"Usability is A Sleeping Giant"
By Wendy M. Grossman, Mon Jun 11 00:00:00 GMT 2001

It's probably the hardest thing in the world to make something simple.

Late last year, several reports were published pointing out the usability problems with WAP services. The most widely covered of these was Nielsen Norman's user study, which asked 20 London-based users to carry out a set selection of tasks. The second was a study conducted by Usable Products Company, which studied the way six people approached using WAP services for simple tasks. Both studies concluded that WAP was difficult, slow, and frustrating. The WAP Forum of course responded by complaining that the sample size in the Nielsen Norman study was too small to be statistically valid.

But even those involved in WAP development compare the current state of WAP, just as Jakob Nielsen did, to the state of the Web in 1994, when it was slow, buggy (in the sense of broken links), and hard to navigate (imagine: no search engines!). Mike Mills, chief interface technology for the software and services company Alter Ego, says "Usability is a sleeping giant." Alter Ego provides middleware to help adapt Web sites into WAP services.

The constraints on WAP design are obvious: slow network speeds, tiny screen, lousy data entry facilities, high costs to users. It's not an easy change for a developer used to working for the Web and the richness of the personal computer environment. What's harder to put together is a set of guidelines that lead to good design for WAP.

Even so, some basics are beginning to emerge - both Alter Ego and microbrowser supplier Openwave have published guidelines, which agree in many places. For example, given the poor data entry facilities on most phones, users are happier with numbered picklists, where choosing a menu option is hitting a number key. In any environment writers learn to be economical with words - in this microenvironment they must learn to be economical with characters.

Openwave recommends, for example, that labels for softkeys (the navigation aids that appear in the bottom left and right of the phone window) be limited to no more than five letters, to make sure they work on all phones. Yet those labels need to be clear and immediately understandable.

Mills says the worst mistake is trying to replicate the Web and squeeze all the information into WAP. More than that, he says, a big problem for users is being able to navigate back to the pages they've already visited. On a site the company adapted that features consumer reviews, he says, "We have a little go-to screen that gives the logical links so they can get back, simulating breadcrumbs. Even though screen space is so precious, just replicating the name of the category of the screen they've come from helps people have a sense of orientation. It sounds really simple, but if you don't do it, you make life hell." It's, he says, common to encounter sites where you have no idea how to get back - much like some companies - voice menu trees - "If it happened to me once, I would never go back." Voice input coupled with visual output may help solve some of this.

A big problem, of course, is the wide variety of phones - as if you were designing the Web back in 1979, when no two computers worked alike. Nokia phones have a little roller for selection, so Nokia users like long scrolling lists of menu items; but if all you have is two softkeys and no roller and you have to push a softkey repeatedly to scroll, you would prefer a shorter, collapsed list from which you could pick a category.

Mills says that there's an important misperception about the Web. "Even with a rich Web site, people are only processing a small chunk of it at once. So don't think of WAP as designing for some impoverished device, but as giving you a chance to reformat and present information in very usable mind-sized chunks."

More obvious things to avoid include splash screens (people are paying by the minute here), finding ways to minimize the number of screens involved in a given task (for example, eliminating extra confirmation steps), and being inconsistent about labeling.

The problem for those designing WAP services is that it's such an early stage. Most technologies have a somewhat forgiving initial audience of early adopters who are so desperate for the new device that they'll put up with anything - even if it means typing arcane strings of gibberish with two fingers over a 1200 baud modem beset by line noise. (I know: I've once spent something like four hours downloading an 11Mb video clip of an exploding whale (now visible to you, in real time) though to be sure, strictly speaking I didn't actually need it.)

Computer software was a growing industry for more than a decade before usability became a serious issue - somewhere around 1991. Many parts of the Web haven't grappled properly with usability yet - it's estimated that some 80 percent of ecommerce shoppers bail on orders before they're completed.

But WAP's problem is more severe much earlier in its history. User-wise, it's back where the Web was in 1994: slow, empty, confusing, trying to prove its value. But the users it's aimed at are the broad mass of the consumer market and time-pressured business professionals, not geeks willing to spend countless hours and dollars exploring something new.

WAP's other problem is that it has too much competition. Say you're talking on your mobile phone, trying to fix a time to play tennis in the next few days and you want to check the weather report. You could hang up, navigate the WAP service, find the weather report, and call or SMS back. Or you could keep walking and buy a newspaper (probably cheaper). Similarly, if what you want is to check real-time stock prices, you could find a pub that's showing CNBC or stop into an Internet cafe, where for a couple of dollars not only can you check your portfolio but do your email as well with a real keyboard, something to drink, and someplace to sit. For WAP to be appealing, it's not just that the network has to be faster - although that's important, too, and a popular reason for its underwhelming take-up so far - but services have to be designed so they are the best and fastest way to do things.

This is a genuinely hard problem, and it's no wonder that so far no one's done very well at it, although eBay and Amazon.com have made what most people think are a pretty good start.

Mills seems optimistic. "It's going to be pretty painful for the next few years," he says, "but in the next six to eight months we should see some people getting it right. Once we have always-on and easy bookmarking, it will start to be useful. And in three to four years we really will start to see the pervasive Internet." He's probably right: one think about technology is you can keep working on it until it gets better.

Wendy M. Grossman is a freelance writer based in London, and author of net.wars. Wendy's new book, From Anarchy to Power: The Net Comes of Age, is out. Sample chapters are available here.