Best Time to Test Usability
Author of this post: Mike Madaio | About Blog Authors »>
You asked your usability questions, and Mike has answered! We invite you to ask more questions here

Hi Mike,
I work for a small web design company, and have been arguing with one of the designers about when we should do usability tests for our projects. I think we should do them close to the beginning, after we have a few main pages done, so we can make changes if there are big usability problems. But he thinks we should wait until the site is considered final so the test covers everything. Which is better? (Keep in mind, I owe him lunch if I’m wrong.)
Best,
Jennifer
Hi Jennifer,
You’re actually both right. And both wrong! ;o)
Unfortunately, there isn’t a magic time to do usability testing that will return a perfect result. You should test as often as you can within the restrictions of your budget and the time frame of your project, making sure to have at least 1-3 rounds of testing throughout the process, depending on project size.
Great times to test include:
1. Before the project starts: This will allow you to determine problem areas, or where your next project should focus.
2. Concepts: User paper prototypes or lo-fi mockups to determine whether users understand your designs.
3. Working Prototype: Once you have a working prototype, you can get more realistic tests and scenarios. However, if you wait until this point to test for the first time, you might miss something big and have to start over from scratch!
4. After Launch: The best test results often come once the product is launched with real data and real scenarios, but we can’t wait this long to find out about major problems. Still, testing after launch will help you get ready for the next project (See #1!)
If you can only afford to test once or twice, going with lo-fi conceptual mock-ups, either paper or HTML (I prefer HTML — it creates a more realistic experience) is your best bet. The advantage to testing at this point is that if your design fails, you still have time to iterate and test again, often several times until you get it right. Though this can delay the project, it is far better to spend a little time up front ensuring your design is right than to spend a lot of time fixing problems once you have implemented.
Anyway, I hope this helped (and that you got a free lunch out of it!)














