JCMC 5 (2) December 1999
Collab-U CMC Play E-CommerceSymposium Net Law InfoSpaces Usenet
NetStudy VEs VOs O-Journ HigherEd Conversation Searching for Cyberspace
Browsers or Buyers in Cyberspace? An Investigation of Factors Influencing Electronic Exchange
Isenberg School of Management
University of Massachusetts-Amherst
Department of Management and Business
Bharat P. Rao
Institute for Technology and Enterprise
Table of Contents
- The Conceptual Model
- Vendor Characteristics
- Perceived Security of Transactions and Concern for Privacy
- Consumer Characteristics
- Likelihood of Electronic Exchange (PUR)
- Model Development
- Additional Analyses
- Implications of the Study and Future Research
- About the Authors
AbstractIn its current form, the Internet is primarily a source of communication, information and entertainment but increasingly also a vehicle for commercial transactions. An understanding of reasons for purchasing on the World Wide Web is particularly relevant in the context of predictions made regarding electronic shopping in the future. In the paper, we focus on some of the antecedents to electronic exchange in the online context. In particular, what are some of the factors influencing online purchasing behavior? What is the role of privacy and security concerns in influencing actual purchase behavior? How do vendor and customer characteristics influence consumers' propensity to engage in transactions on the Internet? We analyze secondary data from an e-mail survey. The study has implications for both theory and practice. The findings extend our knowledge of factors influencing marketing exchange from the traditional setting to the internet context. In addition, the findings regarding factors enhancing the propensity to shop online have implications for internet retailers seeking to enlarge their online customer base.
IntroductionThe Internet has emerged in the recent past as a dynamic medium for channeling transactions between customers and firms in a virtual marketplace. The growth of the Web has been phenomenal, and there has been a corresponding growth in commerce on this robust platform. Varying estimates of its impact on online retail and shopping abound, and the estimates point to dramatic growth. It is projected that online shopping will grow from $11 billion in 1999 to $41 billion in 2002 (National Retail Federation, 1999). The US Commerce Department citing a study by Forrester Research suggests that online retail trade ranging from $7 to $15 billion in 1998 will reach anywhere from $40 billion to $80 billion by 2002 (US Department of Commerce, 1999).(1)
Many companies are already online, exploring and shaping this new opportunity. Amazon.Com offering more than two million titles of books and CDNow are prime examples of companies that are successful in the online space as evidenced by their stock market capitalizations (Zwass, 1999). According to the Travel Industry Association of America (TIA), the number of Americans using the Internet to plan vacation or travel grew from 10% in 1996 to more than 25% in 1998 (Kate, 1998).
The rapid growth of this new medium poses intriguing questions for academic research. To date, researchers have focused on the role of the Web as an information and communications medium (Hoque and Lohse 1999; Lynch and Ariely 1998; Alba et. al 1997; Hoffman, Novak and Chatterjee 1996; Berthon, Pitt and Watson 1996; Hoffman and Novak 1995). Berthon, Pitt and Watson (1996) introduce the concept of conversion efficiencies of a Web site, which refers to the rate at which browsers are converted into buyers. While a number of authors have examined factors that may influence shopping on the internet, (e.g., Alba et al. 1997; Palmer 1997), much of this research is primarily conceptual in nature. Hoque and Lohse (1999) examine the impact of user interface design on information search costs in electronic media. Very little empirical research exists on issues relating to shopping on the internet.
In order to examine the various alternative shopping formats, a brief comparison of Web retailing with traditional retail, TV in-home shopping, catalog shopping on various dimensions of variety, trialability, asynchrony and interactivity is enclosed in Table I. As can be seen in the Table, the internet retailer offers the benefit of asynchrony, i.e., the internet retailer is available for shopping any time of the day or night, and the benefit of interactivity. The disadvantages are the inability to sample the product and the limited variety in terms of merchandise.(2)
Table ITransaction Types Compared
Variety Trialability Asynchrony Interactivity Full Retail High High Low High TV Moderate Low Moderate Low Catalogs Moderate Low Moderate Low Web Moderate Low High High
Scant research exists which examines factors influencing purchasing over the internet, an issue which is particularly relevant in the context of predictions regarding electronic shopping in the future. In particular, the following questions are of interest to the researchers: (1) What factors influence online purchasing behavior?; (2) What is the role of perceived risk in online purchasing on actual purchase behavior?; (3) How do differences in customer characteristics influence their decisions to shop online?
Our primary objective of this study is to investigate factors influencing commercial transactions in the online environment. To address this issue, we develop a model examining the factors influencing electronic exchange. The model will be tested on a sample of internet users. Some of these internet users are likely to exhibit a greater propensity to shop online. Therefore, the purpose of this study will be to examine factors influencing buying among internet browsers. In the following section, the model used in this study will be described and the hypothesis for each antecedent developed. Following that, the data analysis is presented. Finally, results, conclusions and directions for future research are presented.
THE CONCEPTUAL MODELIt has been long known that the exchange process is central to the concept of marketing (Sheth and Parvatiyar 1995; Morgan and Hunt 1994; Dwyer, Schurr and Oh 1987; Bagozzi 1975, 1974). The exchange system has been conceptualized as a set of social actors and their relationships to each other, and as the endogenous and exogenous variables affecting the behavior of the social actors in those relationships (Bagozzi 1974). The theory of exchange has evolved into the theory of relational exchange. Relational exchange theory or relationship marketing has a number of proponents and has been frequently used in the past studies (Morgan and Hunt 1994; Sheth and Parvatiyar 1995). Although the relational exchange literature primarily focuses on the determinants of long-term buyer-seller relationships, some of the concepts from the earlier literature on exchange (Bagozzi 1975, 1974) and the recent work in the relationship marketing area can be drawn upon in identifying factors influencing the likelihood of electronic exchange.
