., O’Carroll, P.W. and Friede, A. (1996), ‘‘E-mail versus phone’’, available at: http://ai.mit.edu/ projects/iiip/conferences/survey96/hertz/ wonder.html (accessed 17 December 2001). Jackson, A. and DeCormier, R. (1999), ‘‘E-mail survey response rates: targeting increases response’’, Marketing Intelligence & Planning, Vol. 17 No. 3, pp. 135-40. Jarvis, S. (2002), ‘‘CMOR finds survey refusal rates still rising’’, Marketing News, 4 February, American Marketing Association, Chicago, IL, p. 4. Kehoe, C.M. and Pitkow, J.E. (1996), ‘‘Surveying the territory: GVU’s five WWW user surveys’’, The World Wide Web Journal, Vol. 1 No. 3, pp. 77-84, available at: http:// cc.gatech.edu/gvu/user_surveys/papers/ w3j.html (accessed 12 December 1999). Kiesler, S. and Sproull, L.S. (1986), ‘‘Response effects in the electronic survey’’, Public Opinion Quarterly, Vol. 56, pp. 402-13. Kittleson, M.J. (1995), ‘‘An assessment of the response rate via the postal service and e-mail’’, Health Values, Vol. 18, pp. 27-9. Mehta, R. and Sivadas, E. (1995), ‘‘Comparing response rates and response content in mail versus electronic mail surveys’’, Journal of

the Market Research Society, Vol. 37 No. 4, pp. 429-39. Noh, G. (1998), ‘‘Motivation, design and personal Web presence’’, Conference Proceedings of the Association for Education in Journalism and Mass Communication, August, Baltimore, MD, available at: http://list.msu.edu/ archives/aejmc.html (accessed October 2002). Opperman, M. (1995), ‘‘Email surveys: potentials and pitfalls’’, Marketing Research, Vol. 7, pp. 29-33. Park, C. and Jun, J.-K. (2002), ‘‘A cross-cultural comparison of online buying intention: effects of Internet usage, perceived risks, and innovativeness’’, in Treloar, A. and Ellis, A. (Eds), Proceedings of AUSWEB 02, The Eight Australian World Wide Web Conference, Southern Cross University, Lismore, 5-10 July, Maroochydore, pp. 472-84, available at: http://ausweb.scu.edu.au (accessed 11 July 2002). Parker, L. (1992), ‘‘Collecting data the e-mail way?’’, Training and Development, July, pp. 52-4. Parks, M.R. and Floyd, K. (1996), ‘‘Making friends in cyberspace’’, Journal of Communication, Winter, Vol. 46 No. 1, New York, NY, pp. 80-97. Rae, N. and Brennan, M. (1998), ‘‘The relative effectiveness of sound and animation in Web banner advertisements’’, Marketing Bulletin, Vol. 9, pp. 76-82. Schaefer, D.R. and Dillman, D.A. (1998), ‘‘Development of a standard e-mail methodology: results of an experiment’’, Public Opinion Quarterly, Vol. 62 No. 3, pp. 378-97. Schuldt, B. and Totten, J. (1994), ‘‘Electronic mail vs mail survey response rates’’, Marketing Research, Vol. 6 pp. 36-9. Smith, C.B. (1997), ‘‘Casting the Net: surveying an Internet population’’, Journal of Computer Mediated Communication, Vol. 3 No. 1, available at: http://usc.edu/dept/annenberg/ vol3/issue1/ (accessed October 2002). Sproull, L.S. (1986), ‘‘Using electronic mail for data collection in organizational research’’, Academy of Management Journal, Vol. 29, pp. 159-69. Tse, A.C.B. (1998), ‘‘Comparing the response rate, response speed and response quality of two methods of sending questionnaires: e-mail vs mail’’, Journal of the Market Research Society, Vol. 40 No. 4, pp. 353-61. Walsh, J.P., Kiesler, S., Sproull, L.S. and Hesse, B.W. (1992), ‘‘Self-selected and randomly selected respondents in a computer network survey’’, Public Opinion Quarterly, Vol. 56, pp. 242-4. Weible, R. and Wallace, J. (1998), ‘‘Cyber research: the impact of the Internet on data collection,’’ Market Research, Fall, Vol. 10 No. 3, pp. 19-31. Williams, A.N., Morphew, C.C. and Nusser, S.M. (1997), ‘‘Some considerations for conducting an electronic mail study with university

Heath McDonald and Stewart Adam A comparison of online and postal data collection methods in marketing research Marketing Intelligence & Planning 21/2 [2003] 85-95

students’’, Annual Conference of the American Association for Public Opinion Research, Norfolk, VA. Zadeh, H.S., Adam, S. and Deans, K.R. (2000), ‘‘A technical response to online marketing research issues’’, in Treloar, A. and Ellis, A. (Eds), Proceedings of AUSWEB2K, The Sixth Australian World Wide Web Conference, Southern Cross University, Lismore, 12-17 June, Cairns, pp. 409-20, available at: http://ausweb.scu.edu.au/ aw2k/papers/zadeh/index.html (accessed 17 December 2001).

Zelwetro, J. (1998), ‘‘The politicisation of environmental originations through the Internet’’, Information Society, Vol. 14, pp. 45-56.

Further reading De Leeuw, E. and Nichola, W. II (1996), ‘‘Technological innovations in data collection: acceptance, data quality and costs’’, Sociological Research Online, Vol. 1 No. 4, available at: http://Socresonline.org.uk/ socresonline/1/4/leeuw.html (accessed 7 January 2002).

[ 95 ]

Luxury branding on the Internet: lost opportunity or impossibility?

Francesca Dall’Olmo Riley Kingston University Business School, Kingston upon Thames, UK Caroline Lacroix Pileje SAS, Paris, France

Keywords Brand names, Brands, Internet, Brand image, France, United Kingdom

Abstract The paper addresses the issue of the use of the Internet for the management of luxury brands. We investigate luxury managers’ attitudes towards the use of the Internet to promote their brands and also assess consumers’ opinions and behaviour concerning luxury brands Web sites. Finally, we present the results of a content analysis of a number of luxury brands sites. Any discrepancies between expectations and reality are investigated. Results indicate that reality falls short of the expectations of both managers and customers and that the interactive potential of the Internet is not exploited. Furthermore, there is evidence that the Internet works best as a communication than as a customer acquisition channel for luxury brands.

Marketing Intelligence & Planning 21/2 [2003] 96-104 # MCB UP Limited [ISSN 0263-4503] [DOI 10.1108/02634500310465407]

[ 96 ]

Introduction The Internet has become an essential part of the branding and communication strategies of organisations, ranging from fast moving (e.g. Nestle and Procter & Gamble), to luxury goods (e.g. Chanel and Dior). Peterson et al. (1997) comment that the Internet may provide an efficient medium for accessing, organising and communicating information, while Chaffey (2000) focuses on the Internet’s interactivity potential and on the opportunity to build relationships with individual consumers. Yet, the content of existing Web sites varies widely, from truly interactive to on-line brochures. Likewise, the level of customer service and personalisation provided by different on-line companies is found to vary widely (Hewson and Coles, 2001). In parallel with the increasing number of brands with a dedicated Web site, there has been an escalating interest in the literature on the use of the Internet for brand management (e.g. de Chernatony (2001) and in the recent Journal of Brand Management (Special Issue) 2001). However, much of the literature on successful brand building on the Internet is either prescriptive in nature (e.g. Chaffey, 2000; Clauser, 2001), or does not go beyond the hype of anecdotal evidence (e.g. Chiagouris and Wansley, 2000). No single empirical paper contrasts planning with implementation issues. This paper attempts to fill such a gap, first, by exploring the motivations that have led manufacturers to create Web sites for their brands. We then contrast the managers’ aims with consumers’ perceptions and their use of brand Web sites. Finally we report on a content analysis of a number of such sites. The focus of our research is the luxury sector, for a number of reasons. First, a study The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

by Nyeck and Roux (1997) on whether the Internet had the potential to become a suitable communication tool for luxury brands sets a precedent and offers a means of comparison with the early days of the technology. Second, following Peterson et al.’s (1997) products typology, the Internet may be an effective transaction and communication tool for luxury goods, because of their high cost, relatively low frequency of purchase, high value and high differentiation characteristics. On the other hand, since many luxury goods (e.g. clothes) are experience (rather than search) goods, the Internet may function better as a communication than as a transaction channel.

Background Internet branding A number of recent papers have confronted the issue of brand management on the Internet. Most authors (e.g. Chen, 2001; de Chernatony, 2001; Clauser, 2001) agree that the fundamental rules of branding and the essence of ‘‘the brand’’ itself are the same offand on-line. However, the new technology may require adjustments at the executional level, for example in the way information is presented (de Chernatony, 2001). Moreover, the degree of control exerted over the brand may need relaxing so that, for instance, unmoderated chat forums and communities are built within the brand’s Web site (de Chernatony, 2001). This would enable companies to engage in a more immediate and unfiltered dialogue with their publics (Ind and Riondino, 2001). The sites’ interactivity features are crucial to Internet marketing success, according to Chaffey (2000). Ind and Riondino (2001) also highlight that while a brand’s presence on the Web The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0263-4503.htm

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

should be conceived in a way that is sympathetic to the technology and the environment, the Web site should be consistent with the whole brand presentation and should reflect the overall brand identity. However, and in spite of the emphasis on the interactive, real-time features of the Net for branding success, studies by Hewson and Coles (2001) and Molesworth and Jenkins (2002) reveal that many organisations fail miserably in the implementation of such features, resulting in customers’ needs and expectations not being met.

Managing luxury brands Nueno and Quelch (1998) define luxury brands as: . . . those whose ratio of functional utility to price is low while the ratio of intangible and situational utility to price is high.

Phau and Prendergast (2000) point out that while ‘‘luxury’’ is a subjective concept: . . . luxury brands compete on the ability to evoke exclusivity, a well-known brand identity, [. . .] brand awareness and perceived quality.

Another concept related to luxury brands is the ‘‘rarity principle’’: the prestige of the brand gets eroded, if too many people own it (Dubois and Paternault, 1995). This creates a paradox for luxury brand management (Roux and Floch, 1996): the company needs to maximise its profits but can never sell or standardise too much. Luxury brands organisations have to maintain a fragile equilibrium between high exposure and awareness but a controlled level of sales. To maintain their dream value and avoid the risk of commoditisation, ‘‘luxury brands must be desired by all, consumed only by the happy few’’ (Kapferer, 1996). Because of their high prices, luxury brands must not only deliver the best but also ‘‘extensively customise them (their products) in order to prove how customer-focused they are’’ (Kapferer, 1996). This will be an issue explored in this research, as the Internet allows a great deal of personal customisation.

The Internet and the luxury sector According to Kapferer (2000), e-commerce presents both an opportunity and a threat for luxury brands. The Web will give an opportunity to buy to those potential customers who do not have nearby access to one of the few distribution outlets of a luxury brand or who feel intimidated from entering ‘‘such temples of luxury’’. At the same time, for brands whose image is based on excellence and on highly selective distribution and information channels, there is a real threat of ‘‘commoditisation’’, arising

from being referenced by other portals and being pirated by those who buy merchandise on the grey market. On the other hand, this ‘‘commoditisation’’ effect may not be as strong in all markets. For example, Phau and Prendergast (2000) note that the popularity of a brand does increase its dream value among Asian customers. An exploratory study by Nyeck and Roux (1997), at the time when the Internet started to take off in Europe, offered the framework for the current study. They carried out in-depth interviews with six luxury brand managers concerning their opinions on the Internet. Some thought that there was no compatibility between the Internet and ‘‘luxury’’ or luxury brands consumers. Others were positive about the Web and believed it would bring an innovative twist to the brand image, especially as they thought that the luxury sector had to be ‘‘avantgarde’’. They pointed out that the site had to be luxurious and have ‘‘perfect aesthetics’’. With regards to selling through the Web, some said it clashed with the very selective distribution channels the brands used and feared the possible increase of fake products sales. Others saw an opportunity to reach far foreign markets in a sector where globalisation is vital. Sales through the Internet were judged as useful for repeat buyers where the overall luxury experience can be slightly less important. As for the consumers, they liked to enter into the shops to purchase luxury goods, but felt they lost the environmental quality when visiting the luxury brands’ Web sites. Managers also mentioned this point. Consumers also felt it was sometimes hard to use the medium and that they could not master it, although they expected luxury brands to be online. Some thought it was not for consumption, except for a repeat purchase or for well-known accessories or products that do not require to be tried on first, such as champagne.

Methodology Because the use of the Internet has become much more widespread since 1997, it is likely that more luxury consumers are now Web users, at least for the younger layers of the target markets. At the same time, most luxury brands nowadays have Web sites. Hence, it seemed appropriate to further Nyeck and Roux’s (1997) research at the current, more advanced stage of development of the Internet, both as a management tool and as a consumer activity. The study is divided into three parts. The first part

[ 97 ]

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

investigates luxury managers’ attitudes towards the use of the Internet to promote their brands. The second part assesses consumers’ opinions and behaviour concerning luxury brands’ Web sites. Finally, we present the results of a content analysis of a number of luxury brands sites. Any discrepancies between expectations and the reality are also investigated.

Managers’ attitudes In-depth interviews were used to elicit the attitudes and opinions of five luxury brands professionals. Because most of the headquarters of the luxury houses contacted are in France, the face-to-face interviews were carried out in Paris, in French. The interviews were tape-recorded, transcribed and content analysed in French. All efforts were made to ensure that the translation into English, for the purpose of this paper, was as accurate as possible. All organisations interviewed had an operational Web site. Company A, company D and company E specialise in clothes but also have a range of perfumes. These three companies belong to the LVMH group and employ, respectively, 250, 150 and 600 people, world-wide. The respondents were respectively a brand manager, a marketing and sales manager, and a multimedia specialist. Company C is a small company that started in 1991 and employs two people in the showroom and workshop in Paris. The respondent was the founder and owner. Company B is a fashion company, created in 1976 and employing 40 people. The managing director was interviewed. The interviews aimed to obtain managers’ insights in the following areas: for how long have they had a Web site; what was the Web site aiming to achieve; are they attempting to sell online; and what impact was it deemed to have had.

Consumers’ perceptions A self-completion questionnaire (in English) was employed to collect data on the attitudes of luxury brands’ consumers towards the use of the Internet both in general and for luxury brands. The only pre-requisites were that the respondents had bought a luxury brand over the previous 18 months and that they had Internet access. The questionnaire was administered in two ways: face-to-face and through the Internet. Twenty-nine responses were collected by means of store intercept in four London department stores. The same questionnaire was sent by e-mail to a convenience sample of 112 acquaintances and addresses authorised by them. This yielded 52 valid questionnaires (a response rate of 46

[ 98 ]

per cent). Checks revealed no significant differences between the responses of the two cohorts, hence the entire sample of 81 respondents will be analysed all together. Sixty per cent of respondents were female, and 75 per cent were between 18 and 44 years old.

Web content analysis A total of 54 luxury brands’ sites were content analysed, following the framework provided by Carroll (2001). The brands were chosen because they were referred to in most academic articles on luxury (e.g. Kapferer, 1996); they were often mentioned by respondents in the piloting of the survey; and the managers of some of these brands were interviewed. The sites were in six categories: ‘‘accessories’’ (three sites), ‘‘automotive’’ (three sites), ‘‘couture’’ (26 sites), ‘‘perfume and cosmetics’’ (nine sites), ‘‘food and drinks’’ (seven sites) and ‘‘jewellery’’ (four sites). The brands that operate in more than one sector (e.g. couture and cosmetics) were classified according to the emphasis given in the home page of the site. Brands with different addresses for different products had an entry for each one. The Web sites were analysed for their general, visuals, promotion, communication, interactivity and sales features.

