 | Level: Introductory Jakob Nielsen (http://www.useit.com), Usability guru, Nielsen Norman Group
01 May 2001 Usability is not simply a matter of making users like your product, even though subjective satisfaction is important and usually increases substantially when you design a system to better match users' needs. Mainly, usability is a matter of how easy a system is to learn and how efficient it is to use after you have learned it. Great usability does not come simply from wishing for it or from promising to be good to your customers. Usability follows from integrating usability methods into the development process and from running user tests as soon as you have a prototype design.
I estimate that good usability methodology can be expected to lead to the following improvements in the main usability attributes for the average development project:
Learnability
(for novice users) | 100% | Efficiency of use
(for experienced users) | 25% | | User errors | 500% |
(Note that a 500% improvement in user errors really means that the number of
errors is reduced to a fifth of the previous number.)
These are estimated improvements for computer software that has not been subjected to systematic usability engineering methods. If the focus is Web sites, the potential for improvements is much greater. Software design is fairly established, so most current software products derive a minimum level of usability from the very fact that they follow a tradition where certain principles have been learned over the years (for example, "Undo" is a good feature that helps users explore a product with less fear of making irreparable mistakes).
In contrast, the early Web sites were wild experiments that often resulted in miserable usability. Even today (where the worst excesses have been removed from most leading e-commerce sites and many corporate sites), it is still true that the Web is much too difficult to use and that people usually fail when they try to use a new site. I expect that leading Web sites in ten years will have at least a thousand percent higher usability than current sites.
But back to software. My estimates in the table above apply to average projects that have not used usability engineering in the past. Such software almost always hides several usability disasters that significantly hurt learners or that make certain operations particularly error prone. That is why the biggest improvements usually are seen on the measures for learnability and user errors. Efficiency of use for the experienced user rarely increases quite as much because humans are so flexible that they can use almost anything once they have learned it.
One way of looking at the improvements is that doubling learnability is the same as cutting the training budget in half. When developing software for in-house use, this one benefit can pay for all the usability costs. When developing software for the open market, even bigger savings can often be realized in the form of fewer calls to the support center due to fewer learning problems and fewer user errors.
After the biggest usability problems have been removed, there is still potential for further improvements in future releases. Research on iterative design illustrates that usability goes up substantially every time one revises a software product after the use of systematic usability methods.
Subsequent usability improvements after the first huge gain are typically on the order of about 30% for each additional iteration. This never seems to end: there is no such thing as the perfect software product that cannot be made even easier to use. So the available budget and the available time are the only limits to the number of iterations that are advisable.
Basic usability methods
Most of usability engineering steps involve some form of user testing. There are several other valuable usability methods that can be used in more advanced projects. I immodestly refer you to my textbook,
Usability Engineering
, for an in-depth discussion of these other methods.
Basic user testing is very simple. There are three things to keep in mind:
- Get real users who are representative of your target audience. You cannot test with your colleagues or others who know too much about the project. The exception is intranets where your colleagues are the representative users. Even so, don't test with other members of the project team.
- Have the users perform real tasks with the design. Don't simply have the users play with the interface. They can always find something to click on, but the true test of usability is whether a user can solve a specific problem. The test tasks should be chosen to be representative of the types of tasks that the product is intended to support.
-
Shut up and let the users do the talking. The goal of the test is to elicit comments from the users and see what they think as they proceed in solving the test task. The goal is not to demonstrate your own superior intelligence or to help the users get their work done faster than they would on their own. Remember that users will not be sitting next to one of the designers when they use your product. You want to see how they use the interface under realistic circumstances (this means without help).
As they progress through the tasks, users should be asked to think out loud. This will show you users' misconceptions and allow you to identify those aspects of the design that cause problems. Sometimes you will discover simple usability problems like the wording of a menu item that makes users think that it does something different than it really does. Other times you will discover bigger issues like a mismatch between the task flow in the system and the way users approach their work. In any case, simply by hearing what users say as they interact with the system, you will build up a long list of things that need to be changed.
Advanced user testing can add several complications such as eye-tracking, use of a professional usability laboratory, or ways of managing a test in a foreign country. However, basic user testing is really quite simple: get real users, have them do real tasks, and let them do the talking. Anybody can do this, and everybody should do this. If you release a software product without some user testing, you will have wasted most of your development budget, and you will have to release an upgrade much too soon. I sometimes like to say that your product will be tested by the users. The only question is whether the test takes place before or after the release. And we all know how much more expensive it is to correct problems after a product has been released.
How many users?
Amazingly, it is enough to test a design with five users to find the vast majority (usually 80%) of the usability problems. Sure, people are different, thus you need to test with more than one person, however; there are great similarities between the types of mistakes people make when using computers. So, after the fifth user, you will have seen the same patterns repeat themselves so often that you will be itching to get back to the drawing board and improve the system.
If you have the budget for more than five users, it is better to save the resources and run a second test later with another set of users. In the time between the two tests, you should redesign the interface to remove as many of the usability problems as possible. When you run the second test, you will gain three benefits:
- You will find those last few usability problems that were not identified in the original test.
- You will discover whether or not your supposed "fixes" were in fact improvements to usability or whether they introduced their own weaknesses to the design.
- You will study users as they progress further into the system because they are now capable of doing more things; this will uncover additional, deeper usability problems.
About the author | | | Jakob Nielsen, Ph.D., is a Web usability guru and a principal of Nielsen Norman Group, which he co-founded with Dr. Donald A. Norman, former Vice President of Apple Research. Before starting his own company, Nielsen was a Sun Microsystems Distinguished Engineer; he has also worked at the IBM T. J. Watson Research Center, Bell Communications Research, and the Technical University of Denmark. Nielsen is the author of several best-selling books; his next book, Designing Web Usability: The Practice of Simplicity , will be published by New Riders on November 29, 1999. Dr. Nielsen has been called "the guru of Web page usability" (The New York Times) and "the smartest person on the Web" (AnchorDesk), and "the next best thing to a true time machine" (USA Today). He holds 50 United States patents, mainly on ways to make the Internet easier to use.
|
Rate this page
| |