The fundamental exchange model (Bagozzi 1974) views the exchange process as a social influence process. Among the characteristics identified in the social exchange model as key antecedents of exchange are social influence, social characteristics of actors and third party effects. Thus, in the context of electronic exchange, the characteristics of the consumers and the vendors should affect the propensity to engage in a transaction. For instance, in regards to consumers, Sheth and Parvatiyar (1995) suggest that consumers’ sociological orientations may play an important role in increasing the propensity to engage in relationships. We investigate the role of consumer characteristics by examining the consumers’ shopping orientations as an antecedent to the likelihood of electronic exchange. In regards to vendors, literature on relationship marketing recognizes the role played by trust (Morgan and Hunt 1994; Moorman, Deshpande and Zaltman 1993). Trust is defined as confidence on the part of the trusting party that the trustworthy party is reliable, has high integrity and is associated with such qualities as consistency, competency, honesty, fairness, responsibility, helpfulness and benevolence (Morgan and Hunt 1994). Vendor characteristics were chosen based on these factors.
Finally, perceived risk has been identified as a key antecedent to relationship commitment in past studies (Sheth and Parvatiyar 1995). In this research, the perceived risk is referred to as the overall perceived security of transactions in an online environment and it is not specifically related to a single vendor. The notion of perceived risk as a key antecedent to consumer behavior has been established in the past and may be a primary factor influencing the conversion of browsers to buyers (Bauer 1960). Connected with this issue is consumers’ concern for privacy, e.g., Bloom, Milne and Adler (1994). This research suggests that consumers may not be willing to give out information on the Internet since they may be afraid that their private information may be sold to someone else and this may prevent them from engaging in e-commerce. We posit that perceived security of transactions and concern for privacy are two other antecedents to electronic exchanges.
The model of antecedents to electronic exchange is presented in Figure I. The model shows the likelihood of electronic exchange as the focal construct of interest influenced by consumer and vendor characteristics, concern for privacy and perceived security of transactions. For the purpose of this study, we define electronic exchange as past purchasing behavior measured in two ways: (1) number of occasions when a WWW user makes an electronic purchase and (2) the total amount spent online in the last six months.(3) The data in this study was analyzed at the individual level. In the next section, each antecedent presented in the model is described and formal hypotheses are developed.
Figure IFactors Influencing Likelihood of Electronic Exchange
Vendor CharacteristicsWe define a "vendor" as any seller who seeks commercial electronic exchange with an Internet user. This should not be confused with electronic service providers like Netcom or America Online who provide computer time for a fee. A retailer with a home page on the WWW, like JC Penney, which provides users with the opportunity to shop over the computer is therefore classified as a vendor. Consumers evaluate these vendors before they enter into electronic exchanges and therefore, the characteristics of these vendors play an important role in facilitating an exchange. These vendors have to be superior to other vendors in alternative shopping modes in order to be noticed and contacted by consumers. We identify the following vendor characteristics as important in the context of electronic transactions: (1) reliability, (2) convenience in terms of services offered, and (3) the perceived price competitiveness and easy access of information offered by Web vendors in comparison to alternative shopping modes.
Reliability is related to the construct of trust. Trust is defined as, "a willingness to rely on an exchange partner in whom one has confidence," (Moorman et al. 1993, p.82) and as confidence that the other party is reliable, honest, consistent, competent, fair, responsible, helpful and altruistic (Morgan and Hunt 1994). Luedi (1997, p.22) argues that vendors should '..fulfill transactions by reliably and securely supporting the full spectrum of electronic commerce from promotional pricing to secure payment handling." The trust in a vendor is likely to affect the consumers’ perception of vendor’s reliability and is therefore identified as an antecedent of an electronic exchange.
The perceived convenience offered by Web vendors is a significant factor in influencing the decision to purchase at home. Shopping convenience is acknowledged to be the primary motivating factor in consumer decisions to buy at home (Gillett 1976). Studies of catalog and telephone shopping have indicated the role of convenience-orientation as a significant predictor of in-home shopping behavior (Gillett 1976; Reynolds 1974). Shopping convenience includes the time, space and effort saved by a consumer and it includes aspects such as an ease of placing and canceling orders, returns and refunds, timely delivery of orders (Gehrt, Yale and Lawson 1996). H1a: The greater the perceived reliability of Web vendors compared to other vendors, the greater the likelihood of electronic exchange. Price competitiveness of an online vendor in comparison to other online vendors should promote Internet purchases. Previous research suggests that the Internet provides consumers with information that allows for price comparisons (Zellweger 1997). Alba et al (1997) states that the Internet increases price comparisons and intensifies competition among the online vendors who try to attract potential buyers. H1b: The greater the perceived convenience of using Web vendors compared to other vendors, the greater the likelihood of electronic exchange. Finally, the wealth of useful information that is readily provided on the Internet by a vendor is likely to enhance electronic transactions (Zellweger 1997). Zellweger (1997, p. 13) states that buyers become extremely frustrated "especially when pages contain irrelevant information." Luedi (1997 p.22) argues that successful Internet marketers should "attract and retain consumers by providing personalized and compelling content coupled with a sense of community relevant to them." This suggests that buying might be the result of encouraging browsers to repeat visit the site. Consumers might consider richness of information as a vendor-specific characteristic. This information may in itself be a reason to return to that vendor. Further, with technologies like personalization used in conjunction with detailed product information, the switching costs of moving to another vendor are increased after the first positive shopping experience at the vendor. Therefore,it is proposed that: H1c: The greater the perceived price competitiveness of Web vendors compared to other vendors, the greater the likelihood of electronic exchange.
H1d: The greater the perceived usefulness of information of Web vendors compared to other vendors, the greater the likelihood of electronic exchange.
Perceived Security of Transactions and Concern for PrivacyOne of the most important and pressing concerns for businesses on to the Internet deals with the level of security in transactions. Many companies are going online, not because of strategic reasons, but due to strong lobbies that push for such an interface with the outside 'computer' world. Most current commercial Web Pages provide consumers with various options to place orders for the products advertised. These include addresses, toll-free numbers, and in many cases, a provision for sending credit card information. Unscrupulous use of such sensitive information cannot be ruled out. Despite advances in Internet security mechanisms like SHTTP, cryptography, and authentication, customers are still concerned about using an impersonal transaction medium for secure transactions. Online retailers have to make concerted efforts to allay these fears by offering clear guidelines to consumers on their online security and privacy policies, limits of consumer liability in the case of fraudulent transactions, and offer alternate payment mechanisms through toll-free phone numbers, customer representatives, etc. if necessary. Still, perceptions of unsatisfactory security on the Internet is one of the primary reasons hindering online purchasing (Zellweger 1997; Communications of the ACM, April 1999, p.80).
