SYSTEM DESIGN AND CONSUMER BEHAVIOR IN ELECTRONIC COMMERCE

Trang 1

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master UMI films the text directly from the original or copy submitted Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer

The quality of this reproduction is dependent upon the quality of the copy submitted Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted Also, if unauthorized copyright material had to be removed, a note will indicate the deletion

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corer and continuing from left to right in equal sections with small overlaps

Photographs included in the original manuscript have been reproduced xerographically in this copy Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge Contact UMI directly to order

Bell & Howell Information and Learning

300 North Zeeb Road, Ann Arbor, MI 48106-1346 USA 800-521-0600

®

Trang 3

SYSTEM DESIGN AND CONSUMER BEHAVIOR IN ELECTRONIC COMMERCE

MARIOS KOUFARIS Information Systems Department Leonard N Stern School of Business

New York University June, 2000

A thesis presented to the Faculty of the Leonard N Stern School of Business, New York University, in partial fulfillment of the requirements for the degree of Doctor of

Trang 4

UMI Number: 9988944

Copyright 2001 by Koufaris, Marios All rights reserved

®

UMI

OMI Microform9988944

Copyright 2001 by Beil & Howell Information and Learning Company All rights reserved This microform edition is protected against

unauthorized copying under Title 17, United States Code

Bell & Howell Information and Learning Company 300 North Zeeb Road

P.O Box 1346

Trang 5

America, except in the case cf passages quoted from

tublisheac scurces; nat I am the scle auther anc

pretrieter ct saic Gissertation; that the Ccissertatic: cortains no matter which, if published, wil

cr ctherwise in

` SU5SC

Ta

reason of the publicaticn

~ibed to before me this

2Ä _Ô ®‹2 cy) 7

Trang 6

Abstract

The impact of the Internet and the World Wide Web on commerce has been immense Business-to-consumer e-commerce, though relatively smaller, is still a large and fast growing segment of the economy with estimates of over 40 million households in the US alone buying online by the year 2003 (Rowen, 1999) The changes that electronic

commerce has brought have been sweeping We have new goods and services and new ways of selling them such as reverse auctions (Priceline.com) and international “yard sales” (Ebay and Half.com) Geographical and temporal boundanes have enabled consumers to buy anything from anywhere at anytime

All these changes have happened at rapid speeds and new and old companies have struggled to stay afloat in this new volatile commercial space One of the biggest challenges that e-stores have faced and continue to deal with is understanding their customers While a lot is known about how consumers think and act in the brick-and- mortar world, there has not been enough time and effort put into understanding the electronic consumer

In this thesis, we have combined the three reference disciplines of Information Systems Marketing, and Psychology to help us understand online consumer behavior We first developed a theoretical framework based on traditional consumer behavior research and environmental psychology We also compared it to the Technology

Trang 7

Acknowledgements

I would like to thank my dissertation committee members for their help and support Specifically, I’m grateful to Dr Ajit Kambil, my original committee chair and advisor, for his constant encouragement, invaluable insights, and unconditional support with monetary, material, and other resources, even after his departure from NYU I also would like to extend my heartfelt thanks to Dr Edward Stohr, my final committee chair and advisor, for his constant positive outlook, immensely helpful comments, and his tireless assistance I am indebted to Dr Priscilla Ann LaBarbera for her unlimited enthusiasm, reassurance, and insightful comments as well as for providing me with all the necessary Marketing education throughout the entire process Finally, I wish to thank Dr Jon Turner for his tremendous help and support and for making sure I always produced the best research possible

I am also indebted to the founding members of Kozmo.com and especially its CEO Joseph Park Even though they were a young struggling company at the time, he allowed me to use their customers for my first empirical study and took on the task of hosting my survey on the company’s servers The company has since grown into a multi-million dollar e-commerce giant I hope the results of my research study with Kozmo.com provided them with some help in achieving that goal

I would also like to thank everyone at Dynamic Logic, Inc and specifically Nick Nyhan, the company’s founder and CEO His enthusiasm and faith in my research encouraged me to continue and without the company’s help, I would not have been able to complete the second empirical study and possibly this dissertation

Trang 8

Finally, I would like to thank the people that supported me during my five years in the doctoral program both in and out of school The people of the Information Systems Department, faculty, Ph.D students, and staff, provided a warm, friendly, and

intellectually stimulating environment I especially want to thank Dr Hank Lucas, Dr Tomas Isakowitz Arnold Kamis, Wonseok Oh, and Maytal Tsechansky My family, Andreas and Elli Koufaris and my sister Christina Koufaris Makrides, gave me my

Trang 9

| ed 3 3 3 4 4 4 4 4 5 Table of Contents

INTRODUCTION AND MOTIVATION 7

SUMMARY OF THESIS 12

CONSUMER BEHAVIOR ON-LINE: LITERATURE REVIEW 15

.[ THEORETICAL STUDIES: - S5 ch HH ng ng HH ưng 1n tà nh nh n9 1 15 2 EMPIRICAL STUDIES s2 Ăn HH HH nh ng 4 nh hen net ng e4 20 3 SUMMARY OF PAST FINDINGS .ccccccccececoceecceeescseseeessecenscecenseensesensccnnucessnssensrcensreserresesscstsrsanasen 27 THEORETICAL FRAMEWORK - HYPOTHESES DEVELOPMENT .-. < <<se« 29 LL INITIAL ASSUMPTIONS .2 -cccccceseeeesceeseeeescreneeeesetescecscsnersnsaesensseeegsrscecenonectecaesecerenscneessnereeneneaes 29 2 REFERENCE DISCIPLINE 1: MARKETING .::ccsesccesscescesseeessenesecenecencneensneeesecsseaaerersseseeessnesenneese 31 3 REFERENCE DISCIPLINE 2: PSYCHOLOGY - 5-2 ĂĂ- 5S Sen HH nh nh nh HH T00 08014 35 .4 REFERENCE DISCIPLINE 3: [NFORMATION SYSTEMS «son nhe nen 36 .5 CONSUMER BEHAVIOR IN ELECTRONIC COMMERCE - A THEORETICAL FRAMEWORK 38 4.5.1 Dependent Variables .:ccseccseccescceseecenennsenseneeenseneecasenseseersccocnesestenssensseeseeeeraeenneesses 39

