Introduction

With the holiday season upon us, online merchants are gearing up. Online shopping seems like an easy way to get through that shopping list, leaving more time for friends and family. After all, online shopping is the same as going to the mall, without the challenge of competitive parking. Or is it?

Electronic survey tools present a similar problem. In recent years, a number of powerful survey tools have emerged on the Web. Some of these tools make it easy to post very sophisticated surveys and collect data on the Web. Distribution is cheap. Data entry, verification, and descriptive analysis can be completely automated. Just-in-time information survey help (e.g., links to definitions of terms) and multimedia resources can be embedded directly into the survey. Since contingent question presentation can be handled invisibly, there is no more: "If you answered, 'Yes,' skip to question 26b". Further, using electronic mail to invite participants to the survey, transitions participant recruiting from a pull to a push activity.

These are all very powerful tools. You can literally implement a survey in a few minutes. It is easy to forget, in this domain of apparently instantaneous productivity, that there is an entire field of knowledge that supports the design of effective and meaningful survey data collection. As with Photoshop, the power of the tool distracts us from the essence of the task.

e-Commerce n'est pas shopping

Although many designers and e-commerce merchants don't appreciate it, transacting on the Internet is very different from shopping at the mall for most consumers. Bricks and mortar establishments provide a variety of explicit cues conveying information about the professionalism and competence of the staff. The age and dress of the staff, the style of sales interactions, and merchandise displays all help convey important information to would-be customers that automatically shapes their expectations for subsequent interactions and service. In most cases, fulfillment is instantaneous: we simply trade payment for goods at the store itself. Finally, the physical existence of the store provides a clear and simple strategy of recourse should problems with the goods arise after the purchase itself. And we all know the drill for returns. Shopping in a bricks and mortar establishment follows a fairly standard and predictable process – even when things go wrong. The evidence that a vendor will hold up his end of the deal is embedded in the purchase process itself.

In contrast, the recognizable social cues and predictable process that reassures customers in physically-based transactions are largely absent in electronic ones. E-commerce transactions are disembedded interactions. There is no building or staff to evaluate. There are only pictures of merchandise. Transactions are not instantaneous and continuous. Payment is often rendered well before a product is delivered. There is not always post-payment assurance that delivery will occur. Transactions typically involve little or no 'dialogue' between the merchant and customer. In addition, critical parts of the sales interaction are ambiguous: shipping costs and processes are often not well presented. Avenues for recourse for problematic transactions are unclear. Newer consumers are not sure which (if any) cues indicate that a vendor is trustworthy. More sophisticated consumers do note specific cues to trustworthiness – however, they are also suspicious about the meaningfulness of those cues.

In addition to the disembedded nature of electronic transactions, the technology used to make them happen is unfamiliar. The technical processes that underlie electronic transactions are, at best, unclear and uncomfortable for most consumers. Concerns about privacy and credit card security increase the perceived risk. While experienced consumers understand that the locus of transactional risk is the vendor, novice users sometimes believe that even trustworthy vendors cannot protect them from the vagaries of online transactions (Riegelsberger & Sasse, 2002).

For many, the process of buying things online is still somewhat mysterious. Like the first time you wrote a check or used a credit card, it doesn't quite seem real. The cues that signify the trustworthiness of a bricks-and-mortar merchant are often missing online. Those that are present are not always valid. All of this makes Internet transactions feel risky.

Researchers are not users

In stores, the perceived risk of transactions is mitigated by social cues embedded in the experience. If the right cues are there, merchant trust comes for free. That is, in the bricks-and-mortar environment, we look for familiar cues, and often accept their mere presence as a heuristic, non-compensatory strategy for establishing trusting in a new vendor. So the challenge is to identify, and virtually embed, cues that will build consumers' understanding and trust in the system, reducing the perceived risk.

Interestingly, research working on understanding consumers' reluctance to buy online focus primarily on how to lose consumers trust, not how to gain it. It's a "physician, heal thyself" phenomenon. For researchers (and designers and e-merchants) the process of shopping online is like shopping in stores. The process is predictable. The technology supporting the transactions is not mysterious. They know when to be suspicious, how to recognize trustworthy vendors, and the flags that indicate an interaction may not be successful, and the possible strategies for recourse. Their experience with and confident understanding of the system (along with its vagaries) means that vendors do not need to gain their trust initially. They are going to transact online. The way to lose their business is to violate their trust. They study the problems that encroach on their world. The good news is that researchers behave a lot like very sophisticated Internet users.

Before you can lose my trust you have to gain it ...

For most consumers, however, the online transaction process is still not so familiar. The process often varies from merchant to merchant and the heuristic cues to trustworthiness are still unstable. Therefore, in addition to researching the parameters that cause consumers to lose trust in online vendors, researchers also need to explore the elements of design that help consumers establish trust in the first place.

