Consumer Trust towards an Unfamiliar Web Merchant : Role of third parties. Inès Chouk, Thema research center (UMR CNRS 7536), France

Transcription

1 Consumer Trust towards an Unfamiliar Web Merchant : Role of third parties TRACK : New Technologies and E-Marketing Inès Chouk, Thema research center (UMR CNRS 7536), France Jean Perrien, Université du Québec à Montréal, Canada Jacques Nantel, HEC Montréal, Canada Abstract Our research intends to provide a better understanding of consumer online purchasing behaviour. It explores mechanisms that convey the trustworthiness of a Web merchant with whom consumers have no prior experience. The core idea underling this paper is that signals provided by third parties could have a crucial role in initial trust development. The focus of this study is consumers response to three types of potential signals: trust seals, testimonies on the web site and partnership with a known web merchant. Key words : consumer trust, third parties, web merchant.

2 Introduction Lack of trust has been identified as the major barrier to the growth of electronic commerce (Jarvenpaa and Tractinsky, 1999; Gefen, 2000; Mcknight and al., 2002; Yoon, 2002, Stewart, 2003, Pavlou, 2003, Bart and al., 2005). Therefore, if trust is a sine qua non condition of e- commerce development, then understanding its determinants should be a primary concern for web merchants. Building trust is essential for the success of electronic vendors especially less known ones. Our research intends to provide a better understanding of trust determinants within the context of business to consumer e-commerce. More specifically, it will focus on the role of third parties in trust building. Indeed, while the majority of recent researches were interested in the role of web merchant characteristics (e.g., company awareness (Yoon, 2002)) and web site characteristics (e.g., site quality (Mcknight and al., 2002)), little attention has been given to the role of third parties (expert organisms, past buyers, partners). These actors could play, however, a key role in trust development mainly when the consumer is not familiar with the web merchant. Fundamental to our analysis is the use of the signaling theory. For Kirmani and Rao (2000), signaling could be an effective strategy under the condition of incomplete information. Indeed, when pre-purchase information is relatively scarce, consumers look for signals to make inferences about the vendor s trustworthiness. Hence, signaling theory could be relevant to enhance our understanding of trust building mechanisms. Different signals provided by third parties could enhance consumer trust. As shown in table 1, in this research, we will focus particularly on three potential signals of the web merchant trustworthiness: trust seals (awarded by a third party organisation), satisfied past buyers testimonies (selected by the web merchant and displayed on the web site) and partnership with a known web merchant. Signals Trust seal Past buyers testimonies reported by the vendor on the web site Logo s partner Third parties Third party organisations Past shoppers A known web merchant Table 1 : Potential signals of a web merchant trustworthiness Precisely, through this research we intend to answer the following key research questions: Do third parties have an effective role in building consumer trust towards an unfamiliar web merchant? Does the consumer s trust have a significant effect on behavioural intentions (the intention to visit the web site in the future and the intention to recommend it to others)?