Risk is faced by individuals when a decision, action or behavior leads to different outcomes (Bem 1980). When an individual's action produces social and economic consequences that cannot be estimated with certainty, the individual encounters risk (Zinkhan and Karande 1991). Risk relates to situations or problems (Bem 1980; Dowling 1986), overall product categories or brands (Dowling and Staelin 1994) or persons' attitude to risk (Zinkhan and Karande 1991). In our context, two types of risk are especially relevant: (1) person's overall risk taking propensity and (2) the perceived risk of online transactions. The perceived security of online transactions varies based on the specific payment procedure. For instance, the perceived security of giving credit card information directly over the Web is likely to be different from the risk of setting up a third party account and using that account number in transactions.
Connected with this issue are consumers’ concerns about the use of their private information by organizations when engaging in Internet activities (Business Week, April 5, 1999). Various surveys show that online shoppers are concerned about privacy (Communications of the ACM, April 1999, p. 80). Rohm and Milne (1999) confirmed the finding that privacy is an important issue to the Internet users although other research on direct mail suggests that privacy may not be of such great importance to consumers (Milne and Gordon, 1993). H2: The greater the perceived security of transactions in an online medium, the greater the likelihood of electronic exchange.
Alternative payment procedures might offer convenience, but offer a limitation to customers who are concerned about transmitting personal information, e.g., credit card numbers, online. P3P and other privacy protocols also represent the initial steps in the evolution of technical standards to combat privacy abuse. We refer the reader to Cranor, et al (1999) for a comprehensive analysis of consumer attitudes to online privacy, technological solutions, and the current debate on privacy issues. (Cranor et al, 1999).Given the current recognition of privacy as a major issue in electronic commerce, we propose that consumers propensity to engage in shopping over the internet is lower if they are concerned about privacy of information.
H3: The greater the concern for privacy, the lower the likelihood of electronic exchange.
Consumer CharacteristicsThe relationship marketing literature suggests that consumer characteristics, e g., sociological orientation, plays an important role in a consumers’ propensity to engage in the Internet transactions (Sheth and Parvatiyar 1995). The retailing literature also suggests that consumer characteristics are important indicators of the probability of making purchase decisions on the Internet. In his pioneering study, Stone (1954) suggested that shopping behavior has social-psychological origin and classified shoppers into four types: economic shopper, the personalizing shopper, the ethical shopper and the apathetic shopper. Another typology was identified by Stephenson and Willett (1969) who grouped consumers into recreational, convenience and price oriented shoppers. Two additional categories that is psychosocializing and name-conscious shoppers were added by Moschis (1976). Bellenger and Korgaonkar (1980) suggest that consumers can be classified into recreational and convenience shoppers. They suggest that the recreational shopper is motivated by the social aspects of shopping. Past research suggests that the Internet is less attractive to consumers who value social interactions since it allows for very limited interactions relative to other retail formats such as department stores (Alba et al 1997).(4) Therefore, it is hypothesized that those consumers who are primarily convenience shoppers are more likely to shop online than those that seek social interaction.
H4a: The likelihood of electronic exchange will be greater among convenience shoppers.
H4b: The likelihood of electronic exchange will be lower among shoppers seeking social interaction.
DATAThis study uses secondary data based on an e-mail survey conducted by the Georgia Visualization and Usability Center at Georgia Tech of approximately 5000 respondents. The respondents were invited to participate in the e-mail survey through announcements on Internet related newsgroups (e.g. comp.infosystem, www.announce, comp.internet.net-happenings, etc.), banners randomly rotated though high-exposure sites (e.g. Yahoo, CNN, Excite, Webcrawler, etc.), banners rotated through advertising networks (e.g., DoubleClick), announcements made to the www-surveying mailing list, a list maintained by GVU's WWW User Surveys composed of people interested in the surveys, and announcements made in the popular media, e.g., newspapers, trade magazines, (http://www.gvu.gatech.edu/user_surveys/survey-1998-10/#methodology). A $ 100 cash incentive was given to approximately ten randomly chosen respondents. One of the limitations of this survey is that respondents are not chosen in a random manner. In order to ensure a random sample, it is essential to have a list of all users of the Internet such that respondents may be chosen randomly using probability sampling. Since such a list is not available, a non-probabilistic sampling procedure described above is used. This may result in self-selection bias and reduce our ability to generalize to the population at large. However, most surveys have some element of self-selection bias due to the refusal by certain respondents to participate in a survey. It is acknowledged that the sampling procedure may be a limitation of the current study but also believe that the insights derived from an empirical analysis of the topic outweigh the limitation imposed by the sampling procedure.
The survey, conducted in 1998, involved data collected in questionnaires addressing each topic: vendor characteristics, security of transactions, concern for privacy, customer characteristics and purchasing behavior. The data available in various databases was matched using an ID number given to each consumer. Of the 5000 respondents that responded to the individual questionnaires, only 428 had completed responses to all the topics. In other words, the final sample size after merging data sets was 428. A description of the respondents in the sample in terms of their age, education, income and gender is presented in Table II. The respondents were qualified to include those who are users of the internet for collecting information and browsing. Some of these respondents use the internet for purchasing. Approximately 15% of the respondents hardly ever purchased anything online.
TABLE IIDescriptive Statistics
Variable Percentage (Sample Size=428) Gender
Assessing Discriminant Validity of the Scale: In order to assess the discriminant validity of the entire scale, a factor analysis was conducted and those factors with an eigenvalue greater than 1.0 retained. This analysis resulted in eight factors. Items with loadings greater than .50 were identified and used in naming the factors. The results from this analysis showed that the factors that were identified corresponded with the various constructs that were measured, e.g., vendor characteristics, privacy, security and customer characteristics. The results of this are presented in Table III. A detailed analysis of each of these factors follows.
