InformationWeek.com

Welcome Guest. | Log In| Register | Membership Benefits

October 11, 1999

TechView:
The Challenge Of Intermediates
By Jeff Angus

Jeff Angus usiness Web sites have emerged from the primordial soup of the Internet as something more evolved than the primitive pages that were so commonplace just four years ago. There is, however, something quite unnatural about the evolution of Web sites, and I encounter it every day among the thousands of sites that just aren't very usable. I call this The Challenge of Intermediates.

In nature, intermediate forms of species don't usually thrive. Consider two grazing mammals: Horses, whose necks are just a few feet long, are well adapted to grazing on the prairie floor. But giraffes, which are in gross terms quite similar, have tremendously long necks for feeding off the tops of trees. What we don't find in nature is a horselike animal with a seven-foot long neck--too short for treetops and too long for low-effort surface feeding. My recent canvassing of business Web sites, however, suggests that most companies still promulgate just this sort of useless seven-foot necked critter, an intermediate form that fails at both simplicity and feature-completeness because it's trying to do both.

The most common problem I noticed among my recent sampling of E-commerce sites was a failure to organize options around user goals. If I need to purchase an airline ticket, I expect to find a page that lets me describe where and when I want to fly. I'd also expect to see the set of questions my travel agent always asks: Will I tolerate connecting flights or changes to travel time to get a better price? Seldom are these sorts of options presented clearly and in the correct place. Users enter a site with a set of expectations, usually derived from the non-Web world. If your site doesn't jibe with them, you're bound to lose customers.

The second problem is the mindless application of technology. While I get a kick out of Java applets that show me a map of the available seats on a particular flight, it's not a very efficient way of selecting a seat. A simple ordered set of preferences (on the aisle is preferred, far forward in the cabin is next, unless there's a seat in back with an empty adjacent seat) is easier to present and faster to process.

This sort of misapplication of technology is even more common among extranet applications. Especially when the application is derived from a conventional Windows application, system architects try too hard to replicate the entire Windows interface in a lower-performing environment. As system architects create these misfits, they forget that most extranet applications aren't running in the same environment as a normal Windows application. There are now conventional menu bars. Right-clicking on objects, a common expectation in a Windows application, seldom occurs to Web users. HTML frames don't really work like the fixed pane on the left-hand side of Outlook. Ditto for drop-down lists.

Trying to build a gazelle by gluing together parts of a giraffe won't succeed. Like natural selection, usability is a systemic process, rather than a set of top 10 traps to avoid. The process isn't all that complicated and need not involve expensive user experiments. There is, however, at least one Darwinian reference that can help most sites get back on the evolutionary track: Jakob Nielsen's www.useit.com. You'll find simple practical advice that will help you keep customers from looking elsewhere.

Back to Labs
Send Us Your Feedback
Top of the Page