4.5.1.1 Unplanned Purchases: 2 : ::cccssessceseneenesssesesseceseecesesseeseescecesnersecssnseessesensesssssenseseraanasaeaananeanars 40

4.5.1.2 Customer Intention to Retum 42

4.5.1.3 Exploratory Behavior and Length of ViSIL .-. -<c+-seeererrerrerrrrerrierriererrrrrrrrrrrrrrrirrrrerreee 44

9.5.2 Mediating Variables .ccccccsseeccceenscennesessscenesensesnneenccesecnerentenesseneeeseressenaessnseeasaenaenanes 45

4.5.2.1 Flow: An Optimal Experience - Literature R€VICW oi eceeeerrrriirerrrrrrrrrirrrerrerrrrrreereererren 48 4.5.2.2 Consumer COntOI: -Ă+ s2 22c HH HH HH HH HH HH TT 72D cm Teen ke HT 0n 4.5.2.3 Shopping Enjoymen( - sec se ren 2 ng ren Thìn ng hề 14774 rnrremtrree 4.5.2.4 Concentratior/Attention Focus

4.5.3 Hypotheses - Impact of Attitudinal Variables on Consumer BtqvIOr -.-«e-«ceeeree 59 4.5.4 Independent Variables - Individual FqCIOFS . << <<-e=eeeererrrrrrrrrrrrrrrrrerrerrrrrree 6i 4.5.4.1 Consumer Need SpeCIẨiCILY c-+scs~creecererh re tr tren tr nh Treo gr cr ren rhnHrgnhrreh 62

4.5.4.2 Customer Tenure (New vs ÑR€pDCAL) .- Q -Ă -< hen HH nh nh ch TH ng 3n chen 63

4.5.4.3 xo ca rốn ố ẽ ẻẽ 64

A544 Demographics . -ĂĂs> set cư nrn nền ch tềnr nhang ng 3g net TRE tren 65 $.5.4.5 Personality Typc . +~~~seren nen hư ren nh HH TH nnennrrrre 65

4.5.4.6 Web Skills

4.5.5 Independent Variables - Environmental Factors:

4.5.5.1 Product Search Mechanisms 4.5.5.2 Challenges

4.5.6 Hypotheses - Impact of Individual and Environmental Factors on Consumer Attitudes 71

FIRST STUDY: KOZMO.COM 80

vn an cm S 80 5.1.1 [nHHpÍ€HI€HIQIÍOH .àc<<<eseerrrerrerrerrrrrrerreerrserererrreerrerrrerrrrrrrrerrrrerrrrTrinneeeereknree 8/ 5.1.2 Survey Instrument and Data Varidbles . -«-~-s==eseeeererrreerrrrrrirrsreeesreereeree 82 5.1.2.1 Unplanned Purchases: 22 ccscccccsceeceeceeensecesesesneesssenessecensnescaseeesecsestesseessanessesensenessescassnenstessaegs 82 5.1.2.2 Need Specificity: eee seccceeeeeeeseessessessesnesesssecsnenceseereneneesecsesorseenssnenseeseessssnsnensanenenaeunaegenenaerses 82 5.1.2.3 Length of ViSIL àà-c ereeeee~rrrrreeeee 83

5.1.2.4 Number of past VỈSILS: -SScS+eherreHtnrrrrg teen HH Trưng 1001n 0 nenereerrene 83

5.1.2.5 Search mechanisms used and search mechanism pref€rr€d: -. <s+sserereerrrrrrrrrrer 83 §.1.2.6 Concentration/Attention FOCUS: - -<-S+Ă can TH nen ng 01 2.0711 me nnnng 84

5.1.2.7 Shopping Enjoy ment: -.-ccceceeeescecseecceeeestesneneennenennenseceecesesasescesesenaeanersreensassensessnsecsarsassansenants 84

SH cổ O2 te ổ nh 6 6 6 1 84 5.1.2.9 Exploratory BchaviOor: cceceeererrrrerrrrrrrrrrrrrerrrerrirerirerrerirrrerreenrerrereerenrerrerernrnrtrrsrerrerre 84 5.1.2.10 8 1 ./ÔẢ)ÔÔÔÔ111 85 S.1.2.11 Challenges: . - - CSKH HC HH ng HH TH ch nh ng 80210098 85 5.1.2.12 xz(2317158110//10/00111 02225 1 85 5.1.2.13 Demographics: 22.-2-cescecesseccensnensenessessncassenessessasesaesecsessnsensnssaeccestessecenensenenesssnenenssesosaennes 86

Trang 10

9 10 a t2 „3 Ri Cc 3 ‹ œ œ 2 .' ¬ 1 38

5.2.2 HypOtÏeses f€SIINg -<~e-eeeeree~aeeeseererkeeerneee th nh nen nh ch hen nhe 90 2? .n< 104

ST S6 an 6 1.1 107

SECOND STUDY: BOOKSAMILLION.COM 109

6.l STUDY DESIGN . - << SH HH HH T0 nh The nh nh nh nh nh Hee 109 TINH, 2 , n6 S6 .111.1.1114 ill 6.1.2 Surmey Instrumenr and Data Varidbles -«-« -<«+seeseereesrrreeseeerxreesrereereerere 113 6.1.2.1 Unplanned and Cross-Selling Purchases: - :ccssscccsssesseseeeteeeeeeeeeaseasensnnesnncesneeeereseeseneaneseaese® 113

6.1.2.2 6.1.2.3 6.1.2.4 6.1.2.5 6.1.2.6 6.1.2.7 Perceived Control: 6.1.2.8 Exploratory Behavior: 6.1.2.9 6.1.2.10 6.1.2.I1 6.1.2.12

6.1.3 Sample and Descriptive Statistics

6.2 RESULTS ccccsesesncescecesesscacceesenscennecscesce nec conssennrsnssssssensnacsaccenecessesaseanesserseesesencererenserececsresorenees

G21 1 1 e 6.2.2 Hypotheses t€StiNg -.+ ~-~~+e~.Te+errhreeernkererensreereeesenenreeeereeeee kh re e¡i1e9.11e 7 ÔÔỎ

