Usability testing is the most powerful, most effective thing you can do for your website.

That's a big statement to make, and people often want to dismiss usability testing because they're under the assumption that it's too complicated and expensive.

Some organizations conduct usability testing before launching a redesign, but many skip that step altogether, citing lack of time, lack of money, or both. But usability testing doesn't have to be that complicated, and almost anyone can do it—from start to finish—in about three weeks, according to Steve Krug, author of Don't Make Me Think and Rocket Surgery Made Easy.

Krug lays out a simple, inexpensive formula for conducting your own usability tests. His underlying premise is the following: If you watch people use your site and hear them think out loud as they're using it, you'll figure out very quickly where people are getting hung up.

Basic Ingredients

Here's what you'll need to create an effective, ongoing usability-testing program that won't kill your budget:

• One morning per month
• Three volunteers and some sort of incentive for participation (e.g., gift certificate)
• A list of core tasks that are consistent with the business goals of your website, and a script to coach them through the process
• A conference room, a webcam, and some inexpensive software
• A group of people from your organization who are involved in your website content, design, and functionality

That's it. No laboratory. No two-way glass. No expensive moderator.

Week 1: Plan, Recruit, and Build Your Team

1. Decide what you're testing

Are you testing your live site, a prototype, or just wireframes? Many people who conduct usability testing for the first time are testing their live sites.

2. Identify the tasks you want to test

Identify at least 10 (but no more than 15) tasks that your users should be able to successfully execute to get the most out of your site. Choose tasks that actual users would engage in, such as buying a specific product in a specific color and size, downloading a whitepaper, learning about what your organization does, or finding a phone number.

3. Recruit participants

Warning! Many companies get too hung up on the process of recruiting participants. Though using some people from your target audience is good practice, you don't have to worry about that nearly as much as you'd think (especially if you're just starting out), according to Krug.

"What you're looking for first of all are the problems that anybody is going to encounter, such as a confusing interface," says Krug. "It doesn't matter whether they're from your target audience or not. Your grandmother could try using the site and she would run into those problems."

4. Get your team together

Who from your organization should be involved in the tests? A good starting point would be to remember who has had strong opinions about the website in the past. An added benefit of usability testing is that watching people struggle to use your site is an extremely effective way to solve internal power struggles.

Week 2: Refine Your Plan

1. Get feedback from your team on the list of tasks

Make sure everyone has a say about which tasks are the highest priorities.

2. Write a script

You'll use your script when working with each participant. Krug provides a sample script (and other helpful resources) on his website. Once you finalize your tasks, adapt your script to include your specific tasks. But doing so is easy, and it can be done the following week.

3. Nail down logistics

How will you conduct the testing? Will participants come to your office, or will they work remotely? Either option would work if you have the right software. If participants are coming to your office, set up a webcam on the testing computer so that your team (who will be assembled in the conference room down the hall) can watch the volunteer and hear her comments.

4. Install screen-sharing software

Screen-sharing software will enable your team to watch as volunteers click through the site. GoToMeeting is also a good solution, especially if you run your tests remotely. GoToMeeting will allow you to not only tune in live but also record the entire session, including the dialog between the test moderator and volunteer and the details of the participant's navigation through your site.

Week 3: Execution and Debriefing

1. Bring in three volunteers for one-hour shifts each

Don't worry that a sample size of three isn't statistically valid. Usability experts who have done testing for years, including Krug, agree that if you watch three people use your site, you'll discover most of the site's serious problems.

If you schedule more than three participants, by the time you get to the fifth person you'll see the same problems arise again and again. Moderate each participant's session exactly the same way, following your script verbatim—even if it feels weird to do so. Krug cautions against ad-libbing, because you could inadvertently provide different information to the participants that could affect the results of your test.

2. Assemble your team in a conference room to watch

The importance of this step can't be overstated. Considering everyone's busy schedule, you may be tempted to have one person conduct the tests and report the findings to the rest of the team. Resist that temptation! Seeing is believing, and people won't doubt the validity of the findings if they watch users struggle to accomplish seemingly simple tasks.

3. Take (silent) notes

At the end of each session, ask the observers to write down the three most significant usability issues they observed. You shouldn't have any discussions because you want to be able to compare observations that are untainted by group influence.

4. Order a tasty lunch, and debrief

Ask your team members to discuss what they observed during the three sessions, and look for common themes. Together, decide on the top three themes; then, make a list of the most serious problems in descending order. That last point is critical.

Easy-to-fix issues can easily distract you and bog you down. But remember the reason for conducting usability testing yourself: you're resource-constrained. Stay focused on the top issues that will make the most measurable impact on your site's usability.

* * *

By now, you're probably wondering whether conducting your own usability testing could really be this fast, easy, and inexpensive. The answer is—unquestionably—yes! You really can get meaningful, insightful, and applicable user feedback in about four hours per month, at very minimal expense.

The most difficult part of the process is shedding your ego long enough to value, rather than dismiss, what you learn. Remember that you and your team members are blinded by what you already know about your site, making it nearly impossible to comprehend why your visitors can't see the gigantic red button you want them to click, or why they don't understand that the cute magnifying glass icon means "search."

Certain things about your website that are obvious to you, either because you thought of them or became used to them over time, may not be quite as apparent to others. And no matter how robust your analytics package is, the only way you'll learn about those disconnects is by watching real human beings use your site.

The magic of usability testing is that it ensures that a website meets the needs of your visitors—which will, ultimately, also ensure that it meets the business goals of your organization.