I often hear the following argument against usability: Just hire a great designer, and you don't have to worry about that pesky user testing. After all, a great designer will create a great design, and that's all you need.

The most common example given is Steve Jobs. Granted, Jobs has been in charge of some great products. He's also produced many duds as well, the most famous being the NeXT machine and the Mac Cube. Even the Macintosh was very nearly a failure, being saved in the nick of time by Adobe and the advent of desktop publishing. (And, of course, the Mac's usability is more properly credited to Jef Raskin and Larry Tesler's user studies in the Lisa group than to Jobs himself.)

In any case, Steve Jobs is a design manager, not a designer. Having top executives who understand interaction design and care about user experience quality is indeed a boon. The willingness to delay or cancel a project because of bad UI is rare in the technology business, but it's necessary if a company wants to build a reputation for good products.

Turning to actual designers, it's certainly true that you're better off hiring a good designer over a bad one. Likewise, a good usability specialist is better than a bad usability specialist, a good programmer is better than a bad programmer, a good writer is better than a bad writer, and a good marketing manager is better than a bad marketing manager.

In all the various disciplines that come together to create a successful interface design, you should hire the best staff you can get.

The Limits of Genius Design

The real question is not whether you should use a good designer, but whether using a good designer eliminates the need for a good usability person. It doesn't.

It's wrong to rely solely on a "genius designer" for several reasons:

  • You must run your project with the team you actually have, not the team you wish you had. In most companies, you won't find one of the world's top 100 interaction designers waiting around to work on your project.
  • Design is an inexact science; even if you have a superb designer, not all of his or her ideas will be equally great. It's only prudent to reduce risk and subject design ideas to a reality check by user testing them with actual customers. (Remember, new ideas can be tested at low cost through techniques like agile paper prototypes.)
  • How do designers get to be good in the first place? By learning which of their ideas work and which don't. This feedback requires empirical data, which usability testing provides.
  • Even the best designers produce successful products only if their designs solve the right problems. A wonderful interface to the wrong features will fail. And how can designers find out what customers need? Through user research.
  • Nobody's perfect. Even a very good design can be improved when you follow an iterative process of continuous quality improvements. For each step of the design, you should conduct a usability evaluation (testing or guideline review), and use the resulting insights as the step-climbing metric to drive your user experience to the next level of quality.

Several decades' experience with quality assurance says that the best results come from following a systematic quality process, including reality checks every step of the way, rather than simply hoping that you got it right.

Best Principles and Practices

As an analogy, consider accounting. As with designers, it's better to have a good accountant than a bad one. But in either case, your accountant should follow GAAP (the generally accepted accounting principles, published by The American Institute of Certified Public Accountants). Best practices exist for a reason, and your risk of failing a tax audit is dramatically reduced when your accountant doesn't just make things up on the fly.

Similarly, user experience and website success benefit from following best practices in the form of documented usability guidelines, as opposed to making up your own, inconsistent UI.

The difference between design and accounting is that under rare circumstances you can get a better design by deviating from the generally accepted usability principles. But how do you know whether your case is indeed one of those rare exceptions? You could guess. But it's much safer to run a study to find out for sure.

To summarize:

  1. For a good starting point, get a good designer.
  2. To reduce risk, ensure that your designer works from usability data, rather than guesses.
  3. To improve quality, use iterative design and polish each round through usability evaluation.