THIRD STUDY: BOOKSAMILLION.COM 136

TN Š 97c S 0Ô 136

7.1.1 Surmey Instrument and Data VariaBles -~<<<-s=eeeeeseeeeeeererneeererrereeeeee 137

7.1.1.1 Personality TY€: - ca T22 nh TH nh nh KH nen nen ng TH HH te H01 137 LẴ N HH (cac sac ae S ÔỎ 138

LH New cac sa na Ô 139

7.1.1.4 Shopping EnJOVm€TL: -< + +~-z se chen Trọn nh ch TH th ng ng HH1 T1 171 139 TALS Perceived se ổjô n 139 7.1.1.6 Exploratory BehaviOr: -.- -Ặ-cSc sen nhưng kh nh ngà g0 LH ng 139

7.2 SAMPLE AND DESCRIPTIVE STATISTICS - - - - - 55 5S Ăn HH nh gi g0 em e 139 r "05 140

TV.Bo1 SCONES coe cececceccecccccecncccccccccccecccccccecccetenectaccscecseccececencetecsscescosscescnsccsssccnsececsececeeccrscsaceseoacensese 140

7.3.2 Hypotheses Testing - ccsccceencceseeeeneeceeenceneneeecneensnneneneeneesseceecencenenseeeseneregenesesssseessonease 141

9 1e9 19 .Ô 145

SUMMARY OF FINDINGS AND IMPLICATIONS 148

8.1 THE SHOPPING EXPERIENCE COUNTS ccccceccceeeccceeeceesecccceccecsceecccecernescccnesenserececssseeensenessecceneeee 149 8.2 HN: ¡77 v00 ,0 v27 (0n 152 8.3 NEW CUSTOMERS ARE DIFFERENT THAN REPEAT ONES 55 Ặ 5S SH HH ng mang ng 153 8.4 VALUE-ADDED INFORMATION AND POSITIVE CHALLENGES ENHANCE THE CONSUMER EXPERIENCE

154

8.5 CUSTOMER NEED SPECIFICITTY CAN DETERMINE CONSUMER BEHAVIOR . -<<~ se seeee 154 8.6 THE IMPORTANCE OF CUSTOMER PERSONALITY IS WEAK . -ẶĂẶ Ăn H« HH nh ng em nh 15S

8.7 ONLINE CONSUMER DEMOGRAPHICS DO NOT DETERMINE BEHAVIOR -. << sen 1S7

FUTURE RESEARCH 158

Trang 11

11 APPENDIX A: SURVEY INSTRUMENT USED IN FIRST FIELD STUDY WITH

KOZMO.COM™ 167

12 APPENDIX B: SURVEY INSTRUMENT USED IN SECOND FIELD STUDY WITH

BOOKSAMILLION.COM™ 173

13 APPENDIX C: SURVEY INSTRUMENT USED IN THIRD FIELD STUDY WITH

BOOKSAMILLION.COM™ 180

14 APPENDIX D: SUMMARY OF RESULTS FOR ALL THREE EMPIRICAL STUDIES 185 15 APPENDIX E: REGRESSION OUTPUTS FOR FIRST EMPIRICAL STUDY ccccceesere 186

15.1 REGRESSION TO TEST HYPOTHESES | A THROUGH 3A (RELATIONSHIP BETWEEN INTENTION TO RETURN AND EXPERIENTIAL VARIABLES): cccccccccsssscssccccsecececsecceeccceccoccecsececereerereccceesesenssenensneaessoae enone 186

15.2 REGRESSION TO TEST HYPOTHESES 78 AND 8B (RELATIONSHIP BETWEEN SHOPPING ENJOYMENT

AND CHALLENGES AND USE OF VALUE-ADDED SEARCH MECHANISMS): cccccecssccocccessnseeeceeseeeeneeeeeeees 190

15.3 REGRESSION TO TEST HYPOTHESES 7C AND 8C (RELATIONSHIP BETWEEN CONCENTRATION AND

CHALLENGES AND USE OF VALUE-ADDED SEARCH MECHANISMS): -s cecseeerecseenseceececcecenseecereseeese eens 193 16 APPENDIX F: REGRESSION OUTPUTS FOR SECOND EMPIRICAL STUDY 000000 197

16.1 REGRESSION FOR HYPOTHESES 1A THROUGH 3A (RELATIONSHIP BETWEEN INTENTION TO

RETURN AND EXPERIENTIAL VARIABLES): S5 - 5Q SSS SH ng HH ng r2 TH nh nen 197

16.2 REGRESSION TO TEST HYPOTHESES 4 AND 5 (RELATIONSHIP BETWEEN INTENTION TO RETURN

AND TAM VARIABLES) = .0 c-sseesescenececcececcesensennscecececsassececececsececenecenseccscssecsecesessneesnnensensssaanseseaeeqneerees 200

16.3 REGRESSION TO TEST HYPOTHESES 1A-3A 4, AND 5 COMBINED (RELATIONSHIP BETWEEN

INTENTION TO RETURN AND EXPERIENTIAL AND TAM VARIABLES): . 200sccccccereccccecsecereenereeesecceeeeoeers 203 16.4 REGRESSION TO TEST HYPOTHESES 7C THROUGH 8C (RELATIONSHIP BETWEEN CONCENTRATION AND a a0 ha ơi 85306 Ơ 207

16.5 REGRESSION TO TEST HYPOTHESES 7B THROUGH 8B (RELATIONSHIP BETWEEN SHOPPING

ENJOYMENT AND INDEPENDENT VARIABLES)? .:0cccccccceensencncccececeneccccccereceeccersesesscssnsnecceneecseneneesroeee® 210

16.6 REGRESSION TO TEST HYPOTHESES 10 (RELATIONSHIP BETWEEN CONCENTRATION AND

INDEPENDENT VARIABLES FOR LOW NEED SPECIFICITY SUB-SAMPLE): -.scccsesseseceeessseeseesneteneenensecroes 213

16.7 REGRESSION TO TEST HYPOTHESES [0 (RELATIONSHIP BETWEEN SHOPPING ENJOYMENT AND

INDEPENDENT VARIABLES FOR LOW NEED SPECIFICITY SUB-SAMPLE): -ccseseeecceesseneeeeeececeereneesnneneeaes 216 17 APPENDIX G: REGRESSION OUTPUTS FOR THIRD EMPIRICAL STUDY sescoeeee 219

