Small user tests

A common misconception about user tests is that it always requires a lot of participants to get valuable feedback about the design. In everyday practice however asking seven to eleven persons can already reveal the biggest problems in a design. Here is the proof why. The reason why we need only so few persons and already get valuable data is because we are looking for the most common problems in an early stage in productdesign. What we want to find out is what are the biggest issues with our new product design at an early stage possible. To do this we can already suffice with a small amount of participants. The results from this test will not help you to tackle all problems but it will help you to find the most common problems in this early stage.


An analogy, say in a fishpool you see mainly orange-white spotted koi-fish, despite the fact that there are full-orange and full-white koi fish existing in nature as well. Without overseeing the entire fishpool, it would be reasonable to assume that most of the other fish in this pool are also similarly orange-white spotted, right?*