The use of usability tests has increased enormously in recent years. There are many types of tests that can be done, all aimed at improving the user experience or increasing conversion. But how many users do you actually need to conduct a successful usability test? According to UX research agency Nielsen Norman Group the ideal number is 5 users.

More participants do not yield more results (but higher costs)

The results that Nielsen presents are based on their own practical experiences and data from 83 usability tests that they have carried out for their customers. The graph below shows the tests performed, with the number of insights/usability problems found on the Y-axis and the number of participants involved in the test on the X-axis.

Credit: NNGroup – https://www.nngroup.com/articles/how-many-test-users/

There is a very small correlation between the number of participants and the number of usability problems found, but it is very small. The optimal number of participants is 5. More than 5 participants do not necessarily provide you with more insights, while the costs for the usability test do increase significantly with more participants. With 5 participants you will find just as many insights as with more participants, but at a fraction of the costs.

Is there more budget for usability testing? Then run multiple tests that focus on a different specific problem per test. This way you use your budget more intelligently than if you let many participants participate in one test.

Other tests require different numbers

The usability test as described above is qualitative in nature. Other tests have a more quantitative nature and therefore 5 participants are not enough. For example, for eye tracking the optimal number of participants is 39, card sorting requires at least 15 participants and A / B tests sometimes require at least 100 participants.