17.1 REGRESSION FOR HYPOTHESES IA THROUGH 3A (RELATIONSHIP BETWEEN INTENTION TO

RETURN AND EXPERIENTIAL VARILABLES): - -S SH HH ng nen HH n9 00h vn 219

18 APPENDLX H: GRAPHICAL REPRESENTA TION OF RESULLTS <-<==<<=<ss 222 FIRST EMPIRICAL STUDY: KOZMO.COM - St ng HH ng 0028 0 00994 222

SECOND EMPIRICAL STUDY: BOOKSAMILLION.COM - Q SH HH ng ng net th n0 1 224

Trang 12

"{Onsale.com] is a transaction site that knows psychology ts the key" Roger Black, Internet World, May 11, 1998

"Superstore book retailers have to (1) be efficient at helping customers find and purchase the items they want, and (2) serve as an enjoyable

‘destination.’ Amazon succeeds at both counts" Louis Rosenfeld, Interent World, March 2, 1998

1 Introduction and Motivation

The World Wide Web (WWW web) has evolved dramatically in a few years from a medium for sharing simple hypertext documents to a complex medium used for entertainment, education, research, and commerce Early on it became clear that the web had some major advantages that made it highly attractive for commercial use: cheap and instantaneous communication with global customers; customization of communication and marketing to individuals; and instantaneous distribution of information products (Burke, 1997a: Peterson et al., 1997) Initially, the publishing nature of web commerce evoked comparisons with catalog-based home shopping However, its interactive nature makes the web the best current alternative to the physical world for commerce since it provides all the benefits and conveniences of catalog shopping while providing an experience similar to physical-world shopping (Johnson et al., 1998) According to Forrester Research, online shoppers will grow to 40.3 million U.S households by the year 2003 (Dykema et al., 1999) and online retail spending is projected to reach over $100 billion by the year 2003 (Rowen, 1999) Still, the web remains a rather poor medium compared to the physical world since it doesn't use the sense of taste, smell, or

touch (Nowlis and McCabe, 2000) and technology limitations restrict its ability to fully

Trang 13

Despite its limitations, web-based commerce is widely seen as the next wave of

commerce Many companies rushed to establish a web presence early on, sometimes with not very successful results (Dutta et al., 1998) Today, on-line businesses like Yahoo!, Amazon.com, and Ebay enjoy high revenues, a fast-growing customer base, and high stock prices With each passing year, the total income generated from on-line businesses grows substantially Books alone are expected to generate online sales of over $3 billion by the year 2003 (Rowen, 1999)

Companies who sell on the web share similar concerns with traditional companies regarding their customer base: how to attract customers, how to get customers to buy their products, how to get customers to return to their stores, and so on The web, however, presents some unique challenges to the companies that use it for business Researchers have only just begun to study the multitude of factors that can impact consumer behavior in web-based and other technology-based commerce (Bellman et al.,

1999; Morrisette et al., 1999; Novak et al., 1998a, 1998b; Alba et al., 1997; Burke, 1997a 1997b; Hoffman and Novak, 1996; Berthon et al., 1996)

Trying to understand the factors that impact online consumer behavior is made more complicated by the fact that the main entities in the commercial world have changed Specifically, in business-to-consumer electronic commerce we have seen two major transformations (Figure 1):

Trang 14

evaluating alternatives, making purchasing decisions, and requesting customer

service However, he is performing all these activities on a computer while interacting with a system, i.e a commercial web site He, therefore, exhibits all the

characteristics of a computer user We expect that the e-consumer will not necessarily behave exclusively like a traditional consumer or a traditional computer user, but instead will develop a behavior that is the result of the interaction of those two roles Understanding this unique behavior is one of the biggest challenges of web based businesses

Trang 15

e-Consumer e-Store " | Information., mem ~ Technology |

Figure 1: Transformations in Electronic Commerce

In order to understand how these two entities, the e-consumer and e-store, interact we must bring together different lines of research If we were to study the interaction of traditional consumers and stores, we would turn to Marketing and Psychology for guidance If we were to study the interaction of computer users and Information

Technology, we would tur to Information Systems and Psychology (again) for guidance Now that these four elements have been combined into the new entities of e-consumer and e-store, we must combine the three reference fields (I.S., Marketing, and Psychology) to understand the dynamics of this new electronic commercial environment

In this thesis we bring the three fields of research together to examine the consumer

experience on the web and its effects on consumer behavior online We will specifically

Trang 16

concentrate on two major consumer behavior variables: customer intention to return and number of unplanned purchases made We expect that our study will lead us to adapt or add to the prior research to better explain the nuances of web-based commerce and more specifically the factors that determine consumer purchasing activity in this new

commercial environment We believe that the study of consumer behavior on the web is a useful and necessary endeavor that can have important implications both for academic research as well as for the growing industry of electronic commerce

Trang 17

2 Summary of Thesis

We begin in Section 3 with a literature review of the published research related to consumer behavior in business-to-consumer electronic commerce We review separately the theoretical work that has been done in the area and the empirical studies that have been published The review lays the foundation for Section 4 where we develop our own theoretical framework for online consumer behavior

We begin the section with a discussion of four assumptions that enable us to perform this study in the turbulent and constantly growing technological and social environment of the World Wide Web Following that, in Sections 4.2-4.4, we discuss the three reference disciplines that our entire study is based on: Marketing, Psychology, and Information Systems Finally, in Section 4.5 we introduce a theoretical framework that incorporates some of the most significant factors in consumer behavior on the web

Trang 18

There are many factors that can impact our two dependent variables For example, customers may choose to retum to a store because of satisfaction with either the product quality or the customer service Some factors that may impact the number of unplanned purchases are the time and money available to the customer for shopping and the

marketing promotions the customer is exposed to (Beatty and Ferrell, 1998) For our study we will concentrate on one common factor that can affect both customer return visits and unplanned purchases: the consumer experience We will examine how

consumer attitudes about their shopping experience in an e-store impact their intention to return and the number of unplanned purchases they make It is important to note that we will not deal with issues of advertising and attracting customers to a certain web site but rather we will study customers while they are visiting an e-store In order to define and measure the consumer experience and attitudes we will use three attitudinal variables: perceived control, shopping enjoyment, and concentration These variables are taken from the literature on flow (Csikszentmihayli, 1975) and they are described in detail in Section 4.5.2 We then develop our first set of hypotheses on the impact of the consumer experience on our dependent variables in Section 4.5.3

