User Testing the Wired Way
At this point you're probably thinking, "If he's talking so much about asking people what they want, why is HotBot that nasty green? And what's up with that flippy frame on the HotWired frontdoor?"
Those are the result of another uniquely Web-related phenomenon, the hybrid product. Unlike a piece of commercial software, the reader of a Web site has not invested hundreds of dollars in buying the content. Unlike a magazine, she won't have had years of experience with the medium. Unlike television, she'll have a dozen identical places to get the same information. Unlike all three, it's doubtful that there will be an ad campaign to help her remember where she got that great knowledge. And so, Web sites need to juggle a number of functions: They need to be useful, they need to sell ads, and they need to differentiate themselves from their competitors (to self-market). Hence, "HotBot Green."
Thus, a Web site has to solve a different set of problems than has ever existed in media: It has to be usable, be memorable, and sell ads (or, if not ads, then itself). Moreover, with the speed of current technological innovation, it has to change continually, updating itself to stay competitive. Without continually testing a site's interface and comparing it to its competitors, there's no way to know which of the emerging technologies to invest in (remember "push"?).
Because of this hybrid, dynamic nature, and the short production schedules, our approach is different from traditional usability testing. What we do at Wired is something between market research and usability research. We don't just test whether people can use something, we also test whether they want to use it - whether it's desirable. Moreover, we test advertising effectiveness, because our advertisers are ultimately our customers, too (we show their ads in exchange for them paying for us to give content away for free). What we end up with is a complex interaction that lets us design products that attain a balance of usability, desirability, and advertising performance.