How easy is your website to use? An organizational structure that might make perfect sense to the Web team might baffle a large percentage of the site's users. For an objective opinion, people outside of the organization should be recruited to review the site for usability issues. Fortunately, this type of testing does not have to be complicated or expensive. The following tips originated from Steve Krug, author of Don't Make Me Think: A Common Sense Approach for Web Usability.
1. The size of the audience sample doesn't have to be large.
The good news is that, unlike surveys, a statistically representative sample is not necessary to get good, usable information. In fact, interviewing just three or four people might be all that is needed, since most of the time the results are the same if 100 or 10 testers are used—what works well on the site works, and what doesn't is usually frustrating for most visitors.
2. Testers don't have to be members of your target audience.
Usability testing gets at the intuitive nature of your site. Having members of the target audience test the site is great but not necessary. It is suggested that people participating in the testing do so alone (as opposed to in a focus-group setting), that testing be performed once per month, and that different people be used each time.
3. The goal is to get testers to use the site and have them talk to you while doing so.
Written surveys aren't recommended. Talk to the testers before they begin, asking questions such as how often they use the Internet and what their favorite sites are. This will tell you about how comfortable respondents are using the Web, establishes that you are in the listening mode, and gets participants accustomed to talking about their thoughts.
4. Start by having respondents talk about your site's front page.
What do they think the main focus of this site is? What do they see as the most important components of the site? The crucial tasks? Pay attention to where your respondents travel first and what they say. The responses should line up with what the organization has determined are the site's most important components.
5. Ask respondents to find something on the site or use one of its components.
Don't make this a difficult task, since the goal is to see how easy the site is to use, not to stump the testers.
6. Have as many people as possible see the test.
It is amazing how obvious and prevalent some problems can be to everyone except management and the people responsible for the site. Don't write a long report. Actually seeing a few usability tests is much more powerful.
7. Pick the worst two or three problems to fix first.
It's too easy to go after the low-hanging fruit and put off tackling the tough tasks. Decide what the biggest problems are and put your energies there. And don't overfix the problem. Often the simple fixes are the best.
To give feedback on the National Oceanic and AtmosphericAdministrationCoastalServicesCenter's website (www.csc.noaa.gov), please contact Sean Ryan at Sean.Ryan@noaa.gov.