If we can show that consumer attitudes about their shopping experience can have a strong effect on their intention to return and their unplanned purchases, the next logical question is: How can we make the consumer experience and attitudes as positive as possible? We examine two sets of variables that can provide some answers to that question The first set includes individual consumer factors and is discussed in detail in Section 4.5.4 The second set of variables includes characteristics of the e-store and is discussed in Section 4.5.5 We believe that the interaction of these two sets of variables

Trang 19

can significantly influence the consumer experience and attitudes which then in turn can have a strong impact on consumer intention to return as well as the number of unplanned purchases made Therefore, we develop our second set of hypotheses on the impact of individual and environmental factors on the consumer experience in Section 4.5.6

In Section 5 we describe our first empirical study: an online field study performed with the company Kozmo.com™, an online video-rental and delivery store We describe the study design, the survey instrument and implementation, and we discuss the results of our data The purpose of the study is to empirically test the theoretical framework we described in Section 4.5

Due to some limitations of the first study and to enhance our empirical testing, we performed a second online field study described in Section 6 For this study, we used the company Booksamillion.com™, an online bookseller We replicated much of the first empirical study but also added the Technology Acceptance Model (Davis, 1989) with the intention of testing its predictive ability against our consumer behavior framework The results of the test are described in Section 6

Finally, we performed a third smaller study described in Section 7 This was also an online field study with Booksamillion.com™ where we tested the effects of personality type on consumer behavior While we had examined this relationship in the first

empirical study, we did not include it in the second one in order to keep the survey instrument from becoming too long

We end the thesis in Sections 8 and 9 with a discussion of our major findings and the major implications of the results for academic research as well as for the industry We finish with a discussion of future research

Trang 20

3 Consumer Behavior On-line: Literature Review

Since the web has only been around for less than a decade and it has only been used for commerce for a few years, there has only been limited research published on the subject of consumer behavior online Early research has tended to be theoretical and speculative in nature As the technology matured and the consumers became more familiar with it, empirical studies started to appear that tested the impact of specific factors on consumer behavior on the web

In this chapter we will review the literature that has dealt with the various issues regarding how consumers think, feel, and act when they use the web for commerce We will first look at the research that has dealt with the theoretical issues in this area Most of these papers speculate on what the impact of the new medium will be on vanous

marketing issues including consumer behavior Then we will review the empirical studies that have attempted to verify some of those theoretical speculations as well as identify other important variables and relationships in web commerce In some cases, the studies we will discuss deal exclusively with consumer behavior on the web In many cases though, how consumers behave on the web is only one of many issues they examine We will not delve deeply into the other issues, but rather we will concentrate only on the parts of the research that are relevant to consumer behavior online

3.1 Theoretical Studies:

One of the earliest papers touched on some of the factors that would be important

when trying to understand consumer behavior online and was written by Quelch and Takeuchi (1981) Years before the invention of the web, the authors discussed the future

Trang 21

of non-store marketing and touched upon some of the issues that are important for

consumers that use interactive cable television as well as interactive information retrieval The latter term refers to systems like Britain’s Prestel and France’s Antiope where

consumers access computer databanks either by phone or through their TV This was an early ancestor of the web and Quelch and Takeuchi predict that such users will take time feeling at ease with the technology and younger consumers will adopt it first They also question whether the availability of more information on more products will lead customers to make immediate purchases

In their paper on the demise of advertising, Rust and Oliver (1994) predict that reaching the customer will be easier due to the customization tnat the new technologies will enable At the same time, they expect that it will be more difficult due to the fact that mass advertising will not be as easy to get to the customer who will now be much more empowered and have more control over what she sees and hears

Hoffman et al (1995) present their view of how the web can be best utilized for commerce They describe six different types of commercial web sites: L) online storefront, 2) Internet presence, 3) Content, 4) Mall, 5) Incentive Site, and 6) Search Agent They point out that there is almost no knowledge regarding online consumer behavior and they identify one of the biggest challenges to be how to present the customer with an experience comparable to that of physical stores In a more detailed theoretical analysis of online consumer behavior, Hoffman and Novak (1996) address the role of marketing in computer mediated environments and specifically on the web They present a structural model of consumer navigation behavior that uses the concept of flow (Csikszentmihayli, 1977, 1975) as a measure of the “optimal experience” of an online

Trang 22

consumer In a later paper, Hoffman and Novak (1997) discuss more marketing issues in electronic commerce, including consumer behavior They identify the ability of online consumers to not only consume content but also provide content in the commercial environment The authors discuss the difficulty for consumers to achieve “telepresence” i.e the perception of being present in the online rather than the physical world They again propose the concept of flow (Csikszentmihayli, 1977, 1975) to measure consumer behavior Once again, the increased level of control the consumer assumes in electronic commerce is pointed out A more detailed discussion of all the papers by Hoffman and Novak as well as other flow literature follows in a later section

In one of the first studies that attempts to build an artifact that addresses some of the issues of online consumer behavior, Baty and Lee (1995), propose an architecture for electronic shopping infrastructures and describe a prototype called InterShop This

architecture addresses some of the issues regarding the vendor/customer interaction in the electronic medium and provides easier navigation, more comparability of products and vendors, and improves the information processing capabilities of customers This is achieved by the provision of a knowledge representation scheme that vendors can use to organize products in a structural hierarchical fashion That architecture is then used to create an easy navigation system that enables the customer to easily search and locate products

Berthon et al (1996) develops a model to help understand how ‘web surfers’ are converted to web customers The model includes a series of efficiency measures at various stages of the conversion process such as an awareness efficiency factor that

measures the percentage of total surfers that become ‘aware surfers,’ i.e surfers who

Trang 23

become aware of a web site The model differentiates between those who are active and those who are passive information seekers In other words, the authors recognize one important difference in how web customers search for information on the web While the model seems robust, it is too simplistic to account for all the factors that convert web users to customers

