It’s astonishing how many organizations never do usability testing on their websites or applications. And it’s funny the excuses you hear for not doing it. Let’s take a look the top 10.

1. “I don’t have a usability lab”

You don’t need a lab. In fact, you don’t even want one. Labs are unnatural settings and can make participants nervous. A more casual setting is better.

2. “I don’t have screen-recording software”

Both TechSmith (makers of Snagit, Morae and Camtasia) and Silverback for Mac (which costs only $69.95 for the full version) offer 30-day free trials. You can also use GoToMeeting.

Still too much cost or hassle? Then scrap the screen recording software and run tests without it. You’ll still learn something.

3. “I don’t know how to run a statistical analysis on the results”

There’s no need, and no point. Usability testing is qualitative in nature. You’re looking for insights, not statistics.

4. “I don’t have enough test subjects”

Since you’re not looking for statistics, you don’t need a lot of subjects. Test on 3 – 5 users. Even 1 or 2 is better than none.

5. “It’s too hard to find test subjects that match my customer demographics”

Then test on subjects that are “close enough”. The only thing that really matters is that you recruit “passionate users”. By this I mean people who might actually use a product like yours.

For example, if you’re developing a “shoot ‘em up” video game, test with people who play such games. Beyond that, don’t worry about demographics.

6. “My website isn’t finished yet”

Good! The earlier you start testing, the better. Test on prototypes. Test on wireframes. Test on napkin scribbles. If you wait until the product is finished, it’s too late: changes are too difficult and expensive.

7. “I don’t have the time”

I have a friend in print production. He always says, “It’s funny how clients never have time to print it right the first time, but they always find time to re-print it.”

Make the time. You don’t need much. At least do some remote, un-moderated tests using a service like You can have results back within an hour.

8. “I don’t have the budget”

Do-it-yourself testing is cheap (or even free if you’re imaginative.) is just $39 per test.

9. “I’ve never done it before. I have no idea how to do it.”

Learn. It’s easy. Get Steve Krug’s book, Rocket Surgery Made Easy: The Do-It-Yourself Guide to Finding and Fixing Usability Problems.

10. “My site is perfect. It doesn’t need testing”

Okay, I’ll admit I’ve never actually heard this excuse before. But that’s because everyone realizes their site isn’t perfect. Which means all sites need testing.

  • Neil B

    The cheapest usability testing is to create a culture in your company where people say what they don’t like in a way that can easily result in an actionable defect report. It’s also a good way to bring the staff closer to the product, which brings other cultural benefits too. No lab needed. Say “Open (or get someone to open) a defect report for that and we’ll see if we can fix it.” a lot! When an admin assistant / janitor / CFO in your company raises a defect report against your product and sees it fixed they will be beside themselves with happiness, and, since you’ve encouraged them to be critical like normal users, you’ve benefited.

  • Tim Moore

    I agree entirely with this sentiment it is in line with the old adage “measure twice cut once”. In addition, I recommend usability testing previous versions of your software and testing the software of your competitors (without divulging who you work for). You cannot get statistics but I recommend keeping figures on the numbers of different types of failure, complaints by category, times to learn how to use the software etc. Numbers help to see the software in a different light and allow you to set up “stop rules” to decide when an application is “good enough” to release. Iterative design is a good approach, especially if embedded in a process of continuous tracking of usability.

  • Ian Hamilton

    Neil, what you’re describing is not usability testing, as your colleagues bear absolutely no relation to the people using the product, they have expert domain knowledge, are highly familiar with the product, and presumably highly IT literate. It can be helpful to have fresh eyes on it, but that’s about the only benefit of what you’re describing.

    Michael, in general good stuff but I completely disagree with your point about demographics. Demographics ARE important. Critically important in fact. A far better answer to that argument would be ‘use external recruiters’. That’s a very cheap and easy way to get good results.

    • Michael Straker

      Thanks for your comments, Ian. Regarding demographics, I guess a more precise way to make my point would have been, “Don’t let the difficulty of finding ideal test subjects stop you from running tests.” In other words, don’t let the perfect be the enemy of the good.

      Yes, it’s always better to use carefully selected test subjects. This is especially true if you’re trying to determine how well your website resonates emotionally with users, or how persuasive the website is. But for identifying simple usability issues, a precise demographic/psychographic match — though always preferable — is less important.

      Using external recruiters is, of course, a great solution. But it adds considerable complexity and expense. If that additional complexity and expense tempts you to abandon the idea of testing, stop and think again.

      Better to relax your recruiting criteria than to abandon testing altogether. That’s all I meant.