The Internet has notoriously democratised direct access to businesses, putting them only a few mouse-clicks away from consumers. However, research indicates that consumers' lack of trust still constitutes a major psychological barrier to the adoption of new forms of online services. It is therefore imperative to identify factors likely to affect a consumer's perception of an online vendor's trustworthiness. Only then can methods be derived that elicit consumer trust requirements and thereby inform the design of the e-commerce user interface.
The emergence of trust assumes situations of risk, uncertainty and vulnerability. Trust-shaping factors idiosyncratic to the electronically-mediated buyer-vendor relationship have been synthesised in the following model of trust for e-commerce (MoTEC), based on Egger (2000). Qualitatively distinct clusters of variables are represented in bold, whereas individual model components are emphasised in italics.
Pre-interactional Filters: First of all, it is noteworthy that individuals differ as to their General Propensity to trust. A second important factor is people's foreknowledge and expectations with respect to a certain industry or company. The latter can be due to Reputation, i.e. the strength of a company's brand name, previous interactions on- and/or off-line, or reports from trusted third parties (hence, Transference of trust).
Interface Properties: The development of trust is strongly affected by one's first impression of a commercial system. Thus, graphic design and layout are encapsulated in the Appeal component. Overview refers to the extent to which the site's commercial offerings and resources are made explicit by organising its content in a manner relevant to the end user. The Usability component refers to the system's reliability, ease-of-use and familiarity in terms of domain model, classification schemes and terminology.
Informational Content: Information about Products & Services should be complete, relevant and structured in a way that reduces user costs. The Company component reflects the need to communicate the merchant's history, values and commitment. Security refers to the completeness and the understandability of information about financial risk and guarantees. Privacy describes the vendor's openness with respect to its privacy policy. The commercial party's competence can be assessed by comparing the first two components, whereas the relationship's inherent risk can be evaluated in the light of the last two components.
Relationship Management: The first type of trust to take place is conversion trust, where users gain enough trust to engage in a commercial relationship with an online merchant. Whether that trust will be honoured in the long term will depend on the follow-up to the initial transaction, e.g. post-purchase communication and customer service. Communication thus reflects the facilitating effect of frequent and personalised vendor-buyer interactions on trust maintenance.
A case study was conducted to examine how well the MoTEC model was suited for the evaluation of an existing site's trustworthiness and the elicitation of trust requirements for redesign. EuroClix is a Dutch membership-based permissive email marketing system that was launched in September 1999. Through emails and the EuroClix web site, members receive offers tailored to their personal profile. In return for reading and voluntarily acting on this information, members receive points that can be exchanged for money. The EuroClix service is particularly interesting as a case study since its brand name was unknown and its service unfamiliar. In addition, as users have to disclose a lot of personal information when registering, privacy was considered to be a very trust-sensitive issue. Besides, since that service promises financial rewards, credibility was also seen as a crucial factor for acceptability and trust development.
An expert walkthrough and user trials uncovered several design flaws that affected people's perception of the service's trustworthiness. On the pre-interactional level, most subjects had a negative image of direct marketing, which made them were very sceptical about the information they saw. The proposed solution was to increase the visibility of the high-profile advertisers to facilitate the transfer of trust from those reputed companies to the mediating party. While only minor usability problems were encountered, subjects had difficulties getting an overview of the service. It turned out that the primary incentive for users to sign up was not the collection of points redeemable for money, but much more the push of relevant product information and special offers. Such a mind frame significantly influenced the subjects' information search strategies. It was thus suggested that details about the nature of the service should be displayed prominently on the homepage, while ensuring that a different stress be put on the benefits gained by members. In addition, subjects did express concerns about what personal information they would have to provide when registering. To make that process transparent, it was recommended that the entire registration form should be displayed on the homepage. When asked whether a privacy seal from a trusted third party would increase their trust in the service, most subjects reported that it would only be effective if it is provided by a company they know and trust, e.g. by the Dutch consumers' association as opposed to an American Internet-only company like TRUSTe.
The findings from the user trials were used to redesign the EuroClix web site before its launch. In its first six months, it convinced more than 30'000 users to sign up. This study clearly shows that consumers' trust concerns can significantly be alleviated by providing relevant information when and where users need it. The user trials further indicate that the MoTEC model constitutes a comprehensive framework to evaluate a system's trustworthiness and produce guidelines for redesign. Although this study was only concerned with qualitative data, the next step will be to develop and validate a measure of trust to allow for objective comparisons of pre- and post-design trust performances.
Egger, F.N. (2000). "Trust Me, I'm an Online Vendor": Towards a Model of Trust for E-Commerce System Design. Proc. ACM CHI2000, The Hague, 1-6 April 2000, in press.