Some people wonder why we only test with 5 users.
Our founder, Robin Dhanwani, uses an interesting study from the Nielsen Norman Group to explain how five users is all you need to get most of your feedback.
Someone asked me recently, why do we just test with 5 people in a Design Sprint?
Is there enough value in getting the data that comes out of testing something with 5 people? Or is it just too [little] to rely on any decisions based on that data?
What it says is that testing something with 5 people gives you 80% of the observations you might get. Anything more than 5 people — the results are very incremental.
Let me explain why that happens.
If you're testing with a unique cohort of people on a unique user journey (that's what we're talking about when we refer to running a Design Sprint), the first user will tell you something interesting that you did not know.
The second user might tell you one more [piece of] information.
By the time you start reaching the fifth person, you will start noticing that everything that they're talking about from the product testing has been already said by somebody or the other before.
You're not getting anything new anymore. You really start understanding the psyche of people and how they are interacting with your product.
By the time you hit 5 users, you should have 80% of your insights already there.
What will be of more value [than testing more people] is to do it very quickly, so that you can do it more often.
You stop at testing with 5 people, go back to your product design, see how you can make it better, solve the problems that came up in the testing, and maybe repeat the process with 5 users again. That way, the insights that you get will be compounded, and you will also have a better version of the product already.
That's why we think testing with 5 users is of immense value, which can help you make the product better.
We're constantly making videos on product design and sharing our experiences of running these tests and running Design Sprints for clients.
If you have any questions around product design, user testing, or anything that you would like us to talk about, do let us know in your comments.