Quelch and Klein (1996) discuss the global aspects of marketing on the Internet In their paper they also point out that tracking consumer behavior and attitudes is easier on the Internet They identify six ways of collecting data on customers: a) online surveys, b) bulletin boards, c) web visitor tracking, d) advertising measurement, e) customer

identification systems, and f) email marketing lists Looking at the state of customer data collection today, we can see that the authors foresaw all the major methods of customer tracking on the web

Alba et al (1997) concentrate on the informational aspects of consumer behavior online and specifically the factors that influence a consumer’s information acquisition and purchase decision online They identify the need to enable customers to quickly and comprehensively inspect a large set of product alternatives with reliable search results They also point out that customers must be able to compare alternative products in a customized way along the variables that are important to each individual consumer Without these capabilities, the authors predict that online shopping will only achieve the penetration of print catalogs

Burke (1997a) identifies the impact of the increasing control of online consumers on impulse purchases He says that since they can choose to go directly to specific products without having to be exposed to marketing promotions, impulse purchases can decrease

Trang 24

He also claims that some of the attributes that communicate a brand such as shapes and colors are replaced by detailed information on the contents, prices, and features now available online In addition, he points out that the enhanced ability to compare prices can detract attention from brand names and commoditize the products

Peterson et al (1997) point out that due to the fact that many consumers use the shopping experience as a source of enjoyment, they may not use the Internet for shopping Other reasons for not using the Internet according to the authors are lack of access fear of the technology, and resistance to change They add that other consumers will use the Intermet to make better purchase decisions by gathering and comparing product information online before buying offline

Sheth and Sisodia (1997) discuss the future of consumer behavior and they identify two major factors that will impact the way consumers behave: new technologies and changing lifestyles and demographics They point out that new technologies will give consumers unprecedented levels of control over the information exchanged between them and vendors The authors describe how the technology will move to time and location independent marketing Consumers will demand convenience in terms of being able to transact anywhere and anytime They identify eight emerging trends in consumer behavior:

e Disintermediation and Reintermediation e Personalization and Reaggregation e Shopping on Demand

e Consumers as Co-Producers e Greater Value Consciousness

Trang 25

e Blurring of Consumer and Business Markets e Power Shift to Consumers

e Automation of Consumption

McGaughey and Mason (1998) examine the classical decision-making model of consumer behavior (Engel et al., 1990) They discuss each of the five steps (problem recognition, information search, evaluation of alternatives, purchase decision, post- purchase behavior) and discuss the impact of the Internet on the entire process They point out that product involvement will moderate the level of information search

customers will engage in They also stress the need for online commerce to create higher levels of arousal to engage consumers in more active levels of information search The authors also briefly discuss the various decision rules consumers may use to evaluate alternatives such as the compensatory rule and the lexicographic rule

Finally, Cook and Coupey (1998), in their paper on regulatory issues in electronic marketing, address the impact of online commerce on consumer information processing They describe three levels of influence of the electronic medium on information

processing: a) Minimal, where the amount of information used is influenced, b) moderate, where the form of information display is influenced, and c) extensive, where the content of information used is influenced They propose that this framework is used in research regarding regulatory issues in electronic marketing

3.2 Empirical Studies

Trang 26

behavior, their results provided some early indications on how consumers act in an electronic shopping environment The researchers tested two types of grocery store simulations, one rudimentary “bare-bones” environment and one more realistic

environment where actual pictures of the products were presented in virtual shelves They tested the differences between consumer behavior in an actual grocery store and that in the two simulated environments They found that for products that can more easily be represented in a computer simulation consumers exhibited a more similar behavior to the real grocery stores They also found that the more realistic simulation caused subjects to make purchasing decisions more similar to the actual store than the rudimentary

simulation On the other hand, they found that subjects were more sensitive to

promotions in the rudimentary simulation than in the realistic one Subjects bought larger quantities of products in the computer simulations than in the actual store They also included fewer brands in their consideration sets and switched between brands less often in the simulated environment than in the actual grocery store

One of the first studies of actual web consumers was conducted by Jarvenpaa and Todd (1997a and 1997b) The authors assigned 220 web users shopping tasks and provided them with pointers to shopping sites After the participants finished their tasks they answered an open-ended survey The study found that attitude towards web

Trang 27

to shop on the web was influenced by product perceptions, shopping experience, and service quality on the web

In a similar study, Rice (1997) conducted a survey of visitors of 87 web sites to determine what factors are important in customer decision to make repeat visits to a web site He found that good, “non-frivolous” content was the number one factor followed closely by how much the customer enjoyed her visit to the web site Other factors include the layout of the site, its uniqueness, and ease of finding information

Westland and Au (1998) conducted an experimental study to determine whether there are any significant differences between three types of digital retailing: a) an electronic catalog that displays products in a linear fashion much like a print catalog, b) bundling, where products are pre-selected and bundled together, and c) virtual reality where a three- dimensional recreation of a store is used The virtual reality condition was implemented in kiosks where the response time was nearly instantaneous (as opposed to the Internet) The authors found that there were no significant differences in the money spent or the number of items bought in the three conditions They did, however, find that subjects in the virtual reality condition took significantly longer to search for products and make a final purchase decision

Trang 28

convenient access to it was the main reason for using the web, and e) users were not interested in electronic commerce because of the lack of trustworthy web stores

Maignan and Lukas (1997) also performed a qualitative study of web users by interviewing 25 subjects They found that web users see the web as four things: a) a source of information, b) a place or object of consumption, c) a communication tool, and d) a social system They found that new Internet users, who accessed the web often were more likely to see the Internet as a place of consumption, i.e see themselves as web consumers They found that though web users welcome commerce, they do not favor traditional advertising and marketing promotions online Web customers strongly associate their experience with fun and convenience and traditional types of advertising may not agree with such associations

Eighmey (1997) and Eighmey and McCord (1998) performed a pilot study and a subsequent field application to examine how web customers respond to commercial sites Both the pilot and the field study yielded three major findings: a) Web customers prefer web sites that provide useful information in an enjoyable context b) web customers favor sites that use “metaphorical structures” that show users how to understand and navigate the site, and c) ease of use from efficient web design is an important issue to web customers

Trang 29

found that the “customer support” interactivity function has a significant positive effect on the quality of the web site

