User testing seems to have this thing about it that it's something that’s tough to pull off. It’s essential to have a lot of equipment such as eye-tracking cameras, mouse tracking software, speech recording and someone wearing a lab coat. Not true.
User-testing shouldn't make you think it isn't worth the set-up or investing in; it should be the opposite. You need to user test often; it should always be part of your product release cycle.
Let me tell you a myth about user testing.
"The more users you test with, the more you'll learn about the challenge you're trying to solve."
According to the largest UX research group in the world - Nielsen Norman (If you haven't heard of either Mr Nielsen or Mr Norman, I recommend you read any of their books based around user research study) the optimal number of people to user test on is 5!
Let's take a look at their findings any why this is the case.
So what does this show? Well for starters we can see that 0 testers equal nothing learned.
Already by having tested with one single user, you have found out a third of what there is to find.
By the time you get to 3 that has shot to just below three-quarters of what there is to discover. The pattern the group found is that the more and more testers you place, the more times you will get results back of things you already knew.
After the fifth user, you start to get so much repetition. There isn't much point in investing in more users for that particular test script.
"But hang on Mr Blogger doesn't that graph go to 100% !?". Yes, the pot of gold sits there about 15 tests in.
So why test with 5? Well, the main reason is that it is advised to spread your testing budget. You don't need to validate a failing point in a user test flow multiple times, instead spread out 15 tests over 3 separate scenarios.
You now have your feedback, and some action points to go back to the drawing board with. Tweak your design as necessary and then … test again! This is important because while you think you might have the skills to knock it out of the park after one round of testing, no one can design the perfect interface. Your mark 2 will now iron out the minor 15% of problems not found in the first round, leaving you with just 2% of problems.
One last thing, while testing with one user looks like it has the best return on investment, it’s best not to trust the instincts on the general behaviour of one single user. The ability to compare is critical here. Three users will begin to give you patterns.
When we run design sprints, we use the industry-leading testing tools. This gives us the ability to always test with the true customer base for the client. We believe that users who are in the comfort of their usual surroundings give more realistic feedback to tests and are more honest about critiquing the solution in front of them.
The golden rule is that it's always better to do some user testing than no user testing.
Find out about how our Design Sprint process can help you validate ideas faster by using user testing.