TABLE IIIFactor Analysis of Scale Items
Vendor Privacy1 Privacy2 Security Privacy3 Privacy4 Customer Privacy 5 Vendor (Q1) 0.64 -0.09 -0.16 -0.07 0.00 0.08 -0.15 -0.14 Vendor (Q2) 0.73 -0.05 -0.03 0.00 0.00 0.10 -0.13 0.00 Vendor (Q3) 0.73 -0.08 0.13 -0.05 0.07 -0.01 0.19 0.11 Vendor (Q4) 0.53 -0.09 -0.13 -0.06 0.02 -0.08 -0.26 0.03 Vendor (Q5) 0.66 0.04 0.12 0.05 -0.19 -0.10 -0.01 0.02 Security (Q6) 0.01 0.14 -0.07 0.85 0.15 -0.03 -0.03 0.04 Security (Q7) -0.08 0.15 0.03 0.85 -0.03 -0.04 0.05 -0.05 Privacy (Q8) -0.13 0.64 -0.02 0.13 -0.09 0.08 0.16 0.07 Privacy (Q9) -0.01 0.82 -0.05 0.02 0.15 -0.09 0.02 -0.14 Privacy (Q10) -0.03 0.74 -0.06 0.10 0.39 0.07 0.03 -0.16 Privacy (Q11) -0.12 0.61 -0.12 0.15 0.19 -0.04 -0.12 0.30 Privacy (Q12) -0.07 -0.07 0.69 0.08 0.17 0.01 -0.16 -0.05 Privacy (Q13) 0.00 -0.08 0.74 -0.04 -0.12 -0.02 -0.03 -0.06 Privacy (Q14) 0.00 0.07 0.59 -0.17 -0.10 -0.15 -0.07 0.43 Privacy (Q15) -0.10 0.34 0.02 0.15 0.72 0.10 0.11 -0.02 Privacy (Q16) 0.01 0.11 -0.04 0.00 0.86 -0.11 -0.04 0.07 Privacy (Q17) 0.08 0.06 0.15 -0.31 -0.24 0.57 -0.19 0.18 Privacy (Q18) -0.05 0.01 -0.09 0.05 0.08 0.85 -0.03 0.09 Privacy (Q19) 0.04 -0.07 -0.07 0.01 0.05 0.25 0.07 0.80 Privacy (Q20) 0.08 -0.18 0.48 -0.05 -0.10 0.31 0.30 -0.32 Social (Q21) -0.06 0.04 -0.12 -0.01 0.00 -0.03 0.79 0.16 Convenience(Q22) 0.25 -0.08 0.06 -0.05 -0.07 0.11 -0.49 0.15
Note: Items with loadings greater than .50 are highlighted
Vendor Characteristics (VENDOR)a. Vendor characteristics were operationalized using a 5-point agree-disagree scale2. To evaluate vendors characteristics, the following scale items were used: (1) vendor perceived reliability - respondents expressed their opinion to the statement ‘World Wide Web vendors are more reliable’, (2) perceived convenience of using Web vendors - respondents expressed their opinion to the two statements ‘(a) It is easier to place orders placed with World Wide Web vendors, and (b) It is easier to contact World Wide Web vendors, (3) price competitiveness - respondents expressed their opinion to the statement ‘World Wide Web vendors offer better prices’, and (4) access to information - - respondents expressed their opinion to the statement ‘World Wide Web vendors offer more useful information about the choices available.
The factor analysis of the scales showed that all vendor characteristics load on one factor with an eigenvalue greater than one. The reliability of the entire scale was 0.69. The vendor characteristics of respondents were averaged to form one score. Since all the characteristics of the vendor related to the vendors’ superiority over others (in regards to reliability, price, information provision or convenience) the construct was named ‘superiority over other vendors’.
Perceived Security of Transactions (SECURITY)b. Theperceived security of transactions was operationalized using a 4-point scale with the following two items: (1) In general, how concerned are you about security on the Internet? and (2) How concerned are you about security in relation to making purchases or banking over the Internet? The correlation between these two items was 0.57.
Concern for Privacy (PRIVACY1-PRIVACY5)c. The privacy scale items used in the survey consisted of 13 items. A factor analysis of the scale items used in the survey indicated that the privacy items loaded on five factors. These five factors emerged which seemed to map onto various aspects of privacy, i.e., use of information, anonymity, perception of direct marketing, privacy laws and control over information. The items which had factor loadings of greater than 0.5 were retained. The reliability coefficients for the first two subscales were .727 and .536. The correlation between items in the last two subscales were .499 and .411.
Customer Characteristics (SOCIAL INTERACTION and CONVENIENCE)d. In order to assess whether customers are motivated by convenience or the social interaction associated with shopping, two questions were posed. The first question asked the respondents whether they preferred dealing with people during shopping (One of the reasons I have not shopped on the Web is that I prefer to deal with people). The second question asked whether convenience affected their choice of the shopping mode (One of the reasons I shop on the Web is convenience). Both were dummy variables that were coded 1 if the response was a ‘yes’ and 0 otherwise.
Likelihood of Electronic Exchange (PUR)Likelihood of electronic exchange was based on past purchasing behavior on the Web and measured in two ways (1) as number of electronic purchases and (2) the total amount spent online in the last six months. The means and standard deviations of the scale items along with detailed descriptions are presented in Table IV.
Table IVVariables and Scale Items
a the complete questionnaire available at: http://www.gvu.gatech.edu/user_surveys/survey-1998-10/graphs/graphs.html#shopping.