3 The remainder of this paper is divided into four sections. Section 1 proposes, first, a short literature overview of the trust conceptualisation in marketing. Next, it depicts the theoretical foundations of the research. Section 2 provides the conceptual development and the underling hypothesis. Section 3 depicts the methodology and presents the results. In section 4, results of the research are discussed. I. Theoretical background I.1. Trust conceptualisation in marketing I.1.1. Literature overview In marketing literature, trust has been defined in many different ways: as a presumption, an expectation, a belief, a willingness (a behavioural intention) and as a behaviour. With such a diversity of definitions, it becomes important to clarify the conceptualisation of trust. One suggestion is to identify a definitions taxonomy : Trust as a psychological state This stream of researches regards trust as a psychological state that precedes behavioural intentions and behaviours. Such a state can be a presumption (e.g., Gurviez and Korchia, 2002; Bidault and Jarillo, 1995), an expectation (e.g., Sirdeshmukh, Signh and Sabol, 2002; Perrien Graf, Ricard and Landry, 1999) or a belief (e.g., Anderson and Narus, 1990; Ganesan,1994). In this class of definitions, conative dimensions are disregarded. Indeed, according to Morgan and Hunt (1994), the trust concept should not encompass behavioural intentions (and behaviours). For these authors, intentions and behaviours are considered as trust consequences, they must be then evacuated from trust conceptualisation. Furthermore, an examination of these definitions shows an implicit or explicit reference to the notion of expectations. Even when it is not regarded as an expectation, trust is defined as a presumption or a belief in reference to expectations about the exchange partner (e.g., Gurviez, 2000; Frisou, 2000). Trust as a behavioural intention or a behaviour In this second class of definitions, a conative conception of trust is advocated. Thus, trust was defined as an action (Deutsch, 1962), a willingness to be vulnerable (Mayer, Davis and Schoorman, 1995) or as a willingness to rely on the exchange partner (e.g., Moorman, Deshpande and Zaltman, 1992 ; Chaudhuri and Holbrook, 2001). Most of these researches have underlined the importance of the vulnerability notion. For instance, according to Moorman, Deshpande and Zaltman (1992, p.82), without vulnerability, trust is unnecessary. Since it implies an exposure to a potential harm (Bigley and Pearce, 1998), vulnerability becomes stronger in high risky contexts. Joining the stream of marketing literature that characterizes trust as a psychological state, we define trust towards a web merchant as the web vendor competence and expertise (1), the integrity regarding the information provided by the merchant on his web site (2), the integrity with respect to the data disclosed by consumers on the web site (3) and the interest for the problems that may arise during and after the electronic transaction (4) (Authors).

4 I.2. Theoretical foundation : A signaling approach Researches on economics of information are based on the well recognized assumption of information asymmetry between transacting parties (Akerlof, 1970). The signaling approach, proposed by Spence (1973, 1974), is one solution to overcome this problem. When adopting a signaling strategy, actors can communicate unobservable elements (e.g., product/service quality, vendor trustwothiness) by providing observable ones: signals. We adopt this perspective in our research. As highlighted earlier, the primary objective of our study is to examine the impact of a number of signals on the consumer trust within the context of on-line transactions. Specifically, we will focus on trust in an unfamiliar web merchant (i.e., a merchant with whom the consumer has no prior interaction). In such a situation, the information asymmetry between the buyer (the consumer) and the seller (the web vendor) is important. Indeed, without previous experiences with the vendor, assessment of the web merchant trustworthiness seems to be difficult. Signals are proposed to be a potential solution to resolve the problem of taking a decision under the condition of incomplete information (Boulding and Kirmani, 1993). Within the context of product quality research, a signal was identified as an action that the seller can take to convey information credibly about an unobservable product quality to the buyer (Rao, Qu and Ruekert, 1999, p. 259). Hence, signals could be perceived as actions taken to reveal information regarding an unobservable state (such as the product quality or the web vendor credibility). According to Kirmani and Rao (2000), signaling is appropriated and particularly effective when the consumer is not familiar with the product or with the firm. Given that consumers lack information, they use cognitive shortcuts to make inferences about the vendor (Kirmani and Rao, 2000). These inferences are made from the information provided by the seller. Under these conditions, it is in the interest of the merchant to send prepurchase signals about his trustworthiness. II. Conceptual development This paper develops a cognitive model of trust building arguing that trust can be inferred from signals provided by third parties. In this research, we will focus on three types of signals: third party trust seals, testimonies of past buyers and partnership with a known web merchant. Third party trust seals Marketing researchers have been showing interest in third party endorsement seals for over 40 years (e.g., Taylor, 1958). A literature overview shows that empirical findings are controversial regarding seals effectiveness (Beltramini and Stafford, 1993; Lala, Arnold, Sutton and Guan, 2002). Most authors support, however, the assertion of a positive impact of third party seals on consumer behaviour (i.e., product choice, purchase intention). Because they are delivered by experts and independent institutions, seals of approval may be very effective in communications (Parkinson, 1975). In advertising literature, Dean and Biswas (2001) found that third party organisation endorsements outperform celebrity endorsements for expensive and utilitarian products. Consistent with Friedman and Friedman s research (1979), the authors demonstrate that third party organisation endorsements are especially

