In one of Oracle's (ORCL) shimmering glass towers in Redwood Shores, Calif., on the first and second floors, are five offices, a conference room, a classroom, and a lounge, complete with a three-seater couch, end tables, and a full-size TV. The complex might be any corporate meeting area were it not for the hidden cameras, two-way mirrors, and scanning equipment that records who does what there -- all part of an 8,000-square-foot usability lab where 65 engineers test, retest, and tweak the design of Oracle's software interface to make its products easier and more efficient to use.
Oracle opened the lab about eight years ago. But its emphasis on usability didn't really gain momentum until 1998, when CEO Larry Ellison began to notice that potential customers frequently cited an inability to figure out how to use Oracle's products as a reason for turning away its sales reps. So Ellison poured money into the lab while making sure that Oracle's product developers knew that usability was a top priority.
"The benevolent dictatorship that's Oracle really helped us," says Dan Rosenberg, the company's vice-president of development for usability and the lab's manager. "We not only got dollars, we got boardroom support for the idea that usability pays."
EVERYONE BENEFITS. That's true now more than ever. Throughout Corporate America, chief information officers have tightened their budgets. No pennies can be wasted -- especially on software that requires lengthy and expensive training to use. Nor is the only goal of usability testing to help customers save money. Software suppliers are benefiting, too, because by paying more attention to usability, they often catch mistakes that otherwise would make it into the final product.
Two decades after Apple Computer (AAPL) revolutionized PC software with its intuitive, easy-to-use designs, dozens of top tech companies are finally getting serious about usability. Microsoft (MSFT) employs 140 usability engineers who are attached to one of the outfit's 20 divisions. Businesses that don't have their own user-interface departments are hiring outsiders to get the job done. The 85 usability professionals at Human Factors International, a consultancy in Fairfield, Iowa, are 100% booked through the first quarter of 2003. Not bad, given the economy's sluggish state.
Not long ago, most software concerns -- and even their customers -- would have thought usability an expensive luxury. Back in the 1970s, it took a multimillion-dollar mainframe to complete tasks that today can be done by the average desktop. Thus, the idea of dedicating valuable processing power and programming to the task of designing screens that simplify and speed up the use of a particular program or even of an entire computer network was laughable.
GOOD INVESTMENT. The arrival of cheap, all-powerful desktops changed that. Suddenly, it became economically sensible to spend time developing resource-intensive user interfaces -- screens that graphically represent the arcane, typewritten commands that would otherwise be needed to make a piece of software run. Today, colorful menu bars and glitzy graphics are just some of the sophisticated tools usability engineers have in their arsenal. They also use scan recorders to track how computer users move through a software product's command screens and where they make mistakes -- then strive for ways to make the task so intuitive that it's idiot-proof.
Usability experts also perform in-depth studies of how businesses run, so as to ensure that software with a six-figure price tag is compatible with existing workflows -- or even better, simplifies them without requiring a sweeping reorganization, which might doom the software installation to failure.
Buyers have become much more demanding about usability because they know what's possible to do now with a high-performance PC and a little imagination. The Web has accelerated the process. After all, Web sites such as Amazon.com are really just complicated pieces of sales and tracking software with a simplified yet very powerful "front end" that an average person can easily use. "The Web has helped because it brought technology to your grandma in Decatur, Ill.," says Eric Schaeffer, CEO of Human Factors International. "You can't train her. And if she can't figure it out, she'll just go away."
CLEARER WINDOWS. How right he is. E-commerce sites lose nearly half of their potential sales because visitors can't figure out how to use them, reports researcher Jakob Nielsen, a principal at usability-research outfit Nielsen Norman Group in Freemont, Calif. At many such sites, a 5% improvement in usability could increase revenues by 10% to 35%, according to studies by NetRaker, a usability and market-research firm. So, for a business the size of Amazon.com, an investment of $200,000 could theoretically translate into an extra $158 million in revenue.
That's a powerful message, even for software makers who have a captive clientele. Take Microsoft. The Redmond (Wash.) software giant is hardly known for programs that are easy to use. And yet, its new operating system, Windows XP, is getting rave reviews from usability experts and XP customers, too.
Nielsen Norman principal Bruce Tognazzini is both. After 14 years as head of Apple Computer's interface-design group, "Tog," as he is known, is semiretired and traveling around the country in a high-tech-outfitted camper. The motor home has two computers, linked together via a wireless router. "The minute I plugged in the router, Windows asked me if I wanted both computers to have access to the Internet," Tognazzini says. "A few clicks, and the task was done. A few years ago, it would have taken a user manual and several calls to tech support to make that happen."
DRAMATIC RESULTS. Microsoft calls it inductive user-interface design. Rather than giving PC users huge menus with options they may never use, XP tries to anticipate what the user wants to do. And instead of providing an icon called "users" to allow someone to set up profiles for multiple people, XP now asks: "Would you like to set up preferences for different users on this machine?" Says John Pruitt, Microsoft's lead researcher on usability: "In the past, we expected computer users to become experts over time. Now, we try to give them only the information they need to make a decision."
Most usability changes are more basic than that -- such as reducing the number of clicks necessary to make a purchase, or changing colors, typefaces and font sizes to make text more readable. The results, however, are just as dramatic. After Dell Computer (DELL) applied basic usability principles to its e-commerce Web site in the autumn of 1999, its Web sales skyrocketed. Online purchases rose from $1 million per day in September, 1998, to $34 million per day in March, 2000.
Another example: After Oracle changed the navigation structure on its database manager, Oracle's Rosenberg says, database administrators were able to perform their duties 20% faster.
LONG WAY TO GO. Mistakes still happen, of course. Human Factor International's Schaeffer says he recently reviewed the Web site of a major auto maker that violated countless usability rules. For one, it had red text on a blue background, a terrible combination because the human eye focuses on red behind the retina, and blue in front of the retina. The result can be a condition called chromostereopsis, which can result in headaches and dizziness.
Another common flaw: Sites use all capital letters. Many engineers think that will draw a visitor's attention, but it actually slows reading time by 14% to 20%. That's because the human eye reads by looking not only at individual letters but at the shape of words, and all caps make every word a standard rectangle. "Many companies think usability means going out and asking users what they think. But if you ask users how to design something, you often end up with a terrible interface," says Schaeffer.
So, usability has come a long way -- but it still has a long road ahead before software becomes a snap to learn. Jared Spool, a principal at Boston consultancy User Interface Engineering and a professor of engineering at Tufts University, says software-interface design is at the same stage that automobile design was back in 1909. That was the year Ford introduced the Model T, a revolutionary car that was affordable and reasonably easy to start, with big wheels that could navigate unpaved roads. In 1909, Ford sold more Model Ts than all of its previous models combined.
But any driver who tried to drive a Model T today would find the vehicle complex and clunky. Says Spool: "We think we've made great strides, but 20 years from now we'll look back and laugh." By Jane Black in New York