A few groups of researchers have begun exploring how virtual re-embedding strategies might be applied to foster trust in the online shopping environment. Virtual re-embedding means using elements such as pictures, movies and interactive chat to provide familiar social cues for developing trust. The simplest approach to virtual re-embedding on the Web is to add pictures of people. Research in other domains of disembedded interactions clearly indicates that face-to-face interaction helps. Including photos of authors can increase the credibility of online articles (Fogg, 2002 ). Exchanging photos of participants enhances collaboration in social-strategy games played over chat (Olson, et. al., 2002). Collaborative interactions of distributed working teams are facilitated through face-to-face meetings and team-building sessions (Rocco, 1998). Print advertisements depict human faces rather than pictures of the touted product to provide social cues to the desirability of ownership, suggesting that a simple picture is sufficient to socialize the relationship.

Can it be that simple on the Web, as well?

Face it, it depends....

Recent anecdotal and research reports exploring the effects of including human photographs on transactional Web sites have returned mixed results.

Steinbrück (2002) conducted an experiment comparing online banking sites with no pictures, with an unlabeled employee picture and with the same picture labeled as a nameless customer service representative. After about fifteen minutes of combined free exploration and transaction simulation tasks, participants rated the trustworthiness of the sites overall. Independent of their Internet experience, participants rated the site with the labeled picture most trustworthy. In addition, the site with the unlabeled picture was considered more trustworthy than the site with no photo whatsoever. Steinbrück concludes that customer service representative photographs on banking Web sites establishes a human point of contact in an otherwise intangible, virtual company facilitating the creation of consumer trust.

Subsequent research exploring the trust building impact of photographs on transactional sites, however, has not universally been positive.

Anecdotal findings from Nielsen (reported in Riegelsberger, Sasse and McCarty, 2003) suggest that embedding photos of "perfect" people on an Intranet – people who are too consistently beautiful to be real employees – actually undermines credibility, and therefore undermines trust.

Research by Reigelsberger and colleagues begins to provide insight to the inconsistencies reported previously.

In their qualitative study, Reigelsberger & Sasse (2002) identified two distinct groups of users (out of four) who responded negatively to photographs in Web sites. Interestingly these users are on opposite ends of the e-commerce use spectrum. Non-shoppers with a very low baseline trust in e-commerce interactions responded negatively because they interpreted the photos as strategic attempts to manipulate their trust. In contrast, very experienced shoppers considered the photos gratuitous design elements, unnecessarily cluttering the interface without adding meaningful functionality.

Humans are especially wired to notice other human faces. Developmental psychology research demonstrates that normal babies orient preferentially to human faces over other types of stimulus. (Crack babies don't.) Neuropsychology research demonstrates that there is a special module of the brain dedicated to face recognition. Damage to that area of the brain results in a disorder called prosopagnosia, or the inability to recognize familiar faces. (See work by Oliver Sacks for interesting studies.) Eye tracking studies exploring the trust-building effects of photos on Web sites demonstrate that while photographs have no deleterious effect on task performance, pictures of human faces do systematically draw the eye on at least the first view of a Web page (Reigelsberger, 2002). If we notice faces automatically, then designers must bear in mind that users will also incur a cognitive cost if they suppress the urge to glance at photos that do not contribute meaningfully to the task flow on Web sites.

Which comes first? The interaction between trust and credibility

Brand credibility also influences trust in online merchants. Consumers new to e-commerce often make their initial purchase on the site of a familiar bricks and mortar establishment. In this progression, confidence in the brand translates into trust in the e-commerce arena. If trust derives automatically from the credibility of the overall brand, will re-embedding strategies like adding photographs provide any benefit for already credible sites?

Reigelsberger, Sasse and McCarthy (2003) explored this question by varying site credibility, the presence/absence of a photograph, and the trustworthiness of the person depicted in the photo. Both the credibility of the site and the trustworthiness of depicted individuals were established through pretest norming studies. In this study, Reigelsberger & colleagues discovered that vendor characteristics (credibility) and person characteristics (trustworthiness of person) interact. In the absence of a photo, participants could accurately assess the relative credibility of sites. However, the presence of photos undermined participants' ability to differentiate credible from less credible sites. Photos increased the trustworthiness ratings of non-credible sites and decreased the trustworthiness ratings of credible sites. In this study, the presence of photos actually undermined the brand of credible vendors. Positive trustworthiness ratings more closely related to professional page design.

So what's a poor designer to do?

Based on these studies it appears that trust in virtual transaction systems may be more about comfort with, and understanding of, the e-commerce system than about the embeddedness of the relationships within it. This means that creating trustworthy e-commerce sites will involve creating predictable service systems in which customers understand the transactions process, in which fulfillment and follow-up services procedures are consistent and predictable, and in which the flags to risky interactions are consistent and generally recognized.

In the mean time, should you include photographs in your Web site? As with most questions in design, the answer is "It depends."

The research suggests some broad guidelines. Pictures of people make virtual transactions more familiar and, as such, sites seem more trustworthy, with the following caveats:

• If your brand is already credible, photographs will not enhance its trustworthiness.
• Photographs do not enhance the trustworthiness of sites for users who are not confident about the trustworthiness of the Web in general.
• Photographs without functional value can undermine overall perception of sites for very experienced Web users by interfering with task completion.
• Social re-emedding cues are perceived as trustworthy only when they are seen as being given unintentionally.