Congrats to my buds at Yahoo for winning ERE’s prestigious 2009 award for best corporate careers website last week. This accomplishment is particularly impressive in light of the type of questions they had to answer as part of the evaluation process. “How has the site has paid off or contributed to improved profits, better employees, and other quantifiable outcomes?” This was no beauty contest; it was about hardcore metrics and making a measurable difference in the greater business strategy.
Since I wasn’t part of the judging process (maybe next year), I don’t have their answers, but as a researcher, I do have lots of questions and so should you. Your company’s career website is the hub for all applicant traffic and you should be asking yourselves the hard questions about how it’s measuring up.
That’s where Web usability testing comes in handy. As the name implies, these are studies that enable you to evaluate your career site by testing it on users, employees, or even yourself.
In a study conducted almost exactly one year ago by Forrester Research entitled “Best and Worst of Career Web Sites,” the 12 websites examined all received failing scores due to poor performance and usability. Common problems included missing content and functions, flawed navigation flows, illegible text, and poor use of space. Do any of those sound familiar?
Step one is admitting you have a problem.
Usability testing can measure just about anything, but here are the key factors you need to consider:
Efficiency: How many steps did it take the user to accomplish pre-specified tasks (did the user develop carpal tunnel or fall asleep)? Was there a call to action? Was it easy to respond?
Performance: Did the user make any mistakes, and could they easily recover from them?
Recall: What information did the user remember afterwards. Does it communicate important information clearly and accurately?
Emotional connection: Would the user recommend the site to a friend? Is the site on-brand (i.e. did the experience match the expectation)? Did it make the user smile or cry?
After my last move, I set up the kitchen and called my kids in: “Quick! find me a bowl and spoon.” If they went for the right cabinet first, I knew I had successfully unpacked. They checked the dishwasher. Try again.
Your usability study will be a lot like that. Pick out a range of employees from various disciplines within your company and watch them perform 7 to 10 pre-determined tasks based on various scenarios. Be prepared with a pen and paper to take notes, and have a stopwatch ready so that you can measure time-on-task. Remember to reassure participants that their jobs are not on the line as you stand behind them with a stopwatch, or you’ll be witness to some rather erratic web browsing and sweaty palms. You won’t need to test it with more than five users. The first test will probably tell you 35% of all you need to know.
Task 1: First Impressions
Task 2: 5 minutes of free exploration (where did they go first, second, and third?)
Task 3: Learn about employee benefits (timed task)
Task 4: Apply for a job (timed task)
Task 5: Replace upload forms
Task 6: Give them a place to go (i.e. “job description”) and check the path they took
Task 7: Return Exploration: let them go anywhere they felt confused about or wanted to revisit. Capture the stops.
Save time at the end for some open-ended discussion on recall, branding and overall user experience. “What did you like best?” “What would you change?” While you’re at it, have some employees perform this test on one of your competitor’s sites. Use that as a benchmark for your site as you track it over time.
If the results are eye-opening (in a bad way), then it’s time for a sit-down with the head of your IT team. Be sure to come armed with the results of your test and a positive attitude, or you may experience some resistance. Have an open discussion about how to make usability improvements and offer your assistance in finding the right vendor that specializes in this work. Worse comes to worst, have your IT manager participate in your next usability study.