Results The managers’ perspective Here we discuss how the interviewees viewed the use of the Internet for brand management.

The Internet The motivations given by the managers for building a Web site varied. Some had put a Web site in place because it was a fashionable thing to do (e.g. companies A and D). For others (e.g. company E), the Web project had been championed by someone within the organisation who had realized the new medium’s potential. The issue of selling on the Internet was then explored. Several problems were highlighted in the responses. The first problem was how to show the garments in a truthful and complimentary way on the Internet. The director of company B remarked that: They [the clients] should not be lost or disappointed by what they are going to see.

Not being able to touch and feel the garments was another problem, for instance: Personally I don’t think I would buy on the Net, you need to see and touch the garments.

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

And . . . there is the sales assistant’s advice (Marketing manager, company D).

The impossibility of touching the product was part of a more general issue, the holistic experience that is put forward by the literature. For the brand director of company A, an ‘‘atmosphere’’ can be communicated through the Internet ‘‘but this requires a very high level of professionalism in this sector. This cannot be put together by an IT person’’. However, all the other respondents felt that the holistic experience could not be transmitted through the Web. For example: You remove the ‘‘pleasure purchase’’, that is to go in the shop, being received there, being recognized . . . All this aspect disappears (Director, company B).

When probed to elaborate what products could be sold online, clothes were generally excluded from consideration. In company C attempts at selling on the Web had been made, but the results had not been satisfactory: Well I sold six or seven pieces when there were about seven or eight thousand visitors on the site . . ..

The director of company B felt that the Web site was catering more for the needs of the buyers of the diffusion lines, than for the couture customers. For him, the Internet suppressed the spontaneity in the purchase, just because you have to make a conscious decision to get on the site.

Impact of the Web site The respondent for company C said that the site does not have much impact on his brand image: It remains very limited. . . . the brand promotion has not really taken off on the Net, maybe later if the phenomenon increases, but here it is essentially the catwalks, the press and the open show-room that bear the brand identity.

In contrast with this view, the managers in companies D and E believed that their sites had a positive influence on the brand image: It is a beautiful site. I rather think that it reinforces the brand image.

When asked if the Web site has had an impact on company E, the multimedia specialist said: Yeah, I believe so, especially for our target anyway. For our target, it allows us to modernize the image. As we attract a new audience, we are going to modernize the image, which is going to be relayed in the new communication campaign, just being worked on. The new campaign reflects this will of

modernizing the brand. It is truly a part of the communication tool.

The final area of interest is the impact of the Internet Web site in the communication strategy of the luxury companies. For the respondent in company D: It is true that the consumers can communicate with us much more easily since we have the site. Still, I believe that the site and the communication are very difficult to manage, once again it is a group project.

In contrast, for the multimedia specialist in company E, the Internet site: is totally integrated, hence the presence of the site here, internally of the communication department, it is because it is not a window site, it’s an integral part of the company’s communication tool.

Company E’s multimedia specialist added that there is a strong attempt at relationship building through the Web site, where a community is created, the users can subscribe and receive a newsletter. Additionally, the Web site offers a magazine that changes every 15 days and ‘‘. . . (the customers) can also take part in the forums’’. In contrast, the other brand managers said that the relationships are established and maintained through human contact. It is the sales assistant and shop managers who learn to know the customers and have sometimes an independent customer database, for example: Relationships are made in the boutiques and in the multi-brand shops, the franchises. This is where the relationship happens (marketing/sales manager, company D).

The consumers’ perspective Use of the Internet Over two-thirds of the 81 respondents claimed to access the Internet on a daily basis, with a further 30 per cent doing so ‘‘at least once a week’’. The main access point was from home (75 per cent), although there was an overlap with ‘‘at work’’ access (51 per cent). The most important reason for surfing the Net was ‘‘to look for information’’ (40 per cent), followed by ‘‘for leisure’’ (19 per cent) and ‘‘to shop’’ (18 per cent). Other reasons included ‘‘to chat’’ and ‘‘e-mail’’. The respondents’ perceptions regarding the advantages and the disadvantages of the Internet were also investigated. The three dominant reasons selected by the respondents were: ‘‘it saves time’’ (20 per cent); ‘‘it contains all the information needed’’ (19 per cent) and ‘‘it is interactive’’ (18 per cent). Among the disadvantages, ‘‘not secure’’ and ‘‘too slow’’ came on top (17 per cent each), followed by ‘‘can’t always find what I am looking for’’ (15 per cent) and

[ 99 ]

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

‘‘expensive’’ (10 per cent). The Internet’s superficiality, coldness and impersonality followed. Other respondents mentioned ‘‘spam and banners’’ among their dislikes. Of respondents 50 per cent did not feel comfortable with leaving their credit card details on the Internet, and a further 14 per cent admitted to leave their details, despite feeling anxious about security. Men were more concerned about security than women.

Use of luxury brands Web sites One-third of the respondents had never visited any luxury brand Web site. Among the respondents who had visited a luxury brand Web site in the previous 18 months, the reasons most cited were ‘‘looking for information on the brand’’ (22 per cent), ‘‘looking for information on the product’’ (21 per cent) and ‘‘seeking personalised advice’’ (12 per cent). Only 5 per cent had accessed a luxury brand Web site for the purpose of making a purchase. The reasons for not buying on-line are reported in Table I. In 40 per cent of the cases, the respondents could not find what they were looking for. This is probably due to the fact that many sites do not have any sales capabilities and the ones that do usually have limited ranges. Consistently, with some of the comments made by the managers, the second reason for not buying online was that people like the shopping experience (35 per cent) and also like to touch and try (11 per cent). The security issues came up again, for 22 per cent of the respondents. Among the ‘‘other reasons’’, respondents indicated that when visiting a site they were ‘‘just looking’’ and not shopping online. Respondents were then asked to indicate what they liked about the luxury Web site they visited most. Sites that were easy to access and navigate were the most liked (33 per cent). The design of the site and the visuals are also important factors and were appealing to 30 per cent of the respondents, indicating that the brands managed to communicate their ‘‘beauty’’ or ‘‘magic’’ (Kapferer, 1996) through the Internet.

Table I Reasons for not buying online

[ 100 ]

Personalisation and interactivity followed, at 22 per cent and 17 per cent respectively. When asked to indicate what they would expect from a luxury Web site, the aesthetics and presentation features again came up at the top, followed by the opportunity to communicate with the organisation via e-mail and, once again, the ease of access and navigation (see Table II). The high scores in Table II indicate that the users have high expectations from luxury Web sites, especially the aesthetics (mentioned by 60 per cent), communication via e-mail (47 per cent) and information on the product (44 per cent). Consistently with other results of this survey, only a quarter of respondents ticked the possibility to purchase the goods. When asked what type of luxury items they would consider buying online, respondents answered as follows (Table III). Here we can see that the first two categories are food and wine and cosmetics. It should be noted that these items are usually the ones carrying the least involvement in the purchase. This is also true to an extent for the accessories category. Finally, respondents were asked whether or not they felt that the luxury brand they purchased most often, online or offline, was involving them in a relationship. The

Table II Features expected in a luxury Web site Feature

%

Aesthetics Communication via e-mail information on products Easy to navigate Information on brands (PR events) Personalised advice (club, newsletters) Secure connection Community (boards, chat) Clear postage and package charges Real time communication World-wide availability Purchase possibility

60 47 44 40 36 33 32 31 28 27 27 26

Table III Luxury items the respondents might buy online

Reason

%

Category

%

Could not find what I wanted I like the shopping experience Concerns over security It looked too complicated I need help and advice Conditions of purchase were unclear Like to touch and try Other reasons

40 35 22 15 13 11 11 55

Food/wine Cosmetics Accessories Clothes Jewellery Decoration Cars Other items

37 33 24 20 19 17 10 11

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

respondents answered as follows: 43 per cent do not feel they have a relationship with the brand, 47 per cent are loyal to several brands they like and only 9 per cent answered ‘‘this brand makes me feel special’’. Only 1 per cent chose ‘‘I feel loyal to one brand’’.

segmentation of the site was found in half of the sites studied. This shows that the marketing opportunities offered by the Internet are not exploited to their full potential by half of the brands studied. These findings are summed up in Table IV.

Web site visuals

Luxury brands’ Web sites content analysis A total of 54 Web sites in six categories (accessories, automotive, couture, food and drink, jewellery and perfumes) were content analysed in August 2001 for their general, visuals, promotion, interactivity and sales features. These are discussed below.

General features A number of aspects were considered, including: ‘‘Is the Web site operational?’’; ‘‘Requirement for a special shareware?’’; ‘‘About us’’ section; ‘‘History’’; ‘‘Product list’’; ‘‘Media relations’’; ‘‘What’s new section’’; ‘‘Site map’’; ‘‘Segmentation’’ (that is the existence of different sections according to the type of consumer); ‘‘Help’’ and ‘‘Press releases’’. Only 70 per cent of the sites were operational, that is, the site was present on the Internet and had a content. For example, Hermes (www.hermes.com) only had one page with name, logo and flashing visuals, saying it was going to become operational soon. Then the items related to the brand itself were evaluated. ‘‘About us’’ was present in 46 per cent of the sites, ‘‘History’’ of the company and ‘‘What’s new’’ were found in half of the sites studied. Some sections relating to the navigation were also explored. The ‘‘Site map’’ and ‘‘Product list’’ buttons were found in 61 per cent of the sites. The help section was found in only 19 per cent of the luxury Web sites visited. The more interactive ‘‘search the site’’ option was found in almost a quarter of the 54 Web sites. Some more marketing oriented features such as ‘‘Data collection’’, and ‘‘Segmentation’’ were also present in the sites. ‘‘Data collection’’, or any device to gather knowledge on the user (signups, subscriptions, surveys, etc. . . .) were found in half of the sites studied. ‘‘Segmentation’’ was found when the site was segmented in several sites, each designed for a different audience, like male/female, or country segmentation. For example, the Givenchy Web site (www.Givenchy.fr) is segmented for fashion and perfumes, male and female, and then a separate site was set up for the oblique perfume, with a much younger image. The

This section is dedicated to the imagery of the Web site, since aesthetics play a major part in luxury, according to the definition of Roux and Floch (1996) and also in the views of the managers we interviewed. Overall, all the sites made good use of visuals. A total of 72 per cent of them possess some continuous pictures. Some sites had more sophisticated illustrations such in www.chanel.fr or www.dior.com, using the Flash1 technology and even used the video possibilities of the Internet (13 per cent), like in www.armani.com. It should be noted that this kind of features increases the perception of interactivity of the medium but can be confusing for the user. For example, in the Chanel site where the buttons are not precisely distinguished, the navigation is made more difficult. Another drawback of the presence of many pictures and animations on the site is the downloading time: about 50 per cent of sites took more than one minute to download. Music is often used as well (32 per cent).

Promotion This section is about the promotions used on the Web site. These are used in a general context to trigger the purchase, but in the Internet environment, they are especially used to render the Web site ‘‘stickier’’. In all the sites visited in this study, general promotions were often limited to a simple communication message. In 20 per cent of the cases, free downloads were offered, such as branded wallpapers, screensavers and

Table IV Luxury brands’ Web site general features General features

%

Web site operational Product list buttons Site map History What’s new Data collection Segmentation Special s/ware required? Abous us Search Help Press releases Media relations

70 61 61 50 50 50 50 48 46 24 19 15 6 [ 101 ]

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

imagery from current advertising campaigns (Dior, Jean Paul Gaultier). If other brands were advertised on the sites, they were brand extensions and never other companies. This is consistent with Carroll’s study (2001). There were no sites with coupons (they are usually faxed, e-mailed or readily printable from the site and allow a discount in the shops). This is not really surprising, as a monetary promotion would clash with the premium nature of the brands studied here.

Interactivity and communication features The interactivity is the feature that differentiates the Internet from the other media. Consistently with Carroll (2001), we consider here both navigation features and the features that allow a two-way communication between the consumer and the brand. About a third of the sites visited offered links to the organisation’s sites in other countries, or to the group’s corporate site (for example www.clarins-paris.com is linked to www.clarins-group.com), or to sites from the brand’s other activities (for example the cosmetics site of Givenchy is linked to the couture site). Pop-ups were found in a further 30 per cent of sites. They were used to draw attention to a message (for example, the popup at chanel.fr for the launch of the new jewellery collection), but often increased the downloading time and could appear intrusive. The drop down menus allow an easy navigation. They are not always present (only 26 per cent of the cases) but the new technologies like Flash1 allowed betterlooking navigation tools. Of sites 11 per cent had restricted areas that could only be accessed with a username and password. They entail a subscription to the site, for example to a brand’s club like at www.shiseido.com. Such clubs allow the creation of an ‘‘elite’’ group, a concept dear to the luxury brands. Finally, the most common two-way communication feature between the consumer and the brand was the ‘‘e-mail us’’, where the users could send an e-mail or HTML message to the company (60 per cent of sites). Sometimes the topic of the message could be selected, facilitating the dispatching of the message when it arrived. This was linked to the other features like ‘‘Webmaster’’ and ‘‘staff e-mails’’, allowing the users to address their messages to a specific person. Those were rather rare, however, as they were found in only 9 per cent and 4 per cent of the sites, respectively. Only one-quarter of sites offered a contact address and phone number for the ‘‘head office’’ of the company. The other

[ 102 ]

possibilities offered for communication were the ‘‘sign-ups’’ (50 per cent of cases), asking for the users’ e-mail address and allowing them to: . receive a newsletter (e.g. Guerlain); . become a member of a club (e.g. Midiron at Dior.com); . obtain personalised advice (e.g. Chanel cosmetics); and . participate in the site’s community (e.g. Kozen.com). Finally, over one-fifth of sites contained some kind of survey. These features are of importance as they are linked to the customer relationship management.

Web site’s selling features While only 15 per cent of the sites visited had a selling facility, two-thirds of sites provided an online brochure showing the products, sometimes in the form of an online catwalk (e.g. www. Jeanpaulgaultier.com), or like a proper catalogue with extensive description of the products, for example at www.montblanc.com. Prices, packaging and postage costs were not always easy to find and were given only at a later stage of the order. The refund policy was clear for only 9 per cent of the sites, but when they were present, they were very detailed and easy to understand (www.jeanpaulgaultier.com). Finally, only 6 per cent of the sites allowed world-wide dispatching. Most of them were solely for the USA (e.g. www.tiffany.com). This is probably due to delivery problems but, as the sites are available from anywhere across the world, it may generate some frustration. All the sites offering selling facilities had a secure transaction declaration, reassuring the consumers about the encryption of their details. As the online selling is not very popular, 67 per cent of the sites had a store locator, giving the user the address of the nearest boutique.