5 useful for products with high financial risk. Hence, in a risky context, third party seals can provide assurance for potential buyers. Within an Internet context, when a web merchant is awarded an assurance seal, it implies that the web site follows a number of principles and recommendations stipulated by the third party having delivered the seal. These principles can cover a variety of applications such as: transaction payment security, privacy protection and vendor authentication. Trust seals connote a certain level of credibility that may influence consumers trust (Endeshaw, 2001). Thus, when a third party seal is affixed on a web site, consumers are more likely to trust the Internet store. Previous works have highlighted the importance of trust seals for the e- commerce development (e.g., Kaplan and Nieschwietz, 2003; Cook and Luo, 2003, Kimery and McCord, 2006). Hence, we postulate that the use of trusted third parties increases consumers trust. In line with this suggestion, we postulate that third party trust seals enhance the consumer s trust towards a web vendor. Hypothesis 1 : The presence of a third party trust seal on the web site has a positive impact on consumer s trust towards the unfamiliar web merchant. Past buyers testimonies displayed on the web site In marketing, testimonials were mainly studied within the framework of endorsement literature. Friedman and Friedman (1979) have identified three basic types of endorsers: celebrities, experts and typical consumers. A typical consumer endorser was defined as an ordinary person who is expected to have no special knowledge of the product class endorsed except that acquired by normal use of the product (Friedman and Friedman, 1979, p.63). He may exert an influence on consumers through the identification process (perceived similarity) and through the expertise dimension (previous experience with the product). At the theoretical level, consumers testimonies selected by the web merchant and affixed on the web site appear to be closely related to endorsement of typical consumers. Hence, through this research, we will investigate the effectiveness of satisfied consumers testimonies as a signal of the web merchant trustworthiness. We assume that testimonies of past buyers will increase consumer trust towards an unfamiliar web vendor. However, a number of consumers may do not believe them since they could be manipulated by the web merchant. So, in order to assess the real effectiveness of the use of the testimonial approach as a promotional tool, we propose to test the following assertion: Hypothesis 2 : The presence of testimonies on the web site have a positive impact on consumer s trust towards the unfamiliar web merchant. Partnership with a known web merchant: Stewart (2003) studied the trust transfer between a known web site and an unknown one. The research showed that perceived similarity and perceived business tie between tow company web sites increases trust towards the unknown web merchant. Extending the work of Stewart (2003), we focus on trust between tow partners (one is unknown and the other is known). We postulate that the presence of a known partner s logo on the web site could have an impact on the consumer trust towards the unfamiliar web merchant.

6 Hypothesis 3 : The presence of the logo of a known partner on the web site increases consumer trust towards the unfamiliar web merchant. Behavioural intentions : Previous academic studies have provided evidence of the impact of trust on behavioural intentions (e.g.,yoon, 2002; Mcknight and al., 2002; Stewart, 2003, Pavlou, 2003). While the majority if these researches have studied the purchase intention, we focused on two other types of behavioural intentions : the intention to recommend the web site to others and the intention to visit the web site in the future. On the one hand, since purchase decision is associated with a high perceived risk, consumers try to cope with uncertainty by seeking recommendations from experienced sources (Bansal & Voyer, 2000). On the other, when a web merchant builds trust towards his web site, consumers feel confident and they are more likely to visit the web site in the future. Thus, a long term relationship could be established. Hypothesis 4 : Consumer trust towards un unfamiliar web merchant have a positive impact on the intention to recommend the web site. Hypothesis 5 : Consumer trust towards un unfamiliar web merchant have a positive impact on the intention to visit the web site in the future. Drawn from the literature, the conceptual model could be presented as the following : Trust antecedents Behavioural intentions - Third party seals - Testimonies displayed on the web site - Partnership with a known web merchant Trust towards an unfamiliar web merchant - Intention to recommend the web site - Intention to visit the web site in the future Figure 1 : Consumer trust towards an unfamiliar web merchant : Role of third parties