Factors and Scale Items Means and Standard Deviations Reliability Vendor Characteristics (a) (5-point scale ranging from strongly disagree to strongly agree) 1. World Wide Web vendors are more reliable 3.380 (1.037) .690 2. World Wide Web vendors offer better prices 3.666 (0.986) 3. World Wide Web vendors offer more useful information about the choices available 2.764 (0.993) 4. It is easier to cancel orders placed with World Wide Web vendors. 3.387 (0.965) 5. It is easier to contact World Wide Web vendors 3.357 (1.137) Perceived Security of Transactions and Concern for Privacy (b) Perceived Security of Transactions (SECURITY) 6. In general, how concerned are you about security on the Internet? 7. How concerned are you about security in relation to making purchases or banking over the internet? 3.030 (1.016) .567 (4-point scale ranging from not at all concerned to very concerned) 3.156 (0.959) Concern for Privacy (c) (5-point scale ranging from disagree strongly to agree strongly) USE OF INFORMATION (PRIVACY1) 8. Web sites need information about their users to market their site to advertisers 2.631 (1.286) .727 9. Content providers have the right to resell information about its users to other companies 4.332 (1.055) 10. A user ought to have complete control over which sites get what demographic information 4.191 (1.145) 11. Third party advertising agencies should be able to compile my usage behavior across different web sites for direct marketing 4.393 (1.015) ANONYMITY (PRIVACY2) 12. I value being able to visit sites on the Internet in an anonymous manner 2.049 (1.207) 13. I ought to be able to visit sites on the internet in an anonymous manner 1.489 (0.782) .536 14. I would prefer internet payment systems that are anonymous to those that are user-identified 1.979 (1.182) DIRECT MARKETING (PRIVACY3) 15. I like receiving mass postal mailings that were specifically targeted to my demographics 3.990 (1.197) .499 16. I like receiving mass electronic mailings 4.666 (0.699) PRIVACY LAWS (PRIVACY4) 17. There should be new laws to protect privacy on the Internet 2.324 (1.370) .411 18. There should be laws to protect children's privacy 2.462 (1.466) CONTROL OVER INFORMATION (PRIVACY5) 19. I ought to be able to communicate over the Internet without people being able to read the content. 2.806 (1.707 .090 20. I support the establishment of key escrow encryption where a trusted party keeps a key that can read encrypted messages 1.224 (0.578) Customer Characteristics (d) 21. SOCIAL INTERACTION: One of the reasons I have not shopped on the Web is that I prefer to deal with people (Yes=1/No=0) .105 (.307) - 22. CONVENIENCE: One of the reasons I shop on the Web is convenience (Yes=1/No=0) (5-point scale ranging from very uncomfortable to very comfortable) .825 (.380) -
b and c the complete questionnaire available at: http://www.gvu.gatech.edu/user_surveys/survey-1998-10/graphs/graphs.html#privacy
d the complete questionnaire available at: http://www.gvu.gatech.edu/user_surveys/survey-1998-10/graphs/graphs.html#general
1 reliability is assessed using Cronbach’s Alpha when more than 2 scale items are present; otherwise the numbers reported refer to a correlation coefficient.
Table IV (cont..)
Variables and Scale Items
Variables and Scale Items Number of Scale Items Mean S.D. Purchase Behavior (PUR 1) (d)
23. On average, how often do you make online purchases from Web-based vendors?
(5-point scale ranging from hardly ever to at least once a day)
1 2.593 0.983 Purchase Behavior (PUR 2) (d)
24. What is the total amount you spent on purchases through vendors on the World Wide Web during the past six months?
(4-point scale ranging from $0 to $500 or more)
1 2.960 1.091
d the complete questionnaire available at: http://www.gvu.gatech.edu/user_surveys/survey-1998-10/graphs/graphs.html#shopping
note: the pearson correlation between PUR1 and PUR2 is .566
MODEL DEVELOPMENTThe constructs described above were measured using various scale items that were reduced to various dimensions using factor analysis. The variables were averaged for each factor and the averages were used as input for each construct. We use multiple regression analysis to estimate the model. The model to be tested is of the following form:
Y(1,2) = b0 +b1 X1 + b2 X2 + b3X3 + b4X4 + e
Y(1,2) = likelihood of electronic exchange
X1 = vendor characteristics (operationalized as (1) number of electronic purchases and (2) the total amount spent online in the last six months) X2 = perceived security of transactions (measured on reliability, convenience; price competitiveness and access to information)
X3 = concern for privacy
X4 = customer characteristics (convenience and social interaction)
e = residual term
Two models for each indicator of likelihood of electronic exchange were estimated. The results with frequency of Web shopping (Model 1) the total amount spent online in the last six months (Model 2) as the dependent variables are presented in Table V. The explanatory power of the models, as indicated by adjusted R2 for Models 1 and 2 is 10% and 13% respectively.
Model 1 shows that perceived superiority of Web vendors positively affects frequency of consumer shopping on the Internet (b1 = 0.22, p<0.01). Thus, reliability of a vendor, convenience of placing orders and contacting vendors, price competitiveness and access to information, have a positive influence on the number of purchases on the Internet. However, Model 2 shows that these characteristics do not influence the amount of money spent on the Internet.
The results of the study show that social interaction as a shopping motivation deters consumers from shopping frequently (b1 = 0.48, p<0.01) and from spending money on the Internet (b1 = 0.64, p<0.01).(6) These consumers are likely to treat the shopping experience as a social experience. On the other hand, consumers who value convenience tend to use the Internet to purchase goods frequently (b1 = 0.55, p<0.01) and they seem to spend more money (b1 = 0.55, p<0.01) in the electronic transactions. Thus, the need for social interaction negatively affects the propensity to engage in Internet transactions, and convenience orientation positively affects the frequency and the size of purchases on the Internet.(5)
Interestingly, perceived security of transactions had a negative marginal effect on the frequency of shopping on the Internet (Model 1) which means that consumers seem less concerned about the security of electronic exchanges (p<0.1). The analysis of Model 2 shows that consumers are concerned about some aspects of information privacy. Consumers who purchased more on the Internet seemed to be more concerned about the creation of laws protecting privacy on the Internet (b1 = 0.13, p<0.01). Another dimension of privacy, i.e., consumers’ beliefs that marketers need information about them for marketing purposes, had a marginal negative effect on the amount of money spend on the Internet (p<0.1). Thus, it seems that consumers who spend more on the Internet have a tendency to believe that marketers do not need more information about them to market their products.