Conclusions In spite of a more widespread use and acceptance of the Internet over the last few years, managers’ opinions regarding its suitability for luxury branding have not changed much since the times of Nyeck and Roux’s (1997) research. The majority of managers we interviewed viewed the Internet as an unavoidable development posing both a threat and an opportunity. Provided that the brand’s holistic experience could be maintained by means of the design

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

and aesthetics of the Web site, managers agreed that the Internet could serve as a communication tool for their brands. At the same time, they thought that the Internet’s impact on the brand image would be less important than the impact of traditional means of communication and, more importantly, it could not substitute the ‘‘pleasure experience’’ of entering one of their shops. Moreover, managers did not believe that the Internet could provide a viable selling tool for their products, particularly for clothes. The Internet was perceived as removing not only the spontaneity of shopping, but also the pleasure of touching and feeling beautiful material and of entering their ‘‘temples of luxury’’. Furthermore, nothing could substitute the level of service and the human contact provided in the stores. To a considerable extent, consumers’ perceptions matched the managers’ opinions. One-third of our respondents never visited a luxury brand Web site and only a handful had actually ever bought a luxury item through the Web. Most would not even consider doing so, even for less involving search goods such as wine. At the same time, consumers’ expectations of luxury brands’ Web sites were high, especially in terms of their visual effects, aesthetics and interactive features. They perceived the luxury brands’ Web sites as a means of getting information and advice about the luxury brands they liked. Over half, however, did not think the Internet meant they had a relationship with their favourite brands. In contrast, the more technologically minded managers (e.g. the multimedia specialist in company E) perceived the Internet as a powerful means of establishing relationships with customers by means of the interactive possibilities available through the medium. However, the content analysis of luxury Web sites revealed that while most sites scored high on visuals, few offered truly interactive customisation and segmentation features. The results from the different parts of our research are consistent in pointing out that luxury brands’ organisations have not yet exploited to the full the interactive potential of the Internet and that consumers themselves are still reluctant and suspicious of the medium, at least as a shopping tool. Most consumers still perceive shopping for luxury items as a pleasure in itself and as a gratifying experience. Our results are also broadly consistent with Peterson et al.’s (1997) analysis of the potential impact of the Internet as a channel

intermediary for different kinds of products. Based on Peterson et al.’s (1997) framework, it would seem that, for luxury brands, the Internet is very likely to fulfil a communication and information role, but less likely to be used as a customer acquisition channel. For luxury brand managers, the problem remains of how to give incentives to people to visit their Web sites. More efforts seem to be needed not only in communicating to existing and potential customers the exclusive benefits of accessing the sites, but also in providing real benefits for doing so (e.g. customer clubs; customised advice, etc.).

References Carroll, C. (2001), ‘‘Enabled communications – a content analysis of the top 100 Irish companies corporate Web sites’’, in String, C. (Ed.), Academy of Marketing Annual Conference Proceedings, Cardiff University, July. Chaffey, D. (2000), ‘‘Achieving Internet marketing success’’, The Marketing Review, Vol. 1, pp. 35-59. Chen, S. (2001), ‘‘Assessing the impact of the Internet on brands’’, Journal of Brand Management, Vol. 8 No. 4, pp. 288-302. Chiagouris, L. and Wansley B. (2000), ‘‘Branding on the Internet’’, Marketing Management, Vol. 9 No. 2, pp. 34-8. Clauser, R.C. (2001), ‘‘Offline rules, online tools’’, Journal of Brand Management, Vol. 8 No. 4, pp. 270-87. de Chernatony, L. (2001), ‘‘Succeeding with brands on the Internet’’, Journal of Brand Management, Vol. 8 No. 3, pp. 186-95. Dubois, B. and Paternault, C. (1995), ‘‘Observations: understanding the world of international luxury brands’’, Journal of Advertising Research, Vol. 35 No. 4, pp. 69-76. Hewson, W. and Coles, F. (2001), ‘‘eCRM: measuring the effectiveness of Web channels’’, Journal of Targeting, Measurement and Analysis for Marketing, Vol. 9 No. 4, pp. 370-87. Ind, N. and Riondino, M.C. (2001), ‘‘Branding on the Web: a real revolution’’, Journal of Brand Management, Vol. 9 No. 1, pp. 8-19. Kapferer, J.N. (1996), ‘‘Managing luxury brands’’, Journal of Brand Management, Vol. 4 No. 4, pp. 251-60. Kapferer, J.N. (2000), ‘‘How the Internet impacts on brand management’’, Journal of Brand Management, Vol. 7 No. 6, pp. 389-91. Molesworth, M. and Jenkins, K. (2002), ‘‘Gratifications obtained by young people from commercial and non-commercial Internet content’’, in Tynan, A.C. et al. (Ed.),

[ 103 ]

Francesca Dall’Olmo Riley and Caroline Lacroix Luxury branding on the Internet: lost opportunity or impossibility? Marketing Intelligence & Planning 21/2 [2003] 96-104

[ 104 ]

Proceedings of Academy of Marketing Annual Conference, Nottingham. Nueno, J.L. and Quelch, J. (1998), ‘‘The mass marketing of luxury’’, Business Horizons, Vol. 41 No. 6, pp. 61-9. Nyeck, S. and Roux, E. (1997), ‘‘WWW as a communication tool for luxury brands: compared perceptions of consumers and managers’’, in van Raiij, B. et al. (Ed.), Proceedings of the Second International Research Seminar on Marketing Communication and Consumer Behavior, La Londe Les Maures.

Peterson, R.A., Balasubramanian, S. and Bronnenberg, B.J. (1997), ‘‘Exploring the implications of the Internet for consumer marketing’’, Journal of the Academy of Marketing Science, Vol. 25 No. 4, pp. 329-46. Phau, I. and Prendergast, G. (2000), ‘‘Consuming luxury brands: the relevance of the ‘rarity principle’’’, Journal of Brand Management, Vol. 8 No. 2, pp. 122-38. Roux, E. and Floch, J.M. (1996), ‘‘Ge´rer l’inge´rable: la contradication interne de toute maison de Luxe’’, De´cisions Marketing, Vol. 9, pp. 15-23.

Web site characteristics and business performance: some evidence from international business-to-business organizations Despina A. Karayanni Department of Business Administration, University of Patras, Athens, Greece George A. Baltas Department of Marketing, Athens University of Economics and Business, Athens, Greece

Keywords Web sites, Business-to-business marketing, Marketing communications, Internet, Business strategy, Interaction

Abstract This paper addresses the effectiveness of Web site capabilities. It is intended to empirically determine the importance of site properties for Internet sales performance. More specifically, we consider the relationship of characteristics such as navigability, interactivity, multimedia design and marketing communications content with sales attributed to the Internet. Advanced econometric modeling of cross-section data on businessto-business firms demonstrates that interactive responsiveness of the site, enabling customized interactive programs and club membership, fast-downloading multimedia, such as frames and animation, corporate information stressing corporate positioning and active information submission on behalf of the target audience may influence sales performance. Important implications for optimization of business-tobusiness Internet strategies are also considered.

Marketing Intelligence & Planning 21/2 [2003] 105-114 # MCB UP Limited [ISSN 0263-4503] [DOI 10.1108/02634500310465416]

Introduction

The organization of the paper is as follows. The next section lays the groundwork by briefly reviewing the relevant literature. In the third section we provide a conceptual framework of site effectiveness and develop research hypotheses. The fourth section is concerned with empirical analysis and research findings. A concluding section summarizes the paper.

In recent years, new technologies have created new and innovative means of interacting with customers. In particular, business-to-business organizations see the Internet as a complement to their existing networks, which diffuses at exponential rates. For example, according to Forrester Research (2000), inter-organizational oe-commerce in North America will exceed $2.7 trillion by 2004. In view of the substantial and rapidly growing Internet audience, several firms have quickly adopted the Internet as a means to conduct marketing communications economically and efficiently. Nonetheless, while most companies have felt compelled to establish an Internet presence, they have failed to create a Web site with substantive features and capabilities (Breitenbach and Doren, 1998). For example, many Internet marketers take the printed corporate brochure and put it onto the screen but that technique simply does not work in this interactive multimedia environment (Monaco, 1997). Firms need to improve the appeal of their Web sites because of the increased online competition in industrial markets. Therefore, a critical question to consider is what methods are effective in reaching the online business-to-business market. As will be seen later, inspection of the relevant literature suggests that very little is currently known about these issues. Given the lack of empirical research on site effectiveness, this study attempts to assess the effectiveness of various Web site elements on sales performance. More specifically, we examine the relationship of Web site design features and information content with sales attributed to the Internet presence of the firm.

The literature suggests that the Internet has provided more than just a new medium through which organizations can communicate with the public. The WWW allows bi-directional marketing and offers wider and deeper material and richer advertisement content (Ellsworth and Ellsworth, 1995). Nevertheless, little progress has been made toward the development of frameworks regarding the use of Internet for marketing purposes, besides some case studies (Hoffman and Novak, 1996; Palmer and Griffith, 1998; Morgan, 1996). Yet, most of these studies fail to recognize the unique properties of the Web, which may be highly synopsized into the attribute of interactivity (Morgan, 1996), whereas very little is actually known about how to use this new technology. Some researchers contend that the role of this new technology in Internet marketing activities may be determined by the interaction among the firm’s market offering, its marketing activities and technical characteristics (Palmer and Griffith, 1998). Thus, optimal site design involves both marketing and technical concerns and is driven by the core market offering of an organization. Therefore both marketing and technical issues should be addressed by the Internet marketers.

The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0263-4503.htm

Background

[ 105 ]

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

Web sales performance The most common measurement of site effectiveness is the number of visits to the site. This, however, will not completely reveal the site effectiveness, as the visitors may be just casual browsers who should not be considered as potential customers or may just have stumbled upon the site (Ghose and Dou, 1998). Hoffman and Novak (1996) argue that, in addition to examining purely exposure measures such as click-through rates, the degree to which the visitor interacts should be also measured. They propose such measures as duration of browsing time, the depth, or number of assessed pages, and the number of repeated visits. Fortunately, there is a way to generate such information through common gateway interface (CGI) programming (Ong, 1995). However, the real impact of promotional content on the Web may not be accurately known because of difficulties associated with such measurement (Murphy, 1996; Dreze and Zufryden, 1997). Although the number of pages accessed and time spent on a Web site should be expected to relate to visitors’ purchase behavior, it would be desirable to use an objective function expressed directly, in terms of sales, or profitability. This could be accomplished by tracking individual purchases in addition to the aforementioned data (Dreze and Zufryden, 1997). Thus, while sites may be judged on a number of criteria, for the purposes of our study we use the Web-based sales in order to measure performance.

Conceptual framework In this section, we shall consider site properties and develop specific hypotheses regarding the effect of such site attributes on site performance. As noted above, firms should move beyond the brochure look and develop rich and well-organized sites (Palmer and Griffith, 1998). It remains to be seen, however, how site capabilities might determine performance and which of them are particularly effective.

Interactivity of the Web site interface Interactivity of the site relates to the optimization of hypermedia attributes by providing custom-made solutions, allowing users to customize preferences, and enabling the organization to provide solutions to users who have provided personal information (Muylle et al., 1999). It is now technically feasible to offer interactive functions to customers who might otherwise require the assistance of staff.

[ 106 ]

The Internet may change the way firms interact with their customers and thus the way they initiate, develop and terminate relationships with them (Blattberg and Deighton, 1991). Some authors view the Net as a means of tying the customers to the company through the development of detailed customer databases and the use of relationship marketing techniques (Shani and Chalasani, 1992; Kara and Kaynak, 1997). In particular, Berthon et al. (1996) have suggested that the level of interactivity in a Web site might be critical in getting surfers involved in the marketing communication process. It would thus seem desirable to explore whether interactivity plays any significant role in affecting the overall quality of business-to-business organizations’ Web sites. Specifically, interactivity may be discriminated into two facets, i.e. memory storage and organizational response to individual needs and preferences. The role of the Internet in this process lies in identifying the address of a customer, initiating the dialogue, maintaining the dialogue and storing data on customer behavior, which can be used to customize marketing strategies. First, the customer databases that may be built via the Internet constitute the memory of the customer relationship – a record of every message and response between the firm and each address (Blattberg and Deighton, 1991). By adding artificial intelligence the system can design new messages, and even product offerings, at the individual level, to reflect everything learned from past interactions. However, the impact of Web-database management upon industrial sales has not yet been empirically tested. We hypothesize that: H1: Memory storage of industrial buyers’ characteristics and buying behavior, as expressed by keeping customer databases and ability of customers to have access to them, will be related to higher industrial sales performed through the Internet. The second facet of interactivity, i.e. organizational response through customized offerings, may be equally supported by the two-way interaction between Web site and the visitor. The Internet enables demassification, a term meaning the degree to which a specific message can be conveyed to each individual belonging to a large audience (Rogers, 1991). Demassification usually implies that a certain degree of control over a communication system has shifted from the message producer to the message consumer. Thus, visitors may control the manner in which they utilize the

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

site. For example, customers may be guided through a catalogue of products and services that is most suited to their profiles. Such an interactive Internet system is cheaper and far more effective than traditional mailshots, since it is customized. As a matter of fact, it is expected that the Web may be the ultimate medium for market segmentation at the individual level, targeting unique visitors with individually designed promotion, content, products and services (Dreze and Zufryden, 1997). Establishment of customer user groups is another way of increasing organizational responsiveness at a unit of one. These newsgroups are many-to-many communication tools, which personify virtual forums where users can exchange ideas. Customers can share their experience with others and also ask questions which may be answered by others. According to Berthon et al. (1996), user groups may build a community atmosphere in the site, which in turn may make the site a satisfying and adaptive marketplace. Such forums would stimulate a feeling of belongingness for users, which, in turn, would solidify their affiliation with company (Breitenbach and Doren, 1998). Thus, Web sites should facilitate customers to join affiliated to the company online customer clubs, in order to exchange information and share stories. In conclusion, it would be interesting to examine the direct relationship between Web site interactive response element and sales performance. Stated formally, H2. Organizational response achieved through the Web site’s interactivity, as expressed by interactive elements, i.e. the Web site’s interactive program resulting in customized company offerings, and online club membership and forums, will be related to higher industrial sales performed through the Internet.

Navigability Navigation refers to the hypertext connection of the multimedia content and determines ease of information retrieval and site well-organization (Bornman and Solms, 1993). Evans and King (1999) suggest that organized, easy-to-navigate pages should be a primary consideration of good Web site design. More precisely, navigability and accessibility are related to information search costs. Hoque and Lohse (1999) argue that subtle changes in the user interface design influence information search costs, thus they should be considered by Web designers in order to create more effective Web sites. The use of search engines, site

maps and indices provides the ability to link pages throughout the Web site and to hyperlink to related information, which enhances the ability of Web site users to navigate the page and identify information quickly. Stated formally: H3. Navigability of the Web site, determined by such facilities as a site index which would be always in display, a search engine, a site map and personal contact possibilities, will be related to higher industrial sales achieved through the Internet.