7 III. Methodology and results III.1 Methodology An experiment using a factorial design was set up to test the research hypothesis. We manipulated the absence or the presence of three signals: trust seal, past buyer s testimonies and partner s logo. For this purpose, a travel web site was created (sejourpascher.com). The use of fictive web site were motivated by two main raisons : 1- to make sure that the consumer is not familiar with the web merchant 2- to eliminate confounding effects from external sources of variance (e.g., web site design, web site perceived quality, navigation, ease of use, brand, ). Eight web sites were presented to a student sample. A series of pre-tests were conducted to check the credibility of 1- the web trust seal, 2- the testimonials and 3- the partner. We have used a total sample of 90 students. Results show that these signals were perceived as credible. Subjects were asked to conduct a transaction on the web site (buying a travel package). They were also instructed to read the testimonials when the latter are present. After finishing the task requested, participants were invited to fill in a questionnaire. Data were collected on students from eleven French universities. Questionnaires of respondents who have not perceived the partner s logo or the trust seal were eliminated. At the end, 304 questionnaires were used for the data analysis (38 subjects were randomly assigned to each web site). III.2. Results 1- Do third parties have a role in trust building? Hypothesis H1, H2 and H3 were tested using analyses of covariance (Table 2). Variables entered as covariate were the following : attitude towards the website, perceived quality of the web site, familiarity with travel web sites, involvement in the web site category, trust propensity and Internet expertise. Results indicate the presence of a trust seal have a positive significant impact on the consumer trust (F = 7,285, p < 0,05). People who are exposed to the web site with a trust seal have a higher level of trust towards the web merchant (M present = 4,29, M absent = 4,07). Hence, H1 is supported. As predicted in H2, there is a significant effect of the presence of testimonials on the consumer trust towards an unfamiliar web merchant (F = 5,447, p < 0,05). Indeed, subjects exposed to a web site with testimonials perceive greater trust (M present = 4,28, M absent = 4,09). Thus H2 is supported. H3 suggested that the presence of a known partner will increase consumers trust towards the web merchant. This hypothesis was not supported (p > 0,05). Consumer trust do not vary even thought a known partner logo is present on the web site. One possible explanation is that partners are not a key element in consumer s evaluation of the web merchant trustworthiness. Indeed, generally, consumers pay a little attention to logos of partners displayed on the main page of the web site.

8 Dependant variable : Trust Source Type III Sum of Squares df Mean Square F Sig. Partner 4,241E ,241E-02,087 s Seal 3, ,566 7,285 s Testimonies 2, ,666 5,447 s Partner * Seal,633 1,633 1,292 n.s Partner * Testimonies 3, ,063 6,257 s Seal * Testimonies,416 1,416,850 n.s Partner * Seal *,322 1,322,658 n.s Testimonies Error 141, , ,955 R 2 = 0,343 * Only effects of manipulated variables are shown in the table. Regarding covariates, results indicate that the attitude toward the web site, the web site quality, the trust propensity and the involvement have a significant impact on trust. Table 2 : Results of the Ancova 2 - Does trust towards the web merchant have an impact on the behavioural intentions? Hypothesis regarding the impact of trust on the behavioural intents were tested using a structural equation model. The chi-square goodness of fit statistic was 232,77 (df = 85, p<0,05). The normed fit index (NFI) was 0,90 and the comparative fit index (CFI) was 0,93. The RMSEA was 0,07 which is above the recommended standard of 0,05 but could be considered as a reasonable error of approximation (Browne and Cudeck, 1993). Overall, the indicators points out an acceptable fit of the model to the data. Results indicate that trust towards a web merchant have a positive significant impact on the behavioural intentions. Consumers who trust the web merchant are more likely to visit the web site in the future (which supports H4). As predicted in H5, trust towards the web merchant is significantly related to the intention to recommend the web site to others. IV. Discussion Role of testimonials One of the key contributions of this research is the role of testimonials. Although previous academic works have studied several determinants of online trust, no research have focused on the role of testimonials on the web site. The study of testimonials could be embedded in works on endorsement. As noted earlier, Friedman and Friedman (1979) have identified three types of endorsers: celebrities, experts, and typical consumers. In this research, testimonials could be viewed as a kind of endorsement of typical consumers. Because Internet is a high risky context, we have expected that the presence of testimonials on the web site will increase consumer trust. This thesis was confirmed by the results of the research. Hence, testimonials could be used by web merchants as a cue of their trustworthiness.