Lohse and Spiller (1998a and 1998b) used a regression model to predict web store traffic and sales in dollars based on certain functions of the web site They found that traffic measured in number of visits, is positively related to the number of products available, the availability of a FAQ section, the number of “store entrances,” the availability of promotions on the store entrance, and it is negatively related to the

existence of a customer feedback section (the authors do not explain this counter-intuitive result) The authors also found that sales are positively related to the existence of a

customer feedback section, better product lists, the number of “store entrances,” and the availability of promotions on the store entrance

In an extensive study, Korgaonkar and Wolin (1999) applied the uses and

gratification theory to understand web usage They identified seven motivations for using the web and studied their impact on consumer behavior They found that time spent on the web was positively related to the motivations of social escapism, information

gathering, interactivity, socialization, and shopping bargains (economic motivation).They also found that heavier users were better educated and had higher income levels As far as the type of web use, the authors found that those who exhibited heavy web use for

Trang 30

socialization, and shopping bargains and it was negatively related to transaction-based security and privacy concems They also found that those who purchased online tended to be older and male Finally, the motivations of social escapism shopping bargains, and interactivity as well as age and income were positively related to the frequency of online purchases

In a report for Forrester Research, an online research company, Morrisette et al (1999) find that unlike in offline commerce, online customers can be satisfied but still not be loyal to online vendors In fact, the researchers found that overall, 75% of satisfied web customers said they would switch to a different web store, highlighting the low levels of customer loyalty in web commerce

Novak et al (1998b) conducted a large study to test an updated version of their theoretical model of measuring online consumer behavior, which they presented in Hoffman and Novak (1996 and 1997) They collected a large pool of data in conjunction with the 9" WWW User Survey that was fielded by the Graphics, Visualization, and Usability (GVU) Center at the Georgia Institute of Technology They then used a structural model approach to test the model that uses “flow” to measure the online consumer experience They found that high levels of skills, challenges, arousal focused attention, and control could measure flow on the web They also found that flow is positively related to interactivity and telepresence

In a panel study (Bellman et al., 1999) conducted as part of the project called the Wharton Virtual Test Market (WVTM) has examined the personal characteristics of online consumers that could predict whether they would buy online or not The

researchers found that searching for product information online was a significant

Trang 31

predictor of online buying They also found that a “wired lifestyle” is also typical of an online buyer Such a lifestyle consists of a long tenure on the Internet a large email volume, working on the Internet, and an improvement of personal productivity due to the Internet Another interesting finding, is that “time starvation,” i.e little discretionary time

for leisure and shopping, is also a predictor of online buying

Xia and Sudharshan (2000) performed an experimental study to test the effects of interruptions on the online consumer decision process They implement interruptions as pop-up window advertisements and they manipulate the level of control subjects have over viewing those ads They find that interruptions can increase the time and

information consumers need to complete a task online but they found no effect on the frequency of interruptions Also, the effects of interruptions were different for concrete versus abstract goals When subjects had control over the interruptions they had less negative attitudes about them Finally, they found that interruptions reduced consumer satisfaction with the shopping experience but not with their decision outcome

In another experimental study, Nowlis and McCabe (2000) tested the effect of the ability to physically touch products on consumer decision making both online and offline They found that consumers were more likely to choose products that have material

texture properties(such as apparel or towels that have different textures) or material part properties (such as pens and cameras that consists of many interlocking parts) in offline

environments rather than in online ones They found no such difference for products with

geometric properties (such as packaged food which comes in fixed geometric shapes)

Trang 32

offline and online consumers They found that indeed there were significant differences in the consumer choice process between the two groups They found that compared to offline consumers, online consumers are less price sensitive, prefer larger to smaller product sizes, have weaker brand loyalty, do more screening based on brand, do less screening based on size, and fewer of them do no screening at all

3.3 Summary of past findings

The limited past research on online consumer behavior provides us with some knowledge on how consumers behave online We have seen that the representation of the product online matters on both shopper attitude towards sopping online as well as their intention to buy (Burke et al., 1992; Jarvenpaa and Todd, 1997a and 1997b; Nowlis and McCabe, 2000) The consumer attitude towards web shopping as well as the intention to buy are also influenced by the quality of the shopping experience (Jarvenpaa and Todd,

1997a and 1997b) Intention to buy online is also influenced by the level of the consumer's wired lifestyle and lack of leisure time (Bellman et al., 1999)

Customer loyalty on the web is low overall (Morrisette et al., 1999) and it is influenced by the shopping enjoyment of the consumer online as well as the availability of good, relevant content on the web site (Rice, 1997; Eighmey, 1997, Eighmey and McCord, 1998)

The locus of power is shifting from the vendor to the consumer (Raman, 1997) who does not favor traditional advertising and promotions online (Maignan and Lukas,

Trang 33

Finally, certain site features such as the availability of an FAQ section or

Trang 34

4 Theoretical Framework - Hypotheses Development In this chapter we will discuss the theoretical framework that we have developed in order to better understand the dynamics of consumer behavior in business-to-consumer electronic commerce We will begin with a brief discussion of four assumptions that underline our framework and our empirical studies We will then discuss three relevant models or frameworks from the three reference disciplines of Marketing, Psychology, and Information Systems Our theoretical framework will be strongly based on the cognitive consumer behavior model from Marketing and the behavioral model from Environment Psychology, as described below We will develop a set of hypotheses from that framework and we will test them using three empirical field studies We will also discuss the Technology Acceptance Model from Information Systems and compare it empirically against our framework of online consumer behavior

4.1 Initial Assumptions

In undertaking this research study we subscribe to the four assumptions about Internet commerce made by Peterson et al (1997)' These assumptions enable us to pursue this research in the face of rapid changes in the Internet world and especially related to the WWW The four assumptions are:

1 Eventually there will be almost universal access to the Internet and the World Wide Web The current population of web users is significantly smaller than and not representative of the general population (in the U.S but also in the rest of the

' The authors thoroughly discuss these assumptions and support them with the relevant literature so we

will not repeat their arguments here for the sake of brevity

Trang 35

world) However, we assume that in the foreseeable future, the vast majority of the U.S and the rest of the developed world will have access to the Internet and the WWW We also assume that the population of web users will be representative of the general population