Multimedia design The Internet offers capabilities similar to those of newspapers (i.e. text and graphics), radio (i.e. audio) and TV (i.e. video), all in one concise package (Breitenbach and Doren, 1998). Thus, the Internet has hybrid characteristics that combine those of the print, broadcast, outdoor and direct response media. Moreover, Dreze and Zufryden (1997) have found that use of frames contributes to the explanation of number of pages accessed, which coincides with prior beliefs that frames should promote the viewing of more pages by more effectively providing links to them. Advancements in graphics technology allow more creativity on the Web, with developments such as animated ads and banners on screen. New inventions in Web technology today such as the so-called Quick Time movies and xlivescreen may allow multimedia techniques to make much more sophisticated and integrated presentations in the form of video, sound, music, graphics and text. This enhanced capability not only makes visitors’ experiences more fun and stimulating (Spalter, 1996) but also gives visitors the flexibility to activate only a needed part of the presentation (Ghose and Dou, 1998). An example is the presence of banners, which are clickable links to a specific advertisement in a company’s homepage and are usually based on creative graphics animation. Ghose and Dou (1998) state that site attractiveness would increase with interactive functions – the term used to describe Web multimedia applications. Based on media richness theory, they contend that the multimedia interactive format should provide more capabilities than text, sales brochures and catalogues. However, Hoque and Lohse (1999) argued that different ways of presenting information alter information search costs. Indeed, the new media are not frictionless. Some information search tasks take longer online, while others are shorter, depending on how the Web is designed. Unfortunately, there is

[ 107 ]

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

no objective technique for measuring information search in the context of site design. Both speed and effectiveness are impaired when information is provided using video or audio, as compared to other subtle interfaces such as animation, text or graphics. Information search in audio or video takes more effort than searching in text or graphics. The additional effort is attributed to more waiting time to load an audio, or video clip. Notwithstanding, system response time is widely acknowledged as one of the major sources of frustration, annoyance, impatience and stress during human-computer interaction (Kuhman et al., 1987; Ducoffe, 1996). Indeed, it has been suggested that there should be significant correlation between Web site download speed and Web user satisfaction (Muylle et al., 1999). Jayawardhena and Foley (2000) identify a couple of additional disadvantages in program downloading (e.g. an audio-video application), which may be synopsized as risk of importing unwanted components, like viruses, as well as unwillingness to sacrifice hard disk space. Moreover, the value of audio-visuals depends on the target market, the product offering, and the target’s equipment, as video and audio hardware, may not be resident in a user’s PC (Evans and King, 1999). In this sense, graphics may be more effective than audio and video, since photos, diagrams and illustrations are sought by more users (Dreze and Zufryden, 1997). In conclusion, designers of commercial sites face a myriad of decisions about how best to organize and present product information, often without knowing how their design may influence information search costs and subsequent buyer behavior. However, the direct relationship between site design characteristics and sales have not yet been empirically tested. We hypothesize that: H4. The use of multimedia, i.e. use of frames, animation, banners and audiovideo elements that are built in an organizational Web site will be related to higher industrial sales achieved through the Internet.

Marketing communication content Communication is a key aspect of successful marketing through the Internet as it may enable a firm to achieve a competitive advantage in the marketplace (Berthon et al., 1996). Business-to-business Internet marketing communications should heavily stress the company identity and interpersonal customer communication (Karayanni, 2001). Thus, we focus on communication activities that may influence both the company image

[ 108 ]

and the customer interaction, i.e. on company, and customer related information exchange, respectively.

Company-related information Drawing on the pertinent literature we elicit that there is a risk associated with using the WWW as a marketing channel (Gelormine, 1995). Defining and illustrating the company’s identity reinforces buyers’ trust in the company, acting as the counterbalance of risk. Indeed, buyers are more likely to buy products from a company they know and with which they can identify (Breitenbach and Doren, 1998). Companies have the opportunity to present the history and the fundamental culture of their organization online. Information concerning corporate financial health, employees’ profile and affiliation ties with other vigorous organizations, may act as a control mechanism which may guarantee stability in future business transactions. Public relations may also contribute to building the corporate image. For example, users may be more inclined to purchase products and services from socially responsible companies. Thus, companies could make users aware of any charitable activities, community efforts, and environmental projects, in which the company is involved. Companies could also provide some support products and services for free. Support products are those products or services that enhance or complement another offering of the company. In particular, links to unbiased news and reviews of company offerings are useful, especially for purchases with high buyer involvement. For example, Federal Express allows users to track packages to ensure their arrival. The more complimentary products a firm can offer, the more industrial buyers will value their original purchase as well as the site (Breitenbach and Doren, 1998). Finally, a site may provide added value services such as instant calculation of required monthly payment for a product, online contests, electronic postcard delivery, and information of general interest (e.g. the weather or stock exchange rates). We thus postulate that: H5. Company-related information provided on the Web site which aims at building the corporate image, will be related to higher industrial sales achieved through the Internet.

Customer-related information exchange Another way of reducing risk and uncertainty is by transferring control of information flows towards the user of information, i.e. the industrial customer. New technology makes customer

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

participation and feedback easier (Lovelock, 1991). Primarily, the WWW has been regarded as a means by which the buyer can pull information (Young, 1996). Customer support and communications can be achieved via e-mail and associated file transfer activities (FTP) and WWW pages. For example, answering frequently asked questions (FAQ) from customers and prospects cuts down the amount of staff time spent on answering inquiries. This type of support offers some advantages over the telephone or fax. Not only may companies find it a cheap, quick and logistically efficient way to provide customer support, but also firms are often able to use the WWW to coordinate users to help each other (Morgan, 1996). The e-mail and the WWW are also effective for the dissemination of timesensitive information, such as product catalogues, product fact sheets, press releases, speeches, research findings, company newsletters, etc. Moreover, the Internet allows companies to carry out some marketing research functions such as customer surveys, product and service perception surveys, advertising effectiveness surveys and some experimental niche activities such as limited testing of new products and services. A few sites enable an online feedback form. On this form, visitors can comment on anything related to the company, either product related, or company related. Berthon et al. (1996) discussed the merits of such functions. This function provides a way of establishing a dialogue between site visitors and the firm. Visitors can express their opinions on a company’s new product. Companies can invite and encourage customers and potential customers to join in a dialogue using e-mail, and can start or participate in newsgroups about their products, industry and applications (Morgan, 1996). In this manner, customers are likely to feel that the firm values their opinions and their perceptions might therefore be positively enhanced. Thus, Breitenbach and Doren (1998) urge companies to design sites that allow users to offer comments, suggestions and complaints via e-mail. In conclusion, greater involvement of customers in information processing through submitting opinions and complaints to the firm would result in greater buyer control, leading to increased sales. Thus, we hypothesize that: H6. Customer-related information exchange available through the Web site will be related to higher industrial sales achieved through the Internet.

Methodology Sampling and empirical analysis As noted above, the principal aim of the present study is to assess the effectiveness of business-to-business Web sites and, in particular, the impact of the Web site’s characteristics on industrial Internet sales. Toward that end, the research study involved a two-stage data collection procedure, i.e. an e-mail survey for assessing Web sales effectiveness and on-line observation for examining Web site technical characteristics and communication content. First, we conducted the e-mail survey, following the guidelines suggested by Avlonitis and Karayanni (2000). Specifically, a random sample of 700 US and European business-to-business companies was drawn from the Yahoo! directory of the WWW (http://www.yahoo.com/Business and Economy/Companies/Industrial_Supplies) which was used as the sampling frame. First, a prenotification e-mail was sent, which was answered appropriately by 320 companies, accounting for a 45.7 percent of the initial random sample. Within the next two months, two waves of e-mailed questionnaires followed, accompanied by a cover letter that was personally addressed (i.e. not a multiplerecipient letter) to the senior marketing director of each one of the 320 companies. The short research instrument comprised only two sets of questions: 1 demographic characteristics (i.e. size, type of industry and years of Web presence); and 2 percentage of total sales attributed to the Web presence. In total, 112 companies responded, producing a total response rate of 35 percent, which compares favorably with response rates of similar e-mail surveys (Avlonitis and Karayanni, 2000). In order to examine the possible presence of non-response bias, the sample was divided into groups of early (first wave) and late (second wave) respondents. The late respondents imitated the nonrespondents, whereas the early respondents simulated the respondents in the sample (Armstrong and Overton, 1977). Then, t-test analysis was performed on the variable ‘‘percentage of sales attributed to the Web’’, using the demographic characteristics as independent variables, in order to examine any differences in these measures between early and late respondents. No meaningful differences were identified, which gave us the base to ascertain that the sample was homogeneous. As a result, response bias seemed to present no problem in the study.

[ 109 ]

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

As a next step, data on Web site characteristics and communication content were collected by empirical observation of the Web sites of the 112 respondent firms, conducted between July 15 and August 30 of 2000. The computer devices used were Pentiums II 366 MHz, with 64 MB of RAM and 4 MB graphics cards, running Windows 2000. Access on the Internet was through a cable modem running conservatively at 300 to 500 Kbps. Web browsing was undertaken using the latest popular browser versions, Netscape Communicator 4.6 and Internet Explorer 5. Sites that were unusually slow or did not load were double-checked at another day. Given the lack of previous studies on the importance of both site characteristics and marketing communication content in Internet marketing, the examined variables were considered in a rather exploratory fashion. To avoid subjectivity, all attributes relate to objective Web site features, whose presence can be assessed with minimal, if any, ambiguity. The distribution of the examined diverse site aspects over the sample units, indicating presence or absence, is presented in Table I. The site characteristics fall in three categories of respective variables. The first group of site attributes is concerned with the interactive capabilities of the Web site. The second group of site characteristics deals with ease of navigation, whereas, the third category of site attributes concerns multimedia design of the site, such as audiovisional capabilities and fast downloading multimedia. The marketing communication content falls in two more categories of respective variables. Thus, the fourth category pertains to corporate information provision, whilst the fifth category concerns information exchange between the firm and the customer. Table I demonstrates that presence frequencies vary considerably over different characteristics, suggesting heterogeneous integration of the examined capabilities into the sites. Therefore, such attributes may, at least in principle, provide differentiating dimensions to the Internet presence of the firm. It remains to be seen, however, if site capabilities affect the sales performance of the firm.

Factor analysis In order to obtain a structural model of implementable dimensions, application of data reduction methods is necessary. The five categories of 26 variables are factoranalyzed and factor scores are extracted by the principal components method. Following standard practice, we select factors with eigenvalues greater than unity for further

[ 110 ]

analysis. Factor analysis results are presented in Table II. For expositional clarity, insignificant variable coefficients (< 0.50) are not reported. Specifically, the performed factor analysis produced: 1 two dimensions for interactivity, which may be labeled as ‘‘memory storage’’ (databases building) (FR1) and ‘‘organizational responsiveness’’ (i.e. customized offerings and online club membership) (FR2); 2 one unidimensional construct for navigability (FR3); 3 two dimensions for the multimedia design, which may be labeled as ‘‘audiovisional multimedia’’ (FR4) and ‘‘fast downloading multimedia’’ (i.e. frames, animation and banners) (FR5); 4 two dimensions for company-related information items, which may be labeled as ‘‘corporate information provided to the shareholders’’ (FR6) and ‘‘image-building corporate information’’ (FR7); and 5 two dimensions for the customer-related information, i.e. ‘‘customer information provision’’ (FR8) and ‘‘customer information collection’’ (FR9).

Econometric analysis and findings The preceding factor analysis has reduced the dimensionality of the data to a level at which statistical modeling of structural relationships can be conducted. We may now proceed to test hypotheses H1-H6, considering the impact of Web site capabilities on company sales arising from the Internet. To that end, we regressed the dependent variable ‘‘percentage of sales attributed to the Web’’ against the nine variables that were produced by the aforementioned factor analysis. The dependent variable contained zero values due to companies lacking online sales capabilities or interorganizational transactions, implying the occurrence of zeros with non-zero probability, which cannot be accomodated by standard regression models, based on standard distributions. In such cases, the Tobit or censored regression model seems to be more appropriate in bundling composite distributions, containing a discrete part for zero and a continuous part for non-zero observations, as is well established in modern econometrics (Maddala, 1983), which led us to use it for our regression analysis. Table III presents the maximum likelihood estimates of the Tobit model of Web site performance. Inspection of Table III reveals the exogenous variables that have a significant impact on Internet sales, thus placing

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

Table I Distribution of site characteristics and marketing communication content Variable name

Percent present

Percent absent

Interactivity Customised user information Customer databases Access to company’s databases Customer’s login Real-time interactive elements (e.g. club membership)

I1 I2 I3 I4 I5

10.7 48.2 20.5 15.2 9.8

89.3 51.8 79.5 84.8 90.2

Navigability Site map Search engine facility Site index Personal contact possibility

N1 N2 N3 N4

24.1 31.3 72.3 48.2

75.9 68.8 27.7 51.8

Multimedia design Audio Video Frame technology Animation technology Banners

D1 D2 D3 D4 D5

8.0 8.0 63.4 20.5 22.3

92.0 92.0 36.6 79.5 77.7

Company information Company history Public relations material Financial status of the firm Company news and plans Added value services of the firm Employee profiles and CVs

C1 C2 C3 C4 C5 C6

86.6 44.6 23.2 70.5 54.5 16.1

13.4 55.4 76.8 29.5 45.5 83.9

Customer information exchanges Collection of customer views and critiques Structured online surveys Adaptive information query Exposition of customer cases Announcements of promotion activities Automatic FAQ/customer support

E1 E2 E3 E4 E5 E6

33.9 8.9 83.9 18.8 16.1 47.3

66.1 91.1 16.1 81.3 83.9 52.7

Characteristics

support in a subset of the hypotheses of the study. More specifically, H1 was not supported, since interactivity resulting in memory storage of customer data was found to be unrelated to Web sales performance. On the contrary, organizational response was highly related to Web sales, thus placing evidence on H2. It appears that the mere collection of customer data may not lead to enhanced sales performance. Rather, it is the implementation of gathered customer data into customised organizational offerings that may increase the effectiveness of the Web sites. H3 was not supported, since the navigation aids of the site (FR3) appear insignificant, perhaps because such features are crucial only for inexperienced visitors, which might be important in the consumer business (Farquhar and Balfour, 1998). H4, was marginally supported. Specifically, audiovisional multimedia (FR4) was

insignificant, suggesting that audiovisual capabilities might not be an important factor in interorganizational Internet marketing. This finding, however, is rather intuitive since the industrial buying process is less, if anything at all, sensitive to such excitement generating features. This is also reflected in the modest presence of audiovisual capabilities, which are found in only 8 percent of the sites (see Table I). However, fast downloading multimedia (FR5) were marginally related to Web sales, thus supporting H4. Regarding H5, this was highly supported by our findings, since both FR6 and FR7, reflecting company information provided to shareholders and image-building corporate information, respectively, are positive and significant. Finally, H6 was partially supported by the results. Specifically, variable FR8, concerning provision of information to the customers, such as promotion

[ 111 ]

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

Table II Factor analysis summary Factor loadings

Factor

Variable

FR1

Interactivity: memory storage I2: Customer databases I3: Access to company’s databases I4: Customer’s login

0.642 0.737 0.857

2.088

Interactivity: organisational response I1: Customised user’s information I5: Real-time interactive elements (e.g. club membership)

0.692 0.851

1.049

Navigability N1: Site map N2: Search engine facility N3: Site index N4: Personal contact possibility

0.776 0.743 0.465 0.512

1.633

Audiovisual multimedia D1: Audio D2: Video

0.969 0.961

1.903

Fast downloading multimedia D3: Frame technology D4: Animation technology D5: Banners

0.508 0.684 0.698

1.210

Corporate information provided to shareholders C3: Financial status of the firm C4: Company news and plans C5: Added value services of the firm C6: Employee profiles and CVs

0.641 0.686 0.679 0.621

1.795

Image-building corporate information C1: Company history C2: Public relations material

0.773 0.719

1.213

Customer information provision E3: Adaptive information query E5: Announcements of promotion activities E6: Automatic FAQ/customer support

0.516 0.627 0.814

1.463

Customer information collection E1: Collection of customer views and critiques E2: Structured online surveys

0.788 0.789

1.303

FR2

FR3

FR4

FR5

FR6

FR7

FR8

FR9

Table III Maximum likelihood estimates of the censored (tobit) regression model Factor

Variable

FR1 FR2 FR3 FR4 FR5 FR6 FR7 FR8 FR9

Interactivity: memory storage Interactivity: organisational response Navigability Audiovisual multimedia Fast downloading multimedia Corporate information provided to shareholders Image-building corporate information Customer information provision Customer information collection Intercept LogL (max) LogL (restricted)

Note: t-values in parentheses [ 112 ]

Parameter 0.273 4.961 –2.799 –0.313 3.530 4.347 3.216 0.292 4.778 2.721 –1.938 –2.973

(0.160) (3.021) (1.529) (0.192) (1.944) (2.418) (2.050) (0.167) (2.978) (1.802) (0.190) (0.194)

Eigenvalue

announcements and adaptive information queries, was insignificant. In light of the present empirical evidence, such site features fail to affect Internet sales performance in a significant and systematic fashion. Inversely, FR9 concerning collection of customer information, such as customer views and on-line surveys, appears to be associated with highly effective sites.