9 Role of trust seals Lala and al. (2002) found that consumers prefer web sites with trust seals to web sites without trust seals. For Head and Hassanein (2002), trust seals are particularly important on the establishment stage of the trust development. Kaplan and Nieschwietz (2003) demonstrated that web assurance services were positively related the consumer trust. In line we these previous works, this research showed that the presence of trust seals has a significant impact on consumer trust. Very few researches have manipulated the presence (versus absence) of a web seal to identify a causal relationship between trust and these seals. Since this research provides evidence for the effectiveness of web trust seals, web merchants are encouraged to use this signal to promote their trustworthiness. It should be noted, however, that organisms delivering web trust seals had to make efforts to enhance consumers familiarity with these seals (Kimery and McCord, 2006). Role of trust Results of the research provide evidence of the importance of trust for the e-commerce development. Consistent with previous works (e.g., Bart and al, 2005; Corbitt and al., 2003), trust is found to be a determinant of behavioural intents. Precisely, consumers who trust the web merchant are more likely to visit the website again and to recommend to it other persons. This means that web vendors are encouraged to think about underling mechanisms of online trust building. Indeed, trust is a corner stone of exchanges on the Internet and a key element to create long-term relationships (Dwyer, Shurr and Oh, 1987). Conclusion This paper has explored mechanisms that convey the trustworthiness of a Web merchant with whom consumers have no prior experience. The primary purpose of this study was to provide a better understanding of the third parties role in initial trust building. The secondary goal was to study the impact of trust on behavioural intentions. Results of the research showed that web trust seals and testimonials displayed on the web site increases consumer trust towards an unfamiliar web merchant. Moreover, the research indicates that trust is a key determinant of the behavioural intentions. Two main contributions could be assigned to this work. First, few marketing researches have examined trust using a third party approach when a consumer is not familiar with the web merchant. To date, most research dealing with trust-building factors has focused on the Web merchant characteristics, the web site characteristics or the consumer characteristics. Second, our research provides an empirical evidence of the role of testimonials and trust seals in initial trust building. This means that they could be used as signals to promote web merchants trustworthiness. Moreover, while a number of recent researches have been interested in the study of web trust seals (Cook and Luo, 2003; Kimery and McCord, 2006), the role of testimonials remain largely unstudied. To sum up, the present work could be viewed as an initial step toward understanding the role that third parties have in the initial development of trust. It identifies ways signaling theory could be relevant to enhancing our understanding of mechanisms underling the initial trustgeneration process. Further research could enlarge upon our theoretical framework. In particular, research in the field of psychology on first-impressions theory could offer new and relevant insights.

10 Key References Akerlof, G. (1970). The Market for Lemons : Quality and the Market Mechanism. Quarterly Journal of Economics, 84, Akshay, R.R., Lu, Qu., & Ruekert, R.W. (1999). Signaling Unobservable Product Quality Through a Brand Ally. Journal of Consumer Research, 36, Boulding,W., & Kirmani, A. (1993). A Consumer-Side Experimental Examination of Signaling Theory : Do Consumers Perceive Warranties as Signals of Quality?. Journal of Consumer Research, 20, Friedman, H.H., & Friedman, L. (1979). Endorser Effectiveness by Product Type, Journal of Advertising Research, 19, Gefen, D. (2000). E-Commerce : The Role of Familiarity and Trust. The International Journal of Management Science, 28, Kirmani, A., & Akshay, R.R. (2000). No Pain, No Gain : A Critical Review of the Literature on Signaling Unobservable Product Quality. Journal of Marketing, 64, Morgan, R., & Hunt S. (1994). The Commitment-Trust Theory of Relationship Marketing. Journal of Marketing, 58 (July), Pavlou, P.A. (2003). Consumer Acceptance of Electronic Commerce : Integrating Trust and Risk with the Technology Acceptance Model. International Journal of Electronic Commerce, 7 (3), Yoon, S.J. (2002). The Antecedents and Consequences of Trust in On-line Purchase Decisions. Journal of Interactive Marketing, 12 (2),