Shopping the other day on garden.com, I selected one item to purchase and continued browsing. After considering a few more items, I decided I'd shopped long enough.

I looked for a link labeled "check out," but there was no such link on the page. I glanced high and low for the ubiquitous shopping-cart symbol, but it was not to be found. Then I carefully studied every link and graphic on the page, trying to find a way to buy the item I had selected. No go. After continuing to bumble around the site, I finally discovered that garden.com's term for shopping cart is "wheelbarrow" and that there is in fact a wheelbarrow text link on each page.

Screenshot: Garden.com's wheelbarrow link.

Obvious? I think not.

Note: After this article was written, garden.com was acquired by Burpee seeds, and it now provides a standard shopping cart icon and text.

I've noticed a strong new interest in usability testing, thanks to the growth of the Web. If an e-shopper gets lost in a maze of clicks and doesn't make a purchase, that's a big problem. If a visitor decides not to register at a Web site due to the lack of a privacy statement, that's a problem too. With proper preparation, usability testing can isolate these costly problems and point the way to solving them.

In its simplest form, usability testing requires only three ingredients: an application or Web site, a usability tester, and an observer. The usability tester navigates the application or Web site as a real user would, and the observer identifies usability problems by observing the tester's efforts. Add a few ingredients -- more testers, more observers, a facilitator, a set of well-crafted test scenarios, and some video equipment -- and you've got a recipe for success.

Note that usability testers are not the same as software testers whose job is to find bugs in the program. These testers' sole concern is the functionality of the application or Web site.

Be prepared

Careful planning helps immensely in getting beneficial results from this kind of test. The key is to have a clear understanding of the target users and their goals in using the application or Web site. With that understanding, you can choose a representative set of users and create an intelligent set of test scenarios.

Five or six testers should be plenty if you plan carefully to ensure that they represent the user base. Pick testers who represent the range of knowledge and abilities of the people who will use the application. Usability testing can help you understand how easy it is for beginners to start using the application or Web site. It can also help you see how well users make the transition from beginner to intermediate. And it can show you how well your power users fare.

During the planning stage, you should prepare, try out, and refine test scenarios. A test scenario describes a particular task and asks the tester to accomplish the task using the application or Web site. The set of scenarios should provide sufficient challenge and time to uncover any significant struggles with the user interface. By trying them out ahead of time, you can make sure they are clearly written and are not too much of a challenge for the allotted test time.

Trial run

Make sure you select a facilitator with excellent people skills. The facilitator's most important job is to make the testers feel comfortable, to help them understand the objectives of the test, and to clarify their role in usability testing. The facilitator should explain that the purpose of the test is to uncover problems with the software's usability -- not to test the tester.

Be kind to your testers by not keeping them any longer than is absolutely necessary. Limit their test time to 30 minutes if possible. Plan time in the schedule for debriefing each tester and then for debriefing the observers after each test session.

If possible, get everyone on the development team involved as observers. After they've finished watching the users struggle through the test scenarios, they'll be your best allies in making improvements.

During a trial, the tester attempts to follow one or more scenarios. Observers watch and take notes on how easily tasks are accomplished; they note the tester's emotional responses as well. The facilitator should ask the tester to think out loud; this makes observers aware of assumptions users tend to make about the user interface.

Observers should try to note the exact user-interface features applied to each task and the time needed to finish. They should also note the problems encountered and the number and type of hints given to help resolve the difficulties.

The facilitator should make sure the test moves along according to schedule. Afterward the facilitator will debrief the tester, relaying questions from the observers and asking for general feedback, and will then debrief the observers after the tester leaves. Immediate debriefing is desirable. The observers' and testers' findings should be consolidated while the test is still fresh.

After all testers have been observed, the facilitator may lead an additional work session with the observers to analyze and report the usability test findings. If a videotape was made, the observers might want to review portions of it.

Time, cost, and videotape

You'll need to decide if it's worth the extra effort and cost to videotape the test sessions. Tapes can come in handy for comparing conflicting observations. Also, seeing is believing. So if you anticipate needing to convince management, you'll definitely want to tape the test sessions. Put together a persuasive set of excerpts from the tape to back up your test findings.

While a standard video camera will suffice, you can also rent or buy specialized usability testing equipment. Typically, this specialized equipment will allow you to videotape the software in action and superimpose the tester's face -- with its positive and negative expressions -- in a corner of the display. If the usability testing equipment is installed in a permanent testing facility, observers may be able to work behind one-way mirrors so that the tester isn't distracted by the feeling of being watched.

People often assume that usability testing costs a great deal of money. Building a permanent usability-testing lab can cost tens or even hundreds of thousands of dollars, and even a portable lab can cost many thousands of dollars. But if cost is an issue, skip the technology and focus on what's important: a representative set of testers, some carefully crafted test scenarios, and a watchful, objective observer or two.

You don't have to expend a great deal of time on usability testing either. If time is an issue, examine only the parts of the design you're least confident about. Note potential trouble spots as the project is coming together. If you design these problem areas incorrectly, they will cost you more in the long run in redesign, recoding, and retesting. Making sure of them early in the process will save the cost of rework late in development or during the next release cycle.

Usability tests do a great job of showing what's not working in a design, but don't get caught in the trap of asking testers to suggest design improvements. Testers are not designers, and their answers will be wrong. Use the test results to expose bugs, but don't go overboard. If you derive the design from testing, you'll see small improvements, but these amendments will come at the expense of innovation and creativity.

Would a usability test have prevented my difficulty in finding garden.com's check-out link? Probably. Is completing the purchase an important task for this Web site? Absolutely. Would others have similar problems figuring out how to check out? If so, then a carefully crafted test scenario would certainly have caught the problem.

Home Articles We've Written Usability Testing