Conclusions and implications The literature on site design and capabilities is expanding but still lacking in empirical work. This study attempted to shed some light on these issues, by empirically testing a number of hypotheses regarding the

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

relationship of specific site characteristics and marketing communication elements with sales attributed to the Internet. Among various site characteristics we selected interactivity, navigability and multimedia elements to be used as independent variables. As marketing communication elements we selected company-related information and customerrelated information exchanges. The explanatory power of the aforementioned variables on Internet sales were examined through censored regression analysis based on maximum likelihood estimation. The research findings provided support for a subset of the study’s hypotheses, resulting in a number of fruitful conclusions and implications for both academics and practitioners. First, as far as the Web site’s interactivity is concerned, we found that memory storage was unrelated, whereas its counterpart, i.e. organizational responsiveness, was related to site effectiveness. It appears that mere database memory storage of industrial customer characteristics and purchasing behavior does not affect sales. Rather, it is the exploitation of information stored in databases, through customized end-user programs and club memberships, that may enhance the Web site’s performance. A profound explanation may be that interactive programs, enabling customized end-user Web site applications, reduce information search costs, thus influencing industrial buyer behavior. At the same time, the aforementioned positive relationship of interactivity responsiveness with Internet sales provides evidence that, apart from business relationships, the Internet many-tomany communication model may also influence actual business performance (i.e. sales). The above results seem to advocate for an Internet many-to-many marketing model, vis-a`-vis the traditional one-to-many marketing model, which introduces new dynamics to industrial marketing. In regard to multimedia design characteristics, we found that sophisticated multimedia, comprising audio and video elements, failed to explain cross-sectional variability of the dependent variable, which is consistent with prior evidence supporting that complex, high downloading-time sites may restrain the site’s visits and buying behavior. Inversely, fast downloading multimedia (i.e. frames, animation and banners) appear to be more effective, resulting in increased site performance. Navigability was found to be insignificant, which implies that industrial buyers are rather conscious and skilled information seekers. Thus, although site organization may be an important property, it does not

appear to be a determinant of site performance, at least in inter-organizational transactions. As far as the marketing communications are concerned, both company-related information variables were found to have high explanatory powers on Web sales. Indeed, provision of favorable companyrelated information promotes the corporate identity and positioning in the market. The empirical findings stress the critical importance of promoting corporate identity which elevates trust among interacting organizations, acting as the counterbalance to the high flux and anarchy prevailing on the Internet. Increased trust, in turn, favorably affects transaction costs of product and financial exchanges and cooperation, thus resulting in increased sales. Finally, between the two dimensions of customer-related information exchange, i.e. information provision and collection, only the later appears to be an important determinant of site performance. The positive relationship of customer information collection with Internet sales may be attributed to two reasons. First, information gathering from customers may facilitate implementation of marketing orientation and identification of customer satisfaction, which are highly related to business performance. In addition, it may result in increased customer involvement, leaning information control towards the customer, which reduces perceived risk of making transactions over the Internet. As the results of this study suggest, the reverse direction of marketing communication flow, that is, from the target audience (the customer) to the firm should be considered an important property of Internet marketing.

Limitations A word of caution, however, might be in order. The present findings are mainly exploratory and it would be inappropriate to generalize too far from a single study. In particular, the direction of the structural relationships cannot be applied crudely to all situations. For example, it might well be the case that some firms with great Internet success are encouraged to invest more on Internet marketing and exhibit sites that are better designed and integrated. In such cases, Internet success may determine, at least in part, site capabilities, rather than vice-versa. As is always the case in social sciences, the definition of causal directions requires experimental manipulation of the datagenerating processes, and might not be feasible in the light of observational data.

[ 113 ]

Despina A. Karayanni and George A. Baltas Web site characteristics and business performance: some evidence from international business-to-business organizations Marketing Intelligence & Planning 21/2 [2003] 105-114

[ 114 ]

References Armstrong, S. and Overton, T. (1977), ‘‘Estimating nonresponse bias in mail surveys’’, Journal of Marketing Research, Vol. XIV, August, pp. 396-402. Avlonitis, G. and Karayanni, D. (2000), ‘‘The impact of the Internet on business-to-business marketing – examples from American and European companies’’, Industrial Marketing Management, September, Vol. 28 No. 9, pp. 441-59. Berthon, P., Pitt, L. and Watson, R. (1996), ‘‘Marketing communication and the World Wide Web’’, Business Horizons, Vol. 39 No. 5, pp. 24-32. Blattberg, R. and Deighton, J. (1991), ‘‘Interactive marketing: exploiting the age of addressability’’, Sloan Management Review, Fall, pp. 5-15. Bornman, H. and Solms, H. (1993), ‘‘Hypermedia, multimedia and hypertext – definitions and overview’’, Electronic Library, Vol. 11 No. 4/5, pp. 259-68. Breitenbach, C. and Doren, D. (1998), ‘‘Valueadded marketing in the digital domain: enhancing the utility of the Internet’’, Journal of Consumer Marketing, Vol. 15 No. 6, pp. 55875. Dreze, X. and Zufryden, F. (1997), ‘‘Testing Web site design and promotional content’’, Journal of Advertising Research, March/April, pp. 7791. Ducoffe, R. (1996), ‘‘Advertising value and advertising on the Web’’, Journal of Advertising Research, Vol. 36 No. 5, pp. 21-35. Ellsworth, J.H. and Ellsworth, M.V. (1995), Marketing on the Internet, Wiley, New York, NY. Evans, J. and King, V. (1999), ‘‘Business-tobusiness marketing and the World Wide Web: planning, managing, and assessing Web Sites’’, Industrial Marketing Management, Vol. 28, pp. 343-58. Farquhar, B.J. and Balfour, A. (1998), ‘‘Consumer needs in global electronic commerce’’, Electronic Markets, Vol. 8, pp. 9-12. Forrester Research (2000), available at: www.forrester.com/ER/Events/Upcoming/ Overview/0,1550,303, FF.html Gelormine, V. (1995), ‘‘Selling in cyberspace’’, Success, May, Vol. 42 No. 4, pp. 61-3. Ghose, S. and Dou, W. (1998), ‘‘Interactive functions and their impacts on the appeal of Internet presence sites’’, Journal of Advertising Research, March-April, pp. 29-44. Hoffman, D. and Novak, T. (1996), ‘‘A new marketing paradigm for electronic commerce’’, Journal of Marketing, July. Hoque, A. and Lohse, G. (1999), ‘‘An information search cost perspective for designing interfaces for electronic commerce’’, Journal of Marketing Research, August, Vol. 36, pp. 387-94. Jayawardhena, C. and Foley, P. (2000), ‘‘Changes in the banking sector – the case of Internet banking in the UK?’’, Internet Research:

Electronic Networking Applications and Policy, Vol. 10 No. 1, pp. 19-30. Kara, A. and Kaynak, E. (1997), ‘‘Markets of a single customer: exploiting conceptual developments in market segmentation’’, European Journal of Marketing, Vol. 31 No. 11-12, pp. 873-95. Karayanni, D.A. (2001), ‘‘A model of relational communication strategies through the Internet, network relationships and business performance: towards a ‘network market concept’’’, Journal of Marketing (paper under review). Kuhman, W., Bouscein, W., Schaefer, F. and Alexander, J. (1987), ‘‘Experimental investigation of psychophysiological stressreactions induced by different system response times in human-computer interaction’’, Ergonomica, Vol. 30 No. 6, pp. 933-43. Lovelock, C. (1991), Service Marketing, PrenticeHall, Englewood Cliffs, NJ. Maddala, G.S. (1983), Limited Dependent and Qualitative Variables in Econometrics, Cambridge University Press, Cambridge. Monaco, B. (1997), ‘‘Marketing on the Internet’’, CNBC Business News (transcript), January. Morgan, R. (1996) ‘‘An Internet marketing framework for the WWW’’, Journal of Marketing Management, Vol. 12, pp. 757-75. Murphy, P. (1996), ‘‘On-line ads effective? Who knows for sure?’’, Marketing News, September, p. 23. Muylle, S., Moenaert, R. and Despontin, M. (1999), ‘‘Steve, Rudy and Marc, introducing Web site satisfaction: an integration of a qualitative pilot study with related MIS research’’, Working Paper, Owen Graduate School of Management, Vanderbilt University, Nashville, TN. Ong, P. (1995), ‘‘Practical aspects of marketing on the WWW’’, MBA dissertation, University of Sheffield, Sheffield. Palmer, J. and Griffith, D. (1998), ‘‘An emerging model of Web site design for marketing’’, Comunications of the ACM, March, Vol. 41 No. 3, pp. 45-51. Rogers, E. (1991), The Critical Mass in the Diffusion of Interactive Technologies in Organizations, Harvard Business School, Boston, MA, pp. 245-63. Shani, D. and Chalasani, S. (1992), ‘‘Exploiting niches using relationship marketing’’, Journal of Consumer Marketing, Vol. 9 No. 3, pp. 33-42. Spalter, M. (1996), ‘‘Maintaining a customer focus in an interactive age: the seven I’s to success’’, in Forrest, E. and Mizerski, R. (Eds), Interactive Marketing, NTC Business Books, Lincolnwood, IL. Webster, F. (1996), ‘‘The future of interactive marketing’’, Harvard Business Review, November-December pp. 151-62. Young, D. (1996), ‘‘Considerations in the design of electronic commerce solutions’’, EC article, Microsoft Internet Explorer, January.

E-relationships – emergence and the small firm

Thomas O’Toole Waterford Institute of Technology, Waterford City, Ireland

Keywords Electronic resources, Small firms, Relational databases, Competitive advantage, Information technology

Abstract The e-business model has become an integral feature of business practice. Its existence has created an extra electronic layer to business relationships. This has facilitated close partnerships to add unique and complex electronic components to their relationships that are not easily copied by other firms. The emergence of e-relationships is detailed and their value creating potential explored. The erelationship concept is then applied to the small firm. The strategic and implementation barriers that inhibit small firms from capitalising on such relationships are examined. Erelationships are found to offer competitive advantage to few small firms. Critical to an erelationship’s success is the relationship orientation of the partners.

Marketing Intelligence & Planning 21/2 [2003] 115-122 # MCB UP Limited [ISSN 0263-4503] [DOI 10.1108/02634500310465434]

E-business is changing market structures and the economics of competition. Technology and business strategy have converged to the extent that e-business is an essential component of strategy (Peppard, 2000). The dot.com phenomenon has become a feature of most businesses. Consumers, buyers, suppliers are all being connected as a product of the e-business revolution. This new business model makes it necessary to focus on the electronic in relationships. The article examines the emergence of the e-relationship and the potential value it offers to a business. The concept is then applied to the small firm where strategic and implementation barriers to its adoption are outlined. E-relationships incorporate a trading relationship that uses both computer and telecommunications technologies, and an added value layer of electronic embeddedness only available to close relational partners. Cunningham and Tynan (1993) capture the spirit of e-relationships in their e-trading concept. This has been added to by Geiger and Martin’s (1999), among others, ideas on Internet relationship marketing. E-relationships can be considered a further layer of interaction that partners can ‘‘wrap around’’ each other to create walls of competitive superiority. To implement e-relationships can require many technologies and systems such as electronic customer relationship management (e-CRM) or virtual private networks. However, e-relationships are not technology specific but are partner-based. Managing value in an e-relationship depends on the level of co-involvement in the relationship. The stronger and closer the relationship the more unique its electronic element becomes.

Electronic relationships that offer benefits to both the seller and buyer are offering a potential source of competitive advantage to partners that can exploit them to create unique social and economic ties. Information technology has had a huge impact on business in general but also has had major effects on the nature of how partners interact (Donaldson and O’ Toole, 2002). Information rich exchanges can be made at the touch of a button. Companies have been able to mass customise their products by selling them on the Internet. Organisations can use virtual transactions to maximise value; whole supply chains are interconnected by technology. Each core relationship variable has been, in some way, affected by technology. Technology, through the routinisation of many operational processes, has minimised conflicts that arise with individual transactions, for example, product availability and delivery can be checked on-line. Information technology is changing the power balance in many relationships. For example, the creation of on-line communities has challenged firms to respond to strong, more vocal, united consumer groups (McWilliam, 2000). The Internet has spawned new businesses that have managed to grow through trust with those with whom they trade, for example, E*trade (www.etrade.com) in share dealing and Amazon (www.amazon.com) in book retailing. The growth of e-relationships is being forced by companies’ adoption of the Internet and e-mail into their business process. Traditional inter-organisational systems (IOSs) are being transferred to the Internet. In 2000, Ford and General Motors announced the switching of their supply management system from electronic data interchange (EDI) to the Internet (Ford’s system is Auto-Xchange (www.auto-xchange.com) and

The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0263-4503.htm

Introduction

[ 115 ]

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

GM’s TradeXchange www.gmtradexchange. com)). The attraction of the Internet over traditional IOS is its openness, flexibility and low cost. The impact of the Internet is also seen through the effect e-mail is having in creating bonds across organisations and between individuals who otherwise would not communicate. The content of electronic communication via e-mail varies significantly across organisations. A closer relationship is likely to have more volume of communication as well as a richer content.

Maximising value in e-relationships Value added in a relationship, or by a relationship, may give it a competitive edge (Ramirez, 1999; Lorenzoni and Lipparini, 1999). A close relationship with a consumer, supplier and even with a competitor can represent a strategic asset (Madhavan et al., 1998; Johnson, 1999). Value creation in a relationship can be seen as a strategic option, or in the case of a close relationship, as a strategic imperative. E-relationship components can add an extra layer of value in a relationship that is difficult to imitate. This electronic value appears to meet, in close relationships, the criteria for sustainable relationship advantage outlined in Dyer and Singh (1998). Much of the initial developments in e-relationships have been to automate existing transactional, production and service functions. This automation adds value in terms of cost reduction and possibly higher reliability in service operations. A lot of this value is probably relatively easy for other firms in any given sector to copy and thus is of a short-term nature. By taking a relationship management perspective to the study of e-relationships we can distinguish other levels of value created. Value can be added in information exchanged that allows knowledge to be pooled for joint advantage, in dense and rich communication patterns, in integrated business processes, through new products and services that are created or developed electronically, and in bonding with consumers that creates extra symbolic value in brands. The potential to transform organisations with technology is real (Brady et al., 2000; Knol and Stroeken, 2001), but in relationships is dependent on another, at least one, partner. Using e-relationships as a tool to better manage partner interaction provides a myriad of possibilities. E-relationships require a trusting culture for exchanges of strategic value to take place. Not all relationships may be worth this but

[ 116 ]

there is considerable potential where it is judged possible. In the absence of absolute co-operation maximising communication and information value using technology is limited. Not all a firm’s relationships will offer similar potential. For example, a manufacturer could share sales and customer data with its retailer who could, in turn, provide more accurate forecasting of the demand for the manufacturer’s products. On the manufacturer’s side, all order processing, inquiry, and service problems could be solved on-line within agreed time frames. Plus the retailer could participate in a product design group on-line due to its knowledge of customers of the manufacturer’s product. However, in other relationships of either partner, the IT integration may be present but not as strategically used or pervasive (O’Toole and Holden, 2001). Therefore, linking technology strategy to overall relationship strategy is vital. Holland and Lockett (1993) attempted to match relationship structures to IOSs in a supply chain management application. The issue of control over the information system dominated many hyopthesised structures. However, the authors did develop a competitive space for a co-operative information system that is analogous to the e-relationship concept of this paper. The core strategic questions when considering an e-relationship strategy are: . How can e-relationships be used to strengthen ties between partners? . Which partners should we use e-relationships with? . How can e-relationships be used to integrate partner firms’ systems? . What effect has electronic integration on traditional and on on-line marketing techniques? . Do e-relationships offer sustainable competitive advantage? . Can firms manage e-relationships or are they optimised through trust? Using relationships strategically, to add value through integrated electronic exchanges with consumers and other businesses, appears to be the preserve of the larger firm. Smaller firms adopt electronic relationships when they have to, for example, if forced by a larger partner, or at a time lag when clear technology standards and off the shelf packages become available that are easily managed. The exceptions to these trends tend to be the case of the very few. The remainder of this paper considers the strategic and implementation barriers in the small firm, which impede its exploitation of

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

the e-relationship concept to create strategic value.