Table VRegression Statistics
Variable Model 1 Model 2 INTERCEPT 0.724 (.459) 2.066* (.526) VENDOR .219* (.069) .097 (.079) SECURITY -.089# (.055) -.024 (.063) PRIVACY1 .018 (.063) -.130# (.072) PRIVACY2 .003 (.061) .064 (.070) PRIVACY3 .067 (.060) -.064 (.070) PRIVACY4 .037 (.041) .126* (.047) PRIVACY5 .066 (.081) .040 (.093) SOCIAL INTERACTION .476* (.150) .639* (.172) CONVENIENCE .552* (.123) .552* (.143) Dependent Variable PUR 1 * PUR 2 * Sample Size 427 423 R2 .12 .15 Adjusted R2 .10 .13 * significant at p < .01
** significant at p < .05
# significant at p < .10
note: total sample size was 428;
Additional AnalysesTo assess the relative contribution of each of the four variables, a stepwise regression analysis was conducted on each of the dependent measures. The incremental R2 for the model with each additional variable included provides an assessment of the relative contribution of each variable. According to this, in Model 1, convenience as a shopping motive accounted for 7% of the variance explained, vendor characteristics accounted for 3%, social interaction accounted for 2% and security for 1% of the variance explained. In the case of Model 2, with the total amount spent online as a dependent measure, convenience as a shopping motive accounted for 7% of the variance explained, recreation accounted for 4%, and privacy concerns such as use of information and privacy laws accounted for 3% and 1% of the variance explained respectively. Thus, across both models, with both frequency of shopping and amount spent online as dependent measures, it appears that customer characteristics dominate all other variables in terms of variance explained.
Logistic Regression TABLE VI
Variable Model 1 Model 2 Dependent Variable Convenience Social Interaction * * Intercept 4.356* (1.525) -3.485* (1.001) Gender -1.450** (.619) .235 .737** (.310) 2.090 Education .039 (.163) 1.039 .524* (.130) 1.689 Income -.102 (.098) 1.107 0.203* (.078) 1.224 Age -.030 (.079) .970 .038 (.067) 1.039 -2 Log L 9.320** 32.229* Sample Size 371 371 One of the factors that emerges significant in explaining the frequency of shopping on the Web is vendor characteristics. Vendor characteristics is a summated scale including reliability of a vendor, convenience of placing orders and contacting vendors, price competitiveness and access to information. One important issue is the relative role of each of these aspects in influencing frequency of shopping. In order to examine this issue, we estimate five regressions with frequency of shopping as the dependent measure and each vendor characteristic, e.g., reliability of a vendor, price competitiveness, access to information, ease of canceling orders and contacting vendors. Each of these is incorporated as an independent variable in a separate regression in the presence of other independent variables such as security, privacy, social interaction and convenience. The results of these regressions suggest that price competitiveness and ease of canceling orders were both significant at the 1% level while reliability and access to information were significant at the 10% level. The only variable that did not emerge significant was the ease of contacting vendors on the Web. In other words, all but one of the vendor characteristics was significant in explaining the frequency of shopping on the Web. It is possible that the ease of contacting vendors on the Web is similar across all vendors in the online environment. a numbers in italics refer to odds ratios
* p <.01
** p< .05
The impact of customer characteristics suggests the possibility that there are distinct segments of consumers who place emphasis on convenience versus the social aspects of shopping. In order to investigate the characteristics of these segments, a logistic regression was estimated. The results of these estimates are presented in Table VI. As shown in the table, Model 1 and 2 included gender, age, household income and education as independent variables. These variables were included primarily on the basis of previous research which suggests that demographic variables that have an impact on internet usage patterns (e.g., Hoffman, Kalsbeek and Novak 1996; Burke 1997) and on the basis of the previous research relating demographic variables to shopping motivations in traditional retailing contexts (Bellenger and Kargoankar 1980). Although some of these demographic differences appear to be declining as internet usage is becoming more mainstream, it is still interesting to examine whether these variables might account for differences in motivations. In Model 1, the dependent variable was whether or not convenience was mentioned as a reason for shopping or considering shopping on the Internet. Model 2 included the same independent variables but the dependent variable was whether or not preference for dealing with people was mentioned as a reason for not shopping or not considering shopping on the Internet.
The results of the logistic regression of these two models indicate that there is a significant relationship between gender and social interaction as a shopping motive (b1 =-1.450, p<0.05). Gender was a dummy variable with females being coded as a zero and males as a one. The result suggests that males might be less motivated than females by social interaction, a finding that confirms what is known about gender differences even in the bricks-and-mortar setting. For example, the study by Bellenger and Kargoankar (1980) suggests that the typical recreational shopper tends to be a female head-of-household. The results from Model 2 suggest that convenience as a shopping motivation is related to gender, education and income. The positive coefficient on gender suggests that males are more convenience-oriented than females. The positive coefficient on education (b1 =.524, p<0.01) and income (b1 =.203, p<0.01) suggests that convenience is a greater motivator at higher levels of education and income. It is likely that these variables are proxies for time-poverty. More research on profiling segments of consumers with different shopping motives is necessary in order to promote online shopping as an alternative shopping medium to various segments of consumers.
Implications of the Study and Future ResearchThis study has both theoretical and managerial implications. The study examines a model of electronic exchange based on a theoretical framework proposed by Bagozzi (1975). The factors examined in this study include vendor characteristics, security and privacy and customer characteristics. By examining a sample of internet users, we are able to examine the impact of these factors in converting existing users from browsers to buyers. The study supports previous theoretical propositions that vendors should be reliable, offer competitive prices, provide useful information on the Internet and easy to conduct services (Zellweger, 1997; Alba et al., 1997; Luedi, 1997; Palmer, 1997). The empirical findings suggest that perceived vendor characteristics, particularly price competitiveness and ease of canceling orders, affect the frequency of purchases on the Internet. The result regarding vendor characteristics provides empirical support for the previous research regarding reliability of an exchange partner (Morgan and Hunt 1994). It highlights the role of trustworthiness of an exchange partner (Moorman et al 1993), thereby extending our knowledge of the importance of vendor characteristics from the traditional to the internet domain.
Interestingly, this study shows an average consumer is not as concerned about the security of electronic exchanges or privacy issues. The concern over security has decreased over the years particularly with developments in internet payment systems that ensure confidentiality. This finding supports previous research, that focused on privacy issues in the context of direct mail, which indicated that consumers are not concerned about privacy issues (Milne and Gordon, 1993). However, this study does show that consumers who purchase frequently on the Internet are interested in creation of new laws protecting privacy on the Internet. They also do not believe that marketers need more information about them to successfully market products on the Internet. This has important implications both from a theoretical and a managerial standpoint. From the theoretical standpoint, this result serves to resolve the debate regarding the level of consumer concern regarding privacy issues (Milne and Gordon (1993) versus Rohm and Milne (1999)). From the managerial standpoint, marketers should be sensitive about privacy issues particularly the perception that information is not necessary to enhance marketing of products on the internet.
