Summary: Having a specialized usability person is best, but smaller design teams can still benefit when designers do their own user testing and other usability work.
At my seminars, I'm often asked whether designers and developers can perform usability activities or whether those activities should be left to dedicated usability specialists. The answer depends on your circumstances; there are several important pros and cons to having designers and developers branch out into usability.
Con: Specialization Drives Performance
We've known since Adam Smith that specialized workers are more productive than people who try to do everything. That's true in the user experience disciplines as well. You can't even talk about "designers" as a single group. There are graphic artists, interaction designers, information architects, writers, and many other professionals, each of whom specializes in designing some aspect of the total user experience. Sure, a single person can do both visuals and information architecture, but such efforts will rarely match the quality of work done by dedicated specialists.
Indeed, even usability professionals often specialize in usability subfields, such as quick qualitative studies, formal measurement studies, field studies, competitive studies, site analytics, surveys, guidelines and standards, and so on.
The more different activities you have to do, the less time you'll devote to learning each one's intricacies — and the less experience you'll build up with each one as well. Lack of experience is especially problematic for usability, because the ability to correctly analyze user behavior is extraordinarily dependent on the experience of having previously observed a wide range of behaviors.
The argument for specialization is particularly compelling for design vs. usability because different personality types tend to excel at each discipline. Design obviously appeals to people with a drive to put things together, whereas usability requires analytic thinking and conceptualization skills.
There are many other fields in which people might branch out into different job categories. In filmmaking, for example, a lead actor could write the script, but it's usually better to have a skilled screenwriter do the job.
Pro: Less Staff Required
It would be great if every project team had ten designers who were each experts in different aspects of user experience design. And it would be great to have a bunch of usability professionals support that design by working on multiple forms of user research and other usability activities.
A big company typically has a user experience department with dozens or even hundreds of such dedicated specialists — once it's moved sufficiently up the maturity scale to invest appropriately in user experience. But a small company (or an immature big company) won't have such a large team.
At many companies, the user experience teams are too small to justify a dedicated usability professional. Indeed, many project "teams" consist of a single person. Luckily, usability basics are easy enough to learn: we take team members through a simple user test of their design in a 3-day workshop.
Not having usability people is no reason not to have usability. Your team will benefit from doing some of the simpler usability activities itself, especially if following an Agile development process. Discount user testing can be done with minimal resources.
Con: Lack of Objectivity
If you test your own design, however, you might be less willing to admit its deficiencies. Designers can be too willing to dismiss user complaints or problems as minor or unrepresentative, when in fact the test indicates the need for a deep redesign. Also, designers can get so caught up in their own theories about how users ought to behave that they forget to test for cases in which people behave differently.
One of the key things we teach in our user-testing courses is how to write good test tasks, because most novice test facilitators use the wrong tasks and get poor data as a result. Designers are susceptible to employing tasks that focus almost exclusively on their own pet features rather than on goals that users really want to accomplish.
Of course, if you know you might lack objectivity, you can proactively work to overcome this deficiency. For example, you can force yourself to include test tasks that go beyond the things you personally care about.
Designers can also increase their testing objectivity by asking colleagues to review their test plans or to listen in on a session or two.
Pro: Higher Credibility, Easier Communication
When the same person does both design and usability, you don't have to worry about the designer dismissing the usability person's findings. People tend to believe in their own work!
When different people focus on different project aspects, they have to communicate, which takes time in meetings and time for report writing. In contrast, when a designer runs a test, he or she knows what happened and can immediately start redesigning to fix the problems that the test identified. No meeting, no report, no communications overhead required, so long as the info is lodged within a single brain.
The downside to the lack of meetings and reports is that the usability findings don't get refined and discussed, which again leads to fewer deep insights and major user-interface reconceptualizations. Even though usability reports are overhead, they're also a good way to build institutional memory that will help new designers and future projects.
Any Test Is Better Than No Test
If you can afford it, it's better to have dedicated usability specialists perform your project's usability activities. But, the choice is not between this ideal and doing nothing. For many projects, there's a middle road: let the designers or the developers do double duty and take on some usability work. This is much better than having no usability at all.
More details in the course on The One-Person UX Team Tool Box.