E-relationships and the small firm The definition of a small firm varies from country to country. However, common characteristics include less than 50 employees and independent ownership and management. Most decision-making functions in small firms rest with the ownermanager. Small and medium sized firms are estimated to comprise over 90 percent of all firms and are, therefore, an important focus for the direction of strategic and marketing planning. The physical and knowledge resources available to plan and execute strategy in the small firm are limited (Rothwell and Dodgson, 1992; Nooteboom, 1994). The skills needed to assess strategic opportunities at a macro level are beyond the capacity and competence of most small firms. Strategy in small firms originates at the level of the firm and with issues that reflect its immediate environment. Relationships and network perspectives on strategy present smaller firms with real opportunities, as they are dependent on the network to which they are connected (Bryson et al., 1993; Perry, 1999). Their relationships with customers, suppliers, employees can define their business. Therefore, they would seem well placed to take advantage of electronic relationships to add value to their business and to further copper fasten their ties to their network of partners. Small firms have lagged their larger counterparts in adopting and gaining value from information technology (Yap et al., 1992). However, the Internet and using it as a marketing tool was seen as a mechanism for small firms to use their advantage of flexibility and adaptability to minimise their barrier of scale and geographic reach. This use of the Internet by small firms seems to be the case of just a few firms rather than the many. Characteristics of the smaller firms that have used Internet marketing successfully appear to parallel the rationale of their larger counterparts – they saw a first mover advantage and went for this opportunity. The exploitation of first mover advantage depends on many factors but especially on management commitment and technological sophistication. Many of the successful smaller firms have been a long time developing their product and thus honing their Internet marketing skill. Small firms have adopted the Internet but have not universally managed to extract

competitive value from it. The fact that most small businesses have not used the Internet to transform their business is not surprising considering the pace at which small firms have adopted information technologies. They have been found to be at the early stages in the adoption of information systems in general, see, for example, Bridge and Peel (1999) and Poon and Swatman (1999). Therefore, small firms are often placed in the early stages of accepted models of IT evolution, for example, at the initiation stage of Nolan’s (1979) information systems growth model, or at the very early stage of Venkatraman’s (1994) transformation model that is enjoying a low level of benefits and potential to transform business processes through information technology. This is illustrated in a study by Chaffey (2000) that found low levels of e-commerce integration among small firms that are a lack of full integration of back office to customer processes. Basic marketing tasks such as electronic availability of brochures and ordering online are not backed up with business integration. Larger firms, on the other hand, appear to have the financial and technological resources to achieve integration. For example, Dell computers (www.dell.com) on-line ordering and configuration of personal computer requirements are fully linked to back office and production processes. This allows Dell to earn more from its value chain. Indeed, this approach has been used by some firms to capture further sources of revenue such as Ryanair’s (www.ryanair.com) use of its online booking and ticketing system to lower brokerage fees and ticketing costs and redefine its relationship to its customer markets. The small firms that have had the greatest Internet marketing success are those who owe their existence to the Internet such as ebay (www.ebay.com), the auction portal. The research in this paper suggests some reasons as to the lack of strategic success by small firms in exploiting the potential of electronic relationship marketing. The research for this paper was based on consultancy and case research by the author and is supported by a benchmark study on e-commerce usage in small and medium sized enterprises (Clancy, 2001). The case research incidents reported involved small companies – two precision engineering firms, two professional services firms, and two specialised food companies – and the research and validation procedure followed Yin (1993) and Patton (1990). The benchmark e-commerce study audited the information

[ 117 ]

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

technology and e-business practices of 100 small and medium sized enterprises (regional random sample included manufacturing, primary processing, and service firms) and confirmed the usage findings of the more in-depth case studies.

Electronic relationship strategy Through the series of case studies, three key issues emerged that determined the strategic approach and potential of e-relationships to the small business. They are manager/owner commitment, technological sophistication and relationship orientation. Manager/ owner commitment and technological sophistication are recurring themes in information technology adoption literature. They also apply to electronic relationships. These issues along with the extent of relationship orientation in the business set the boundaries for the strategic potential of electronic relationships for the small business. A relationship orientation is needed to maximise value including joint communication and information sharing in an electronic relationship. Given that general strategy development is more limited in small firms, it is not surprising that many have not realised the potential of e-relationships. In the small enterprise, the owner/ manager characteristically plays a dominant role in the organisation, thereby shaping the organisational culture with respect to the use of IT (Thong and Yap, 1995). Often they are the only ones in the organisation with the authority, responsibility and access to the information needed to evaluate opportunities for strategic or competitive purposes. Owner/managers, unless they come from an IT background, may not see the benefits of e-relationships. The lack of knowledge of e-business models impedes the development of electronic relationship forms in the small business. The majority of firms in the sample had adopted e-mail with customers but few had integrated electronic relationships into their demand or supply chain. Indeed, there appears to be an industry effect, high technology industries and service businesses had more use of relationship technologies, especially Web pages with functionality including ordering. However, the firms that used electronic relationships to their fullest extent were pioneers, found in all industry types, and saw a potential into the future. These enterprises were placing future bets and were not sector specific but had owners/managers who were totally committed to a new business platform.

[ 118 ]

In one case, a company had limited sales through its electronic relationship presence but believed that the industry future was to go that way – they were ahead of the game. In another case, a company had used technology with an overseas partner to add value to its product offering. Instead of just supplying a basic raw material to a pre-specified low technology design they were able to design and make complete units through a close e-working partnership. The case studies in this research confirmed existing studies (see, for example, Poon and Swatman, 1997) that without a certain level of firm technological sophistication it is difficult to integrate information and communication technologies back into the business, and in the case of this research, back into the relationship processes within the firm. For example, there was little back-end automatic order fulfillment. The confinement of firms’ electronic business application to e-mail and Web brochures limits their ability to exploit the linkages that technology can create. There are many issues that compete for the attention of the owner/manager and e-relationships may not be a high priority. Outside the lack of expertise, many small businesses won’t move until the industry does. This follower status, which is often reactive, does not contribute to the build up of technology competencies in the small firm. In the successful e-relationship small businesses tend to have long experience and are first movers in their sector. The technology sophistication needed to integrate is much higher than the purchase of frontend packages. The business case is not there as most firms still find it is possible to process orders manually and just don’t see the financial case for integration even though orders may be taken, manually completed, and then transferred to a computerised financial system by different people. Lack of technology sophistication is a significant barrier to developing integrated e-relationships as there is little integration between the Internet and internal information technology in organisations and between the companies studied, and for this to happen it would appear that a considerable organisation-wide change would be necessary. Small businesses often become focused in order to maintain the specialised needs of a few customers. They are therefore likely to suffer from forced adoption of information technologies; whether the adoption of such technologies will be of long term strategic benefit to the SME depends on the trading relationship with their dominant partner.

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

This was mainly the case in this research and confirms existing work on the adoption and response to e-business being driven by factors external to the small firm (Poon, 2000). For example, Sillince et al. (1998), in a study of 360 UK small and medium sized enterprises, found that e-mail was used primarily in response to external pressures of the customer or supplier. The small firm that develops e-relationships with its suppliers or customers from its own strategic intent could exert significant competitive advantage. Therefore, an additional level of e-relationship strategy is relationship orientation. A firm’s belief and actions in a relationship can influence how an e-relationship develops. In a hierarchical relationship structure, e-relationships will be delineated by strict rules and procedures and the smaller firm will continue to seek means to balance its electronic dependence on its powerful partner. A relationship based on a more inclusive basis can exploit higher order value from e-relationships, for example, through the integration of business processes across organisations or through the direct involvement of customers in the design and configuration of products/services. There were few cases in the study that had realised significant benefits. In the absence of absolute co-operation (a strong relationship orientation) maximising strategic value using technology is limited.

Implementation-level barriers There are many implementation issues that small businesses must address when implementing electronic relationships. These implementation issues parallel general IT implementation issues found in previous studies (for example, Yap et al., 1992; Palvia et al., 1994; Thong, 1999). Firms need to reassess their business processes to exploit e-relationships. However, the most pervasive use of information technology is to automate existing processes and to place this automation on top of existing processes without any evaluation. The strategy is based on reducing costs of processes rather than increasing the value of these processes to the relationship. It is transaction orientated rather than relationship orientated. E-relationships must be linked to the business processes such as production, sales, service, and supply to be effective. E-relationships require people and processes to support them. Often, getting the back office support right requires considerable changes to current ways of working. The information intensity of

e-relationships places huge importance on managing volumes of data to best understand the relationship. Having systems in place to do this is critical, for example, well-managed databases. Again the costs for a small business to get the back office right can be prohibitive, for example, the need for network expertise, data mining capability, etc. This factor makes many small firms wait for standard solutions rather than bespoke investment (for example, off the shelf e-CRM packages). This can minimise flexibility and responsiveness rather than create advantage in the e-relationship. Only relationshipcentred firms are willing to take long-term views of the cost of developing erelationships. When short-term benefits do not appear to accrue, a ‘‘wait and see’’ attitude seems prevalent. This approach may be linked to the lack of strategic planning in general in the small firm and has many parallels in the literature (Bridge and Peel, 1999). There are numerous control issues when other independent organisations/individuals have access to your information. Given the independence mentality of many small business owners, control over information flows was a major issue in implementing e-relationships. Relationships based on high trust may not perceive the risk of information abuse to be that great. However, many firms aim to monitor information exchanged and to put in place safeguards to manage access. Security and access to perceived strategic data of a firm are a major concern. Security of information exchanged between partners and access to each other’s systems remains a barrier to implementing e-relationships. Considering the information that can be exchanged on e-mail across companies without any potential control gives one the idea of the need for security. In addition, consumers are concerned about the type of information disclosed about them and the security of electronic monetary exchanges. There is also much doubt about what electronic relationships can deliver, as there are problems with quality of connectivity. Telecommunication services for high speed and content rich exchanges are not universally available or competitively priced. This can cause partner friction and hinder e-relationship development. Electronic marketing and technology expertise and competence is mainly not found in the small firm. For example, Web site management is ongoing, yet, often in the small business the development and design are completed by external software providers. Therefore, there is no internal

[ 119 ]

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

[ 120 ]

expertise to continue to develop and actively engage in the use and promotion of the site. Small firms in the study were dependent on external expertise or on the internal person assigned to the task for most of their IT expertise. There is little technology expertise diffusion, thus limiting its application. E-relationships as a communication channel are one-to-one which can lead to a splintering of overall marketing messages. This poses huge risks for the marketing function in its contribution to e-relationship management. Fragmentation of message and associated media makes it very difficult to communicate standard advertising messages and retain consistency in brand and product images. Given that smaller firms have limited resources devoted to marketing, adding another communication medium does pose challenges to its integration into the other elements of brand and company communication. Many of the companies studied stressed the importance of salespeople to complement e-relationships. The development of e-relationships has been used to reduce sales costs, especially sales visits and number of salespeople – there was considerable dissatisfaction with this approach in the case companies. Firms still put salespeople first as their most effective communication medium. In fact, e-relationships should free the salespeople to work on the relationship, as the time spent on the order fulfillment process should be much reduced. Customers, in business-tobusiness relationships, believed that any reduction in salesforce contact and communication would have a negative effect on the relationship. The evolution of electronic business relationships is focusing more pressure on small businesses to develop technology and electronic marketing skills and to integrate these into their existing marketing communication. The overall picture from our sample is one of limited integration of e-relationships into the business system and processes of small firms, with few exceptions. The patterns emerging may be expected given similar findings in small business IT literature in general. Competitive opportunity available to the small firm through e-relationships may be limited, as an additional layer of expertise is needed to fully exploit this medium. Far from promising to make the small firm competitive with its larger counterparts, e-relationships may have widened the performance gap. However, a similar barrier faces all organisations – to exploit e-relationships requires a close relationship that may only be available to the few.

Conclusions and managerial implications The emergence of e-relationships has been hastened by technological developments aimed at linking firms to their partners in their demand and supply chains. However, partnerships that are close have most to gain through implementing electronic relationship strategy. E-relationship advantage is sustained in the value created in a co-operative relationship. A joint value maximisation approach creates uniqueness through information and communication rich exchanges not easily copied, through linked business processes, new products and services, and adds an extra element to customers’ relationship to the brand. E-relationship advantage may only be available to the few as a close relationship is the basis for creating such sustained difference. An e-relationship strategy requires management commitment and resources. These may not guarantee success unless a relationship orientation is present in the first place. A relationship orientation is a co-operative, trusting partnership for mutual value creation. For smaller firms, there are many strategic and technology barriers limiting their e-relationship potential. However, some small firms have developed unique customer relationships using e-technologies even though these relationships may not be linked to back office processes and, therefore, not fully exploited for the firm. Nonetheless, they represent asset bases that can be developed further. How can this latent potential be harnessed? A long-term perspective is necessary to justify the costs and benefits from fully integrated e-relationships. Therefore, management awareness and commitment to this platform for competing is a prerequisite to success. Training of owner/managers will be a critical issue with the objective of exploiting e-relationships as a strategic asset. This training impetus may have to originate from, in the case of the small firm, a co-ordinated national policy response. It could be implemented using industry associations, consultants, government agencies, universities or a network of such providers. One idea would be to use a virtual e-business laboratory as a demonstrator of potential. This would enable firms to visualise the longterm benefits of e-relationships to the costs and performance of their supply chains and to their relationships with their customers. The apparent high usage of some e-commerce tools (for example, e-mail and Web pages) by the small firm clouds the reality of a significant technological lag

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

compared to its larger counterpart. This, coupled with a lack of strategic resources in most small firms, makes e-commerce and e-relationships a policy issue. From a noninterventionalist economic perspective, the marketplace would determine the survival or not of firms in a new technology age. A more minimal policy case may be to ensure the infrastructure for these technologies is made more widely and cheaply available and no more. However, as small firms represent the bulk of enterprises and are often the source of much economic innovation and growth, use of our network of policy agents, agencies, and institutions can be advocated. For example, universities can act as providers of discussion forums and training for companies in exchange for data and participation in new technology development. Indeed, some such programmes are already in place. Competing, using an e-business model, is still at a nascent state for the small firm. Therefore, as a corollary, the opportunities suggested in this paper for e-relationships have yet to be fully exploited. The strategic questions presented earlier can form the commencement debate in any company as to whether e-relationships are part of the competitive competencies a firm wishes to use. Nevertheless, the strategic and implementation barriers to e-relationships for the smaller firm are numerous and outlined in detail in this article. However, the key question facing all businesses is whether or not a relationship approach represents a credible strategy for competing. If it does it can be complemented by an electronic relationship platform that adds strength and depth to the core partnership.