N Marketing on the Internet and the WWW will not increase overall consumer spending There is no apparent reason why electronic marketing will in and of itself make consumers spend more Instead, we assume that good electronic marketing can redistribute consumer spending more heavily toward electronic commerce

3 Any potential changes in the infrastructure of the Internet and the WWW (telecommunication technologies, access devices, etc.) will not significantly change the impact of the Internet and the WWW on consumer marketing Also problems with the infrastructure of the Internet and the WWW will eventually be solved and they will not cause the demise of the Internet

4 Issues of transaction security and privacy on the Internet and the WWW will be solved in the short-term and so research on electronic consumer marketing should be independent of them

We believe that these four assumptions are valid and reasonable Also, they enable us

to study consumer marketing on the Internet and the WWW free of the constraints of short-term problems in the rapidly changing on-line world

Trang 36

4.2 Reference Discipline 1: Marketing

Figure 2: The traditional framework for Consumer Behavior

INDIVIDUAL FACTORS

Consumer Resources Motivation & Involvement Knowledge Attitudes Personality, Lifestyle, Demographics ENVIRONMENTAL FACTORS Culture Social Class Personal Influence Family Situation DECISION PROCESS e Need Recognition e Search e Alternative Evaluation e Purchase e Outcomes

The purpose of our study is to examine the factors that influence consumer purchasing activity in web-based commerce In traditional commerce, these questions are dealt with in a specific area of Marketing known as consumer behavior Therefore, we tum to traditional consumer behavior research as the first part of the foundation of our

theoretical framework The area of consumer behavior was born in the late 1950s - early 1960s and was based on the marketing concept which states that “marketers must first define the benefits consumers seek in the marketplace and gear marketing strategies accordingly” (Assael, 1992) Since then, the study of consumer behavior has expanded into a major part of the field of marketing A good definition of consumer behavior is provided by Engel et al (1990):

31

Trang 37

We define consumer behavior as those actions directly involved in obtaining consuming and disposing of products and services including the decision processes that precede and follow these actions

The most influential approach for studying consumer behavior has been the cognitive approach which views the actions of the consumer as the result of a decision process There are different variations of a complete framework for the study of consumer

behavior under the cognitive approach (Howard, 1994; Assael, 1992: Engel et al., 1990) Most of these frameworks share three components The first component and core of the framework is the customer decision process that consists of five steps:

t

Need Recognition: At this stage the customer becomes aware of an unfulfilled need Need recognition can be the result of different stimulants such as

advertisements, comments by friends, or simply the absence of a needed product Search: Having recognized a need, customers search for information and/or product alternatives to fulfill that need The search can be internal, i.e in an individual's memory, or external, i.e in the outside world

Alternative Evaluation: Having searched for and found alternatives that can fulfill their need, customers proceed to evaluate them based on certain criteria and use different evaluation methods The result is usually a decision to purchase one of the alternatives

Purchase: Customers purchase the product that they have chosen through the means provided by the seller and the transaction environment

Trang 38

The other two components are the factors that can influence the consumer decision process: individual factors and environmental factors The main individual factors in the traditional framework are:

1)

3)

4)

Consumer Resources: Consumers have three types of resources:

a) Time: Consumers have a limited amount of time to allocate for shopping since most of their time is spent at work or at other leisure activities

b) Economic resources: Money is also a limited resource for consumers It can take the form of income, wealth, or credit

c) Cognitive Resources: The ability of the consumers to process information is bounded by the limitations of the human brain

Motivation and Involvement: Customers are also different with respect to how

motivated they are when they purchase a product A consumer’s level of involvement and interest in a product can have an impact on their decision process

Attitudes: Consumers can have different beliefs and preferences towards products, brands, retailers, advertisements, and so on These consumer attitudes can have an effect on peoples’ intention to buy certain products

Personality, Lifestyle, and Demographics: Consumers can display different personalities that are behaviors that are persistent over time They may also be characterized by different ways of living commonly called lifestyles Finally, customers can be clustered under different demographic variables such as age,

income, and race All these variables can extensively influence the consumer decision

process

Trang 39

are:

1)

3)

4)

5)

The main environmental factors in the traditional consumer behavior framework

Cultural and Ethnic Values: Various traditions, customs, and values that are part of the consumers’ cultural or ethnic groups can influence purchasing behavior

Social class and status: The consumer's place in the social class system can also have an effect on the decision process whether it is at a national level or the micro-society of a neighborhood

Personal influence: The consumer decision process can also be affected by various reference groups such as friends colleagues, as well as by strangers, commonly referred to as word of mouth

Family and Household influence: The family and other members of a consumer household are a very important influence group for consumers A spouse’s opinion or a child’s demand can strongly sway a customer decision

Situational factors: Finally, a consumer can be affected by a variety of situational factors One example is the retail environment that includes the layout of products as well as the music playing in a store Another example is the information environment that involves the availability and format of product information

Figure 2 presents the traditional framework of consumer behavior drawing heavily from the frameworks by Engel et al (1990) and Assael (1992)

Trang 40

4.3 Reference Discipline 2: Psychology

The study of online consumer behavior inevitably will involve a number of psychological variables Therefore, we tum to psychology, and more specifically environmental psychology, as the second part of the foundation of our theoretical framework Environmental and individual characteristics are known to influence

emotional responses that, in tum, affect behavioral responses as seen in Figure 3 Three

THE PRIMARY ENVIRONMENT EMOTIONAL RESPONSES BEHAVIORAL e Pleasure >| RESPONSES e Arousal e Dominance PERSONALITY

Figure 3: Behavioral Framework from Environmental Psychology (Adapted from Mehrabian and Russel, 1974, p 8)

commonly identified emotional responses are Pleasure, Dominance, and Arousal (Mehrabian and Russel, 1974) These factors may also be important determinants of consumer behavior, both off-line and on-line In a consumer context, Pleasure can be defined as the enjoyment of the shopping experience, an important variable off- and on-

line Dominance is the equivalent of perceived control felt by customers, also equally important in both the physical and the electronic worlds Finally, Arousal deals with more physical sensations such as being jittery, sluggish, or relaxed and is therefore more

important for the time being in the physical world However, with technology advances, physical arousal could become an important factor in on-line consumer behavior