Research has shown that the main thing Web surfers look for in a Web site is good manners and, whether you're a developer or a site owner, your company's success now depends on how your site/s rank in social behaviour. Kevin Fogarty investigatesStanford researchers Byron Reeves and Clifford Nass make the case that people dislike some Web sites - not because they are badly designed, but because the sites behave badly during their visit. In their 1996 book The Media Equation: How People Treat Computers, Television and New Media Like Real People and Places, the two argue that humans don't react favourably to rude or unsupportive behaviour regardless of whether the slight is real or virtual.
As officers at NetSage Corporation in San Francisco, the two put their research to work analysing software interfaces and developing socially conscious help agents like the animated paper clip that gives usability pointers in Microsoft's Office applications.
We asked them to analyse four leading Web sites - none of which is a NetSage client. Here are their views on which ones charm users into doing repeat business and which do more to put off potential customers than they do to lure them in.
Bottom Line: Successfully applies social rules to create a bookstore rather than a warehouse or a library.
Befitting its reputation as the premier e-commerce player, Amazon's book-buying site follows many social rules to great effect.
The consistent style and tone throughout the site communicates a reliable personality that builds comfort and trust in the business relationship. Appropriate for its products, the tone of the site is casual.
For example, Amazon tells users, `For now, you just need to . . .
Visitors have a sense that the same person is communicating with them consistently throughout their visit at the site.
This promotes a feeling that customers have a single personal assistant rather than a confusing group of merchants - all with different methods and personalities - to help with purchases. Amazon maximises personalisation with minimal information by offering suggestions based on previous purchases, discussing what people in geographic areas are buying and offering one-click shopping that uses information previously stored on the site.
It continually tells users where they are in the ordering and registration process, particularly when they're about to purchase something.
A confirmation to customers that they're `doing the right things' to accomplish a transaction - commonplace in real-life transactions - is used effectively at Amazon.
Amazon also effectively uses physical places on its site to let people know what to expect of the information presented in those places. For example, the largest column of information (the middle two-thirds of each page from the top to the bottom) is devoted to product information.
And regardless of whether a shopper is looking for books, music or electronic gear, the function of the space is unchanged. Even the details of price, shipping and discounts are identical among products.
Amazon-generated book reviews and postings from customers are mixed with book reviews (from, for example, The New York Times) to produce commentary relevant to many items.
Through careful use of language calling for related groups or customers, (for example, Purchase Circles rather than, say, folksier Neigh_bors), and the avoidance of visual clutter around the buttons that execute the purchase, they remind the customer that this isn't a library.
Bottom Line: ETrade feels comfortable but follows rather than leads.
Stock-trading sites demand access to the most personal information we possess - how much we're worth.
For customers to be comfortable with this request, the site must constantly manifest trustworthiness and competence. Even though the tone of this site is casual (in contrast to the stereotypical image of a stockbroker), its language is internally consistent. This encourages trust.
Third-party information is very clearly delineated, both graphically and through text; thus, ETrade ensures that it's responsible only for the information it controls and distances itself from more speculative or potentially low-quality information. Treating people as special because of individual attributes suggests intelligence and responsibility. There's good personalisation for heavy traders at ETrade.
Because the most valuable customers receive special service, they are flattered, and everyone is a sucker for flattery. Unfortunately, occasional customers are marked as second-class citizens. That problem could be ameliorated by encouraging customers to move to the most valuable status by providing intermittent rewards, especially when those rewards are given out of `kindness'.
Research shows that when a site helps users, they will feel a subconscious obligation to reward the site - in this case, by trading more actively. A deficiency on this site is the lack of proactive interaction.
In information-intensive activities like stock trading, it's extraordinarily worthwhile to have information presented to the customer `just-in-time'. But information provided on ETrade is 100 per cent demand-driven, meaning the customer must know what to ask for.
A help system that answers any question a user is clever enough to ask can't take the place of a stockbroker working as a teammate in what for many is a scary and lonely process.
Live financial assistants tell people what is about to be done, when the transaction is complete and, most important, that they did all the right things to make the transaction a successful one.
Bottom Line: A very disconcerting place.
Our natural social reaction when entering complex situations is to search for clues about how to behave. No one wants to look foolish.
Yahoo can frustrate people entering its site because pages are so cluttered with information that users must constantly wonder, `What's most important here? What should I do next?'
The Yahoo site seems to say, `I'll do anything you want if you can figure out how to ask.'
That may work for users who know exactly what to do, but if they don't, the passive stance feels more like a passive-aggressive one.
Most users want some notion of what to do - less is often more. Because people think of Web sites as places, it's useful for people to keep track of when they move from place to place.
Yahoo often frustrates users attempting to do this by automatically moving people from its site to someplace else. A tour guide who says, `Here's a great place', and then walks away without explanation would be in trouble with the tourists.
This practice also dilutes the Yahoo brand because there is no way to distinguish links within the Yahoo site from those that will take people elsewhere.
When users have to spend time thinking about what to do, they pay less attention to the content on the site - including the advertising.
The distractions of complexity are not only annoying; they also undermine the basic business model of a portal site (attentive eyeballs). When people divulge a lot but are greeted anonymously, it can make an impersonal and even impolite impression.
Although Yahoo gathers substantial information about user preferences and interests, there are few signs that this information is used to personalise information. Yahoo manifests knowledge of its users only in the most obvious ways (for example, fulfilling requests for particular news items). While large-scale personalisation might be technically difficult, it has enormous social benefit.
Cleverly distributed and subtle adaptations to users, such as age-specific or gender-specific language, can create a feeling among users that they have been there before and are welcomed and familiar friends.eBayhttp://www.ebay.comBottom Line: An expert auctioneer that doesn't behave like one.
People labelled as experts, whether by others or themselves, are perceived as more competent, more trusting and more likely to provide unique knowledge and expertise. But eBay's reputation as an expert suffers because it doesn't correct sellers' mistakes and is considered to be complicit in these errors. For example, typographical errors in product descriptions reduce credibility, yet eBay doesn't edit that text.
There are tradeoffs among image size, picture quality and details of the background in product presentations, but eBay doesn't make suggestions or offer ways to improve them. A lesson here is that people hopelessly confuse the errors of the `message' with the competence of the `messenger'; poor presentation undermines eBay as well as the sale items.eBay is also impolite. For example, signing up for an account can take as long as 24 hours for confirmation. eBay users also aren't alerted when they omit a field during registration.
Instead of dismissing people during that period, eBay should invite customers to browse. A good social partner tries to own the problems and makes an attempt at resolution. eBay also fails to carry through on the notion of a `personal' shopper.
It's unclear how to submit the initial form that activates the personal shopper, and the shopper doesn't save a list of items on which the customer might want to bid. Having someone remember things for you is key in personalisation.
Finally, eBay users must traverse many pages to find the personal shopper - the exact opposite of what a `personal' shopper should be. The idea of an automatic bidder, someone working on your behalf, is a social plus.
However, eBay should place more attention and emphasis on making the bidding process personal rather than simply automatic.
Byron Reeves is a professor of communication at Stanford University and a specialist in the psychological processing of media in the areas of emotions, attention, learning and physiological responses. Clifford Nass is an associate professor in communication at Stanford, specialising in the social and psychological effects of human-computer interaction. Both are also senior vice presidents for interface research at NetSage.