References Brady, M., Saren, M. and Tzokas, N. (2000), ‘‘The impact of IT on marketing: an evaluation’’, Management Decision, Vol. 37 No. 10, pp. 758-66. Bridge, J. and Peel, M.J. (1999), ‘‘A study of computer usage and strategic planning in the SME sector’’, International Small Business Journal, Vol. 17 No. 4, pp. 82-7. Bryson, J., Wood, P. and Keeble, D. (1993), ‘‘Business networks, small firm flexibility and regional development in UK business services’’, Entrepreneurship and Regional Development, Vol. 5, pp. 265-77. Chaffey, D. (2000), ‘‘Achieving Internet marketing success’’, The Marketing Review, Vol. 1 No. 1, pp. 35-59. Clancy, S. (2001), Wirecom e-Commerce Reports on SMEs, WIT/SEBIC. Cunningham, C. and Tynan, C. (1993), ‘‘Electronic trading, interorganisational systems and the nature of buyer-seller relationships: the need

for a network perspective’’, International Journal of Information Management, Vol. 13, pp. 3-28. Donaldson, B. and O’ Toole, T. (2002), Strategic Market Relationships – From Strategy to Implementation, John Wiley and Sons, Chichester. Dyer, J.H. and Singh, H. (1998), ‘‘The relational view: cooperative strategy and sources of interorganisational competitive advantage’’, Academy of Management Review, Vol. 23 No. 4, pp. 660-79. Geiger, S. and Martin, S. (1999), ‘‘The Internet as a relational marketing tool – some evidence from Irish companies’’, Irish Marketing Review, Vol. 12 No. 2, pp. 24-36. Holland, C. and Lockett, G. (1993), ‘‘Forms of association in business markets: the impact of inter-organisational information systems’’, Advances in International Marketing, Vol. 5, pp. 125-43. Johnson, J.L. (1999), ‘‘Strategic integration in industrial distribution channels: managing the interfirm relationship as a strategic asset’’, Journal of the Academy of Marketing Sciences, Vol. 27 No. 1, pp. 4-18. Knol, W.H.C. and Stroken, J.H.M. (2001), ‘‘The diffusion and adoption of information technology in small- and medium-sized enterprises through IT scenarios’’, Technology Analysis and Strategic Management, Vol. 13 No. 2, pp. 227-46. Lorenzoni, G. and Lipparini, A. (1999), ‘‘The leveraging of interfirm relationships as a distinctive organisational capability: a longitudinal study’’, Strategic Management Journal, Vol. 20, pp. 317-38. McWilliam, G. (2000), ‘‘Building stronger brands through online communities’’, Sloan Management Review, Vol. 41 No. 3, pp. 43-54. Madhavan, R., Koka, B.R. and Prescott, J.E. (1998), ‘‘Networks in transition: how industry events (re)shape interfirm relationships’’, Strategic Management Journal, Vol. 19 No. 5, pp. 439-59. Nooteboom, B. (1994), ‘‘Innovation and diffusion in small firms: theory and evidence’’, Small Business Economics, Vol. 6 No. 5, pp. 327-48. Nolan, R.L. (1979), ‘‘Managing the crises in data processing’’, Harvard Business Review, Vol. 57 No. 2, pp. 115-27. O’Toole, T. and Holden, M.T. (2001), ‘‘The centrality of communication in relationships and to electronic trading – a Waterford Crystal-retailer case study’’, Irish Marketing Review, Vol. 14 No. 2, pp. 19-25. Palvia, P., Means, D.B. Jr and Jackson, W.M. (1994), ‘‘Determinants of computing in very small business’’, Information and Management, Vol. 27 No. 3, pp. 161-75. Patton, M.Q. (1990), Qualitative Evaluation and Research Methods, 2nd ed., Sage, Newbury Park, CA. Peppard, J. (2000), ‘‘Customer relationship management (CRM) in financial services’’,

[ 121 ]

Thomas O’Toole E-relationships – emergence and the small firm Marketing Intelligence & Planning 21/2 [2003] 115-122

[ 122 ]

European Management Journal, Vol. 18 No. 3, pp. 312-27. Perry, M. (1999), Small Firms and Network Economies, Routledge, London. Poon, S. (2000), ‘‘Business environment and Internet commerce benefit – a small business perspective’’, European Journal of Information Systems, Vol. 9 No. 2, pp. 72-81. Poon, S. and Swatman, P.M.C. (1997), ‘‘Small business use of the Internet’’, International Marketing Review, Vol. 14 No. 5, pp. 385-402. Poon, S. and Swatman, P.M.C. (1999), ‘‘An exploratory study of small business Internet commerce issues’’, Information and Management, Vol. 35 No. 1, pp. 9-18. Ramirez, R. (1999), ‘‘Value co-production: intellectual orignis and implications for practice and research’’, Strategic Management Journal, Vol. 20, pp. 49-65. Rothwell, R. and Dodgson, M. (1992), ‘‘European technology policy evolution: convergence towards SMEs and regional technology transfer’’, Tecnovation, Vol. 12 No. 4, pp. 223-39.

Sillince, J.A.A., Macdonald, S., Lefang, B. and Frost, B. (1998), ‘‘E-mail adoption, diffusion, use and impact within small firms: a survey of UK companies’’, International Journal of Information Management, Vol. 18 No. 4, pp. 231-42. Thong, J.Y.L. (1999), ‘‘An integrated model of information systems adoption in small business’’, Journal of Management Information Systems, Vol. 15 No. 4, pp. 187-214. Thong, J.Y.L. and Yap, C.S. (1995), ‘‘CEO characteristics, organisational characteristics and information technology adoption in small businesses’’, Omega, Vol. 23 No. 4, pp. 429-43. Venkatraman, N. (1994), ‘‘IT-enabled business transformation: from automation to business scope redefinition’’, Sloan Management Review, Vol. 35, Winter, pp. 73-87. Yap, C.S., Soh, C.P.P. and Raman, K.S. (1992), ‘‘Information systems success factors in small business’’, Omega, Vol. 20 Nos. 5-6, pp. 597-603. Yin, R.K. (1993), Applications of Case Study Research, Sage, Newbury Park, CA.

E-trust: the influence of perceived interactivity on e-retailing users

Bill Merrilees Griffith University, Gold Coast, Queensland, Australia Marie-Louise Fry University of Newcastle, Newcastle, New South Wales, Australia

Keywords Electronic commerce, Interaction, Internet, Retailing, Consumer behaviour, World Wide Web

Abstract E-retailing has received considerable attention in recent years, especially with the help of consumer behaviour concepts and tools. This particular paper focuses on e-trust and its antecedents. Special attention is given to the proposition that the more a site is perceived to be interactive, then the greater the trust that users attach to that site. Previous research has demonstrated this relationship with respect to a particular site. The current paper re-examines this relationship with respect to another site. More importantly, the paper tests whether the nexus between interactivity and e-trust applies to both pure ‘‘click’’ and ‘‘brick and click’’ e-retailers.

Review of literature examining consumer interactions with Internet purchasing reveals an evolutionary approach. An initial wave of e-commerce research has focused on the adoption of the Internet from the perspective of either the consumer (Eastlick

and Lotz, 1999) or the firm (O’Keefe et al., 1998). Typically, these studies demonstrate that perceived risk of using the Internet has been a major deterrent to the take up of the Internet. The problem areas include concerns about credit card security, the risk of not being able to return the product if unsatisfied and the risk of having to evaluate a product without handing or seeing it. We can reexpress these problems as one of trust. Presumably, if trust can be assured, then a greater proportion of consumers would be willing to use the Internet for online shopping. Another wave of literature has addressed a greater understanding of the consumer behaviour aspects of the Internet. How do consumers search the Web for information? (Rowley, 1996, 2000). What criteria do they use for selecting goods on the Internet (see especially Morganosky and Cude, 2000). What is their experience on the Web and do they experience a ‘‘flow state’’ (i.e. a state of immersion in which the challenge of the task dominates the sense of presence of the participant (Hoffman and Novak, 1996; Novak et al., 2000). How do consumers form brand attitudes about e-retailers? (Merrilees and Fry, 2002) What is the level of interactivity? (Wu 1999; Merrilees 2001) How is trust developed on the Internet? (Urban et al., 2000). The level of trust seems to be of particular importance given the perceived risks of using e-retail sites. The above-mentioned body of research has highlighted two key issues relevant to consumers’ uptake of e-shopping. The first issue refers to the ability of firms designing Internet sites that are easy to use, easy to navigate and allow the user to interact with the site to make decisions about ‘‘what will be presented and how it will be presented’’ (Andrisani et al., 2001). The second key issue refers to the importance of firms developing

The Emerald Research Register for this journal is available at http://www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at http://www.emeraldinsight.com/0263-4503.htm

Introduction The recent literature on the Internet has begun to investigate how consumers behave in this new medium. Particular emphasis has been given to factors that impede their use of the Internet, especially shopping online. Other studies have examined how consumers search for information online and form brand attitudes towards online retailers. A small number of studies have attempted to fit the various pieces of the Internet experience together, but much work remains to be done. The current paper highlights two key concepts that relate to Web usage, namely interactivity and e-trust. It is suggested that these may be among the most important variables that explain consumer behaviour on the Web. Surprisingly, there are only a limited number of previous papers that have empirically analysed these two variables. In part, the paper contributes to the measurement of these two variables. More importantly, the paper proposes that the level of perceived interactivity is a major determinant of e-trust. That is to say, the paper proposes that e-retailers wishing to augment the amount of perceived e-trust that users and consumers have of them should ensure their site is interactive. We begin by reviewing the literature, presenting a research design, outline the sampling frame, present the results, discuss the managerial implications of the results and conclude.

Literature review

Marketing Intelligence & Planning 21/2 [2003] 123-128 # MCB UP Limited [ISSN 0263-4503] [DOI 10.1108/02634500310465461]

[ 123 ]

Bill Merrilees and Marie-Louise Fry E-trust: the influence of perceived interactivity on e-retailing users Marketing Intelligence & Planning 21/2 [2003] 123-128

[ 124 ]

trust-based strategies to build positive relationships with their customers while increasing market share and profits (Hoffman et al., 1988). We will elaborate on each of these as they form the central focus of the paper. Interactivity refers to the interaction between the site and a user of that site and goes to the core of a computer-mediated communication environment (Hoffman and Novack, 1996). Despite its central place in the Internet world, interactivity seems to be under-researched. Often the term is poorly defined or too narrowly defined. Regarding the latter, Hoffman and Novak (1996) use a simple three-item measure of interactivity, which is narrowly expressed as the speed of interaction. Elsewhere, Hoffman and Novak (1996) implicitly refer to the ‘‘flow-state’’ as another type of interaction, but again this is a very narrow and extreme sub-set of interactivity. A number of other papers have suggested that interactivity has the potential to be an important influence in building online relationships (Cuneo, 1995; Berthon and Watson, 1996; Ha and James, 1998; Lindstrom and Anderson 1999), though they were conceptual rather than empirical. Most of the few papers that have actually measured interactivity have largely been exploratory (Wu, 1999). This literature has been summarised by Merrilees (2001) who notes the following features of interactivity: . The most common impression of interactivity from the literature is that it is primarily communication-based. Infrequently, other activities, such as downloading software or making a purchase, are noted as forms of interactivity that seem to have something else other than communication as the driving force. In such cases, the communication or information aspects of those activities can be highlighted. . Interactivity is about two-way communication. This spans communication from the viewer to the firm and from the firm to the viewer, as well as simultaneous interaction between the two parties. . A special feature of Internet interactivity is the ability to personalize the situation for an individual. . A number of writers have emphasized the ability of the individual to control the communication and learn as a noteworthy feature of the interactive process. . A broader, more holistic role of interactivity has been noted as an important contributor to building up the total shopping experience on the Internet.

Merrilees (2001) has incorporated the above features of interactivity into a multi-item scale for interactivity. This scale was shown to be valid and is therefore appropriate to use in the current paper. Notwithstanding, validity checks will be made to ensure that the scale is valid for the samples in this paper. E-trust captures specific cues of Internet sites important to consumers in visiting retail Web sites and for organisations to develop brand equity. Recent industry reports and academic literature highlight the emergence of providing consumers with e-trust building components as a key driver to e-retail, especially in the light of transaction based financial information continuing to be a major inhibiting factor for online consumers’ success (Forrester, 2000b; Nua, 2000; Urban et al., 2000). Forrester (2000a) found that many consumers consider the lack of security as the main inhibiting factor to adopting on-line purchase behaviour, and thus perceive e-purchasing to be risky. Importantly, the perception of risk is a pivotal aspect of consumer behaviour because of the perceived pain and anxiety associated with negative outcomes is real to the individual (Salam et al., 1998). However, Fry and Merrilees (2001) found that those consumers with high levels of perceived security risk were more likely to reduce their level of risk perception via in-depth experience with the Internet site. Importantly, innovative e-retailers are recognising the importance of developing trust-based strategies to build positive relationships with their customers while increasing market share and profits (Hoffman et al., 1998; Urban et al., 2000). Urban et al. (2000) identify the drivers to building Web site trust as: 1 building trust in the specific Web site, ensuring customers of the security of transactions, customer privacy of data collected; 2 providing unbiased and complete information concerning availability of stock, competitive prices, delivery time and reliability; and 3 ensuring fulfillment and satisfaction guarantees of stock, service support, error free billing and credits on return items. A measure of e-trust has been taken from Merrilees and Fry (2002) and includes items relating to level of credit card security, consumer privacy details, safety and security of the e-site, trust of the e-site and credibility of communications relating to the e-site.

Bill Merrilees and Marie-Louise Fry E-trust: the influence of perceived interactivity on e-retailing users Marketing Intelligence & Planning 21/2 [2003] 123-128

Research design The authors have developed a questionnaire based on the literature, deriving issues important to Internet users/shoppers relevant to their attitudes towards purchasing from e-malls. Australian respondents were asked to familiarize themselves with three pre-determined Internet CD sites, namely CD Now, Sanity and Amazon, and explore the sites’ features before answering a set of likert scale based questions. In constructing the survey items specific to this study, we reviewed prior studies to extrapolate items examining Web usage research. A survey was designed with 12 items intended to capture the underlying multi-item constructs of e-interactivity and e-trust. Each survey item was measured on a seven-point scale of (1) disagree very strongly to (7) agree very strongly. The study consisted of 202 Australian undergraduate and post-graduate marketing students from a large regional area in New South Wales. Given the nature of the study, only those who had indicated that they had used the Internet were selected to participate in the study. The student sample met the minimum conditions needed for this study in that they were: . familiar and experienced with the Internet for information search; and . of an age (mostly 18-25) that comprises a major segment in the compact disc buyer market (Forrester, 2000b). Importantly, as the study focuses on purchase intention rather than actual purchase, it was not critical that respondents had previously purchased via the Internet. All respondents had access to the Internet, either privately or through university facilities. The sample consisted of an (almost) equal number of males (91) and females (111) with a median age of 23 years.