Finally, this study shows that consumers who are primarily motivated by convenience are more likely to make purchases online. Those who value social interactions are less interested in the Internet use for shopping and thus shop less frequently on the Internet and spend less money on e-commerce. This in itself is valuable because it suggests that retailers have to fine tune their offerings, and provide very specific solutions to each segment of customers at the aggregate level, and individual customers if possible. With advances in personalization and customization, the notion of one-to-one marketing can be realized. These findings should be valuable for marketers for the purpose of segmentation and targeting their prospective buyers.
The result regarding the importance of convenience as a motivator of internet shopping is interesting from the perspective of enhancing our understanding of shopping motivations in the internet context. In addition, it also provides a basis for marketers to differentiate themselves from competitors. The role of social interaction as a deterrent of internet shopping is also an important result. While it may not be possible to mimic all the features of a physical store and the ability of the physical world to provide unique shopping experiences (Rao 1999), marketers should take advantage of developments such as discussion groups, chat forums etc. in order to enhance the level of social interaction at various sites. This is being achieved by some internet retailers who provide links to various discussion forums, thereby tying in aspects of social interaction with visits to their Web sites. These developments may go a long way in enhancing the level of social interaction during internet shopping. Future research should focus on creating a typology of internet shoppers and linking various demographic and psychographic variables with types of internet shoppers. Such a typology along with a profile of the various internet shopper types may provide actionable guidelines to marketers interested in targeting various segments of consumers.
It is important to note the limitations of this study. This research is based on an e-mail survey and therefore a selection bias might have affected our findings. Only online respondents participated in the study. Therefore, the self-selection bias may limit the generalizability of the findings. Extensions of this study in other settings and using other data collection methods should provide additional evidence to support and expand our findings. An interesting future research could focus on the empirical comparison of factors that affect online shopping versus more traditional formats of retailing. The results we present here are an important step forward in enhancing our understanding of the future of electronic commerce.
FOOTNOTES(1) Note that although rough estimates suggest that, while sales ordered and paid for online were in the region of $11.0 billion, sales to consumers that were ordered online, but paid for off-line were more than $15 billion. Further, the value of off-line orders influenced by the Internet was approximately $51 billion (cite: Maryann Jones Thompson (1999), "Only Half of Net Purchases are Paid for Online," The Industry Standard, March 1, 1999. URL: http://www.thestandard.com/metrics).
(2) This is however, dependent upon the product category. Amazon.Com offers a vast selection of books. In other cases, e.g., online grocery stores, there are limitations in the variety of products that can be offered.
(3) Note that the emphasis here is on commercial transactions implying monetary exchange. This definition precludes those transactions where information is collected online but the transaction is completed offline.
(4) We acknowledge that online communities, discussion groups, and review boards (like the one on Amazon.com), are mechanisms that are being employed by marketers to encourage social interaction among consumers. These allow consumers to interact as they scan for information, make buying decisions, or report feedback on their purchases. They definitely provide incentives to customers to visit sites, and, if implemented properly, might do a better job of turning browsers into shoppers.
(5) Additional analyses were conducted to examine which of the vendor characteristics such as reliability of a vendor, convenience of placing orders and contacting vendors, price competitiveness and access to information had a greater influence. Results of these are discussed in the next section.
(6) Please note that social interaction was reverse-coded such that a positive coefficient indicates a lower propensity to shop online.
REFERENCESAlba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A. & Wood, S. (1997). Interactive home shopping: Consumer, retailer and manufactuere incentives to participate in electronic marketplaces. Journal of Marketing, 61 (3), 38-54.
Bagozzi, R. P. (1974). Marketing as an organized behavioral system of exchange. Journal of Marketing, 38 (October), 77-81.
Bagozzi, R. P. (1975). Marketing as exchange. Journal of Marketing, 39 (October), 32-29.
Bauer, R. A. (1960). Consumer behavior as risk taking. In D. F. Cox (Ed.), Risk taking and information handling in consumer behavior (pp. 22-23). Cambridge, MA: Harvard University Press.
Bellenger, D. N., & Kargaonkar, P. K. (1980). Profiling the recreational shopper. Journal of Retailing, 56 (3), 77-82.
Bem, D. (1980). The concept of risk in the study of human behavior. In J. Dowie & P. Lefrere II (Eds.), Risk and Chance (pp. 1-15). Milton Keynes, England: The Open University Press.
Berthon, P., Pitt, L. F., & Watson, R. T. (1996). The World Wide Web as an advertising medium. Journal of Advertising Research, 36 (January-February), 43-54.
Bloom, P. N., Milne, G. R., & Adler, R. (1994). Avoiding misuse of new information technologies: Legal and societal considerations. Journal of Marketing, 58 (January), 98-110.
Burke, R. R. (1997). Do you see what I see? The future of virtual shopping. Journal of the Academy of Marketing Science, 25 (4), 352-360.
Business Week (1999, April 5). The backlash from e-snooping. i3623, p. 130
Communications of the ACM (1999, April). Building consumer trust online, 42 (4), p. 80.
Cranor, L. F., Reagle, J., & Ackerman, M. S. (1999). Beyond concern: Understanding net users' attitudes about online privacy. AT&T Labs-Research Technical Report TR 99.4.4, Available: http://www.research.att.com/projects/privacystudy/.
Dowling, G. R. (1986). Perceived risk: The concept and its measurement. Psychology and Marketing, 3 (3), 193-210.
Dowling, G. R., & Staelin, R. (1994). A model of perceived risk and risk-handling activity. Journal of Consumer Research, 21 (June), 119-134.
Dwyer, R. F., Schurr, P. H., & Oh, S. (1987). Developing buyer-seller relationships. Journal of Marketing, 51 (April), 11-27.
Gehrt, K. C., Yale, L. J., & Lawson, D. A. (1996). The convenience of catalog shopping: Is there more to it than time? Journal of Direct Marketing, 10 (4), 19-28.