Results A number of tests were undertaken to ensure that the scales of e-trust and interactivity were meaningful and valid. The tests were reapplied to the CD Now site as a benchmark, as well as the two additional sites, namely Sanity and Amazon (Table I). First, Cronbach Alphas were calculated. As expected, the five-item e-trust measure for CD Now had a reasonably high score of 0.75, indicating reliability. Similarly, the e-trust Cronbach Alpha for the other two sites, Sanity and Amazon, were also above the 0.70 research standard, with scores of 0.79 and 0.77 respectively. The seven-item

interactivity measure for CD Now had a high score of 0.83, indicating reliability. Similarly, the interactivity Cronbach Alpha for the other two sites, Sanity and Amazon, were also well above the 0.70 research standard, with scores of 0.85 and 0.82 respectively. We conclude that the constructs of e-trust and interactivity used in this study are reliable. Second, separate confirmatory factor tests were undertaken to verify validity of the two constructs, namely e-trust and interactivity (see Table I). For all three e-retail sites, the e-trust construct and interactivity construct were valid, as indicated by an insignificant chi-square statistic. Further, the two constructs confirmed high levels of fit between the model and the data; for example the adjusted goodness of fit were high, in excess of 0.90. Moreover there were low levels of misfit between the model and the data for both constructs; for example, the ratio of C-min to degrees of freedom were appropriately well under the acceptable standard of 3.0. Third, both the e-trust construct and the interactivity construct indicated face validity in relation to the manner that the items relate to the way the broad concept of e-trust and interactivity are discussed in the literature. The literature refers to the broad essence of e-trust and the ideas of credibility, integrity and honesty. Moreover, the literature often discusses the specific issues of privacy of information provided by the user to the e-retailer as well as the endemic problem of credit card security concerns. Literature pertaining to online interactivity refers to the notions of communication, two-way communication, responsiveness of the site to user queries, control and personalisation. Based on these three sets of evidence, we conclude that the two constructs – e-trust and interactivity – presented in this paper are reliable and valid. In other words, we can have confidence in the across-site comparisons of e-trust. Additionally, we can have confidence in using the e-trust variable in further analyses later in this paper, when we run a multiple regression analysis with e-trust as a function of interactivity.

The statistical nexus between interactivity and e-trust A previous study undertaken by the authors (Merrilees and Fry, 2002) examined the relationship between e-trust and interactivity. In their study, the authors identified that while a combination of interactivity and e-trust are critical in the formation of favourable attitudes towards a site, interactivity was clearly seen to be more

[ 125 ]

Bill Merrilees and Marie-Louise Fry E-trust: the influence of perceived interactivity on e-retailing users

Table I Confirmatory factor tests for the constructs of e-trust and e-interactivity in three Australian online CD sites

Marketing Intelligence & Planning 21/2 [2003] 123-128

Construct/site Trust (CD Now) Trust (Sanity) Trust (Amazon) Interactive (CD Now) Interactive (Sanity) Interactive (Amazon)

Chi-square p-value

GFI

AGFI

C-min/df

Rmsea

0.604 0.689 0.627 0.445 0.118 0.159

0.99 0.99 0.99 0.98 0.98 0.98

0.98 0.98 0.98 0.96 0.94 0.95

0.68 0.56 0.58 1.00 1.57 1.41

0.00 0.00 0.00 0.002 0.053 0.045

important partly because of a greater direct contribution and partly because of its additional indirect contribution via e-trust. While there are no strong theories of what causes greater interactivity with an e-retail site, we suggest the analogy with the sensory/stimulus/motivation literature (Korgaonkor and Wolin, 1999; Alpert, 1994). Sites that are more enjoyable and entertaining are more likely to engage the user in two-way communication and dialogue with the retailer. To explore the notion of an interaction between interactivity and e-trust among three sites that incorporate a mixture of brick and mortar and brick and click stores, the next stage of the analysis was to run a simple regression equation in which the interactivity variable was postulated as a determinant of the e-trust variable. The latter variable was treated as the dependent variable. Separate regressions were run for each of the three e-retail sites, CD Now, Sanity and Amazon. The regression results are given in Table II. There was a positive and significant relationship between interactivity and e-trust in all three e-retailer sites. In fact, the standardised beta coefficients were very high at 0.66, 0.70 and 0.68 respectively. The t-values were very high, in excess of 12.0, indicating that the results were very highly significant (p = 0.000), formally significant at the 1 percent level. The positive beta coefficient indicates that the higher a rating in terms of how

Table II Regression analysis, determinant of e-trust, three Australian online CD e-retailers Dependent variable: e-trust

CD Now

Sanity

Amazon

Beta coefficient on interactivity independent variable t-value Constant term t-value Adjusted R2 p-value

0.70 13.73 1.59 6.50 0.48 0.000

0.66 12.34 1.62 6.74 0.43 0.000

0.68 13.09 1.62 5.88 0.46 0.000

[ 126 ]

interactive the site is, the greater the user perceives e-trust towards the site.

Discussion The first set of results, concerning the development of reliable and valid measures of e-trust and e-interactivity, is important for the development of marketing knowledge in understanding what is important to the e-retailing experience from the perspective of the user. The two particular concepts studied in this paper are not just any two Internet concepts but are fundamental to users having a positive and fruitful online shopping experience. The lack of e-trust is likely to deter any purchase over the Internet, while the lack of interactivity is likely to deter a positive brand attitude and repeat purchases. These results are also of relevance to the marketing practitioner. Assuming that e-retailers are trying to build interesting and effective sites, then the design of those sites should factor in ways to build both e-trust and interactivity. Practitioners can also use the measures of e-trust and interactivity alluded to here as ways of monitoring their sites. For example, our measures of e-trust indicate that, of the three sites examined in this paper, Sanity has the lowest level of e-trust, CD Now has the second highest, while Amazon has the highest level of e-trust in our particular sample. This pattern may vary depending on the sample, but the point is that the scale provides a practical measure that firms can use to monitor how well they are delivering on e-trust or e-interactivity, among other variables. The regression results are also of interest in that they contribute to marketing knowledge. This is done by building up a submodel within the overall user experience of the Internet. Our particular sub-model links two key Web variables, interactivity and e-trust, in a causal way. We now know that e-trust can be developed by increasing the interactivity of the site. This relationship is very robust in that it is present in all three CD sites examined in this paper. The

Bill Merrilees and Marie-Louise Fry E-trust: the influence of perceived interactivity on e-retailing users Marketing Intelligence & Planning 21/2 [2003] 123-128

magnitude of the regression coefficient is very high, of the order 0.65-0.70, indicating that interactivity is a powerful influence on e-trust. Note that the constant term is still positive and significant, suggesting that there are other factors that also influence e-trust. These additional factors need to be identified and incorporated into the modeling of e-trust. However, in the interim it is clear that all future studies of e-trust need to incorporate interactivity as a starting point. A special aspect of our results is a comparison of the nexus between interactivity and e-trust across both ‘‘click’’ and ‘‘brick and click’’ sites. The original study of the nexus reported in Merrilees and Fry (2002) referred to one site only, namely CD Now. At that point we might have been left with the impression that the nexus between interactivity and e-trust was confined to pure ‘‘click’’ e-retailers. Indeed, one might have prophesised that brick-based e-retailers (that is, brick and click) had less need for interactivity because they already had trust built up from their brick operations. In other words, it is only the click e-retailers that have to work hard at interactivity in order to build up e-trust. Given these expectations, our results are somewhat of a shock. We had expected that the constant term would be larger for the brick and click e-retailer (Sanity is a leading ‘‘brick’’ retailer of CDs in Australia) and that the beta coefficient would be a lot smaller for the same site. In fact, the constant terms and the beta coefficients were very similar for both the one brick and click e-retailer (Sanity) and the two click e-retailers (CD Now and Amazon). How can we interpret these phenomena? The similarity of the results across the two business models of eretailing seems to suggest that the trust built up from brick retailing does not automatically carry over to the e-retailing operations of the same firm. In other words, consumers may trust Sanity in a brick context, but still need to be re-assured (via interactivity) that its online activities are trustworthy. This, of course, is a generalisation that may not apply to all segments of users. There may be some segments where there is carryover from brick experiences to trust in the click context of the same firm. Further research is required to examine the size of this segment and to find out what is special about such a group. All of these lessons are equally important for the practitioner as well as the marketing academic. In practice, e-retailers need to know how to build e-trust. An important route, indicated in this paper, is e-

interactivity. Given that some practically oriented academics have suggested that e-retailers need to build trust as the centre of their Internet strategy (see Urban et al., 2000), then developing e-interactivity has been shown to be an important step in pursing such a strategy. In Australia at least there has been a big swing in favour of brick and click firm, with pure click e-retailers (like d-store and e-store) struggling. With an increasing role for brick and click stores, they should keep in mind that they can not simply rely on loyal customers from their brick operations switching, without questions to the click context of the same firm. Rather, brick and click firms should build on their brick knowledge of customers and enhance this in the click context. Brick and click firms still need adequate e-interactivity, otherwise their trusting customers from the offline situation may not trust them in the online context. Trust is most likely a key factor in any company’s reputation, whether e-commerce or not (Merrilees and Fry, 2002). This seems likely because trust is akin to being reassured that the specific company will look after the interests of the user or consumer and protect them in all respects. It is a form of meta-guarantee, one that goes beyond just a product guarantee. Given the virtual nature of the on-line shopping experience, trust is likely to be even more important in the context of e-retailers. This is because major consumer concerns over the adequacy of credit card security or the way that privacy matters are handled could be detrimental to a company’s reputation. To overcome the ‘‘e-trust’’ barrier firms are developing e-trust strategies such as clearly articulating explicit policies regarding credit card security and the way privacy issues are handled to consumers.

Conclusion The current paper has contributed to the measurement of e-trust and e-interactivity, mainly by demonstrating that certain scales are now quite robust against an increasing number of studies of e-retail sites. Reliable and valid measures of e-trust and einteractivity were found in relation to three e-retail sites, namely CD Now, Sanity and Amazon for a convenience sample of Australian users. Equally importantly, the paper has demonstrated that there is a nexus between interactivity and e-trust. Higher levels of interactivity lead to a positive and large increase in e-trust. If the perception of

[ 127 ]

Bill Merrilees and Marie-Louise Fry E-trust: the influence of perceived interactivity on e-retailing users Marketing Intelligence & Planning 21/2 [2003] 123-128

interactivity increases by three points (on a seven-point scale), then the perception of etrust increases by two points (also on a sevenpoint scale and using an average betacoefficient of 0.67, found in Table II). This is a very significant, important and strategic relationship. Of interest, the nexus between interactivity and e-trust was found to be similar across both brick and click (Sanity) and pure click (CD Now and Amazon) e-retailers. Retailers that have built up trust through their brick operations cannot assume that these feelings automatically carryover to the online context. Brick and click e-retailers also have to work hard at building up online interactivity. More research with other sample, in other sites and other countries is needed to back these conclusions. Placing trust at the centre of your Internet strategy is the catch-cry heralded by Urban et al. (2000) in their recent Sloan Management Review paper. Such a view has much merit. However, for this approach to be really effective, firms need to place interactivity at the centre of their e-trust strategy.

References Alpert, F. (1994), ‘‘Innovator buying behaviour over time: the innovator buying cycle and the cumulative effects of innovations’’, Journal of Product and Brand Management, Vol. 3 No. 2, pp. 50-62. Andrisani, D., Gaal, A.V., Gillette, D. and Steward, S. (2001), ‘‘Making the most of interactivity online’’, Technical Communication, Vol. 48 No. 3, pp. 309-23. Berthon, P. and Watson, R. (1996), ‘‘The World Wide Web as an advertising medium: toward an understanding of conversion efficiency’’, Journal of Advertising Research, Vol. 36 No. 1, pp. 43-54. Cuneo, A. (1995), ‘‘Internet world show spurs online commerce debate’’, Advertising Age, 17 April. de Chernatony, L. (2001), ‘‘Succeeding with brands on the Internet’’, Journal of Brand Management, Vol. 8 No. 3, pp. 186-95. Eastlick, M. and Lotz, S. (1999), ‘‘Profiling potential adopters and non-adopters of an interactive electronic shopping medium’’, International Journal of Retail & Distribution Management, Vol. 27 No. 6, pp. 209-23. Forrester Research (2000a), ‘‘Young consumers ignore offline brands’’, available at: www.forrester.com Forrester Research (2000b), ‘‘The online retail and media profile dons a new fac¸ade’’, available at: www.forrester.com Fry, M.L. and Merrilees, B. (2001), ‘‘E-shopping: the role of experience on security perceptions’’, International Quarterly Journal of Marketing, Vol. 1 No. 2-4, pp. 225-37. Ha, L. and James, E. (1998), ‘‘Interactivity reexamined: a baseline analysis of early business Web sites’’, Journal of Broadcasting & Electronic Media, Vol. 42 No. 4, pp. 457-70.

[ 128 ]

Hoffman, D. and Novak, T. (1996), ‘‘Marketing in hypermedia computer-mediated environments: conceptual foundations’’, Journal of Marketing, Vol. 60, pp. 50-68. Hoffman, D., Novack, T. and Peralta, M. (1998), ‘‘Building consumer trust in onlineenvironments: the case for information privacy’’ (Project 2000), Owen Graduate School of Management, Vanderbuilt University, Nashville, TN, available at: www2000.ogsm.vanderbilt.edu Korgaonkor, P.K. and Wolin, L.D. (1999), ‘‘A multivariate analysis of Web usage’’, Journal of Advertising Research, Vol. 39 No. 2, pp. 53-69. Limayem, M., Khalifa, K. and Frini, A. (2000), ‘‘What makes consumers buy from the Internet? A longitudinal study of online shopping’’, IEEE Transactions on Systems, Man and Cybernetics, Vol. 30 No. 4, pp. 421-32. Lindstrom, M. and Anderson, T. (1999), Brand Building on the Internet, Hardie Grant Books, South Yarra. Merrilees, B. (2001), ‘‘Interactivity design as the key to managing customer relations in ecommerce’’, Journal of Relationship Marketing, Vol. 1 No. 3-4. Merrilees, B. and Fry, M.L. (2002), ‘‘Corporate branding: the case of e-retailing’’, Corporate Reputation Review, Vol. 5 No. 2. Morganosky, M. and Cude, B. (2000), ‘‘Consumer response to online grocery shopping’’, International Journal of Retail & Distribution Management, Vol. 28 No. 1, pp. 17-26. Novak, T., Hoffman, D. and Yiu-Fai Yung (2000), ‘‘Measuring the customer experience in online environments: a structural modeling approach’’, Marketing Science, Vol. 9 No. 1, pp. 22-42. Nua (2000), ‘‘Seniors’’, available at: www.nua O’Keefe, R., O’Conner, G. and Kung, H. (1998), ‘‘Early adopters of the Web as a retail medium: small company winners and losers’’, European Journal of Marketing, Vol. 32 Nos 7 and 8, pp. 629-43. Salam, A.F., Rao, H.R. and Pegels, C.C. (1998), ‘‘An investigation of consumer perceived risk on electronic commerce transactions: the role of institutional trust and economic incentive in a social exchange framework’’, Proceedings of Conference for the Association of Information Systems (AIS). Rowley, J. (1996), ‘‘Retailing and shopping on the Internet’’, Internet Research: Electronic Networking Applications and Policy, Vol. 6, pp. 81-91. Rowley, J. (2000), ‘‘Product search in e-shopping: a review and research propositions’’, Journal of Consumer Marketing, Vol. 17 No. 1, pp. 20-35. Urban, G.L., Sultan, F. and Qualls, W.J. (2000), ‘‘Placing trust at the center of your Internet strategy’’, Sloan Management Review, Fall, pp. 39-48. Wu, G. (1999), ‘‘Perceived interactivity and attitude toward Web site’’, paper presented at the 1999 Annual Conference of American Academy of Advertising, Albuquerque, NM.