Gillett, P. L. (1976). In-home shoppers: An overview. Journal of Marketing, 40 (October), 81-88.
Hoque, A. Y., & Lohse, G. (1999). An information search cost perspective for designing interfaces for electronic commerce. Journal of Marketing Research, 36 (August), 387-394.
Hoffman, D., & Novak, T. P. (1996). Marketing in hypermedia computer- mediated environments: Conceptual foundations. Journal of Marketing, 60 (3), 50-68.
Novak, T. P., & Chatterjee, P. (1996). Commercial scenarios for the Web: Opportunities and challenges. Journal of Computer-Mediated Communication [Online], 1 (3).
Hoffman, D.L., Kalsbeek, W. D., & Novak, T. P. (1996). Internet and Web use in the United States: Baselines for commercial development. Special Section on "Internet in the Home," Communications of the ACM, 39 (December), 36-46. Available: http://ecommerce.vanderbilt.edu/baseline/internet.demos.july9.1996.html
Kate, N. T. (1998). Kaleidoscope: Surfing for travel. American Demographics, 20 (2), 36-37.
Louis Harris and Associates Inc. Equifax/Harris Consumer Privacy Survey (1996).
Lynch, J., & Ariely, D. (1998). Interactive home shopping: Effects of search cost for price and quantity information on consumer price sensitivity, satisfaction with merchandise selected, and retention. Working paper, Marketing Department, Duke University.
Luedi, A. F. (1997). Personalize or perish. Electronic Markets, 7 (3), 22-25.
Milne, G. R., & Gordon, M. E. (1993). Direct mail privacy-efficiency trade-offs within an implied social conract network. Journal of Public Policy and Marketing, 12(2), 206-215.
Moorman, C., Deshpande, R., & Zaltman, G. (1993). Factors affecting trust in marketing relationships. Journal of Marketing, 57 (January), 81-101.
Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58 (July), 20-38.
Moschis, G. P. (1976). Shopping orientations and consumer uses of information. Journal of Retailing, 52 (Summer), 61.
National Retail Federation (1999). 1998 holiday sales data. Available: http://www.nrf.com/hot/holiday/dec98/ [5/15/1999].
Palmer, J. W., (1997). Retailing on the WWW: The use of the electronic product datalogs. Electronic Markets, 7 (3), 6-9.
Reynolds, F. D. (1974). An analysis of catalog buying behavior. Journal of Marketing, 38 (3), 47-51.
Rao, B. (1999). The Internet and the revolution in distribution: A cross-industry examination and synthesis. Technology in Society, 21, (3).
Rohm, A. J., & Milne, G. R. (1999). Consumers’ privacy concerns about direct marketers’ use of personal medical information. In J. F. Hair, Jr. (Ed.), Proceedings of the 1999 Association for Health Care Research Conference (pp. 27-37). Breckenridge, CO.
Sheth, J., N., & Parvatiyar, A. (1995). Relationship marketing in consumer markets: Antecedents and consequences. Journal of the Academy of Marketing Science, 23 (Fall), 255-271.
Stephenson, R. P., & Willett, R. P. (1969). Analysis of consumer's retail patronage strategies. In P. R. McDonald (Ed.)., Marketing involvement in society and economy (pp. 316-322). Chicago: American Marketing Association.
Stone, G. P. (1954). City and urban identification: Observation on the social psychology of city life. American Journal of Sociology, 60 (July), 36-45.
Thompson, M. J. (1999). Only half of net purchases are paid for online. The Industry Standard, March 1, 1999. Available: http://www.thestandard.com/metrics.
US Department of Commerce (1999). The emerging digital economy II. Available: http://www.ecommerce.gov/ede/.
Zellweger, P. (1997). Web-based sales: Defining the cognitive buyer. Electronic Markets, 7 (3), 10-16.
Zinkhan, G. M., & Karande, K. W. (1991). Cultural and gender differences in risk-taking behavior among American and Spanish decision makers. Journal of
Social Psychology, 131 (5), 741-742.
Zwass, V. (1999). Structure and macro-level impacts Of electronic commerce: From technological infrastructure to electronic marketplaces. Available: http://www.mhhe.com/business/mis/zwass/ecpaper.html.
Acknowledgments:The authors would like to thank the Georgia Tech Research Corporation and the Graphic, Visualization, & Usability Center for making the data available for academic research.
About the AuthorsVanitha Swaminathan is Assistant Professor of Marketing at the University of Massachusetts, Amherst. Professor Swaminathan's research interests focus on branding strategy and electronic commerce. Professor Swaminathan has published in the Journal of Marketing Research, Advances in Consumer Research and in the proceedings of the American Marketing Association, Association of Management and Relationship Marketing conferences.
Address: Department of Marketing, Isenberg School of Management, University of Massachusetts-Amherst, MA 01003. Tel: (413) 545-5665.
Elzbieta Lepkowska-White is an Assistant Professor at Skidmore College. She received her Ph.D. from the University of Massachusetts. Her other areas of interest include international advertising and public policy issues. She is a member of the American Marketing Association and American Academy of Advertising.
Address: Department of Marketing, Department of Management and Business, Skidmore College, 815 North Broadway, Saratoga Springs, NY 12866. Tel: (518) 580-5113.
Bharat Rao is an Assistant Professor of Management at the Institute for Technology and Enterprise, Polytechnic University, in New York City. He leads several strategic research initiatives for the Institute for Technology and Enterprise in the areas of electronic retailing, supply chain management, strategic alliances and new product development. He earned a Ph.D. in Marketing and Strategic Management from The University of Georgia, and was a post-doctoral Research Associate at Harvard Business School. His research has been published in the International Journal of Electronic Markets, Technology in Society, International Business Review, Journal of the Academy of Marketing Science, and in conference proceedings of the IEEE Engineering Management Society, American Marketing Association, Association of Management and TIMS-ORSA. He is also the author of several business case studies, in both paper and digital formats, published by Harvard Business School and the Institute for Technology and Enterprise.
Address: Department of Management, Institute for Technology and Enterprise, Polytechnic University, 55 Broad Street Suite 13B, New York, NY 11201. Tel: 212-547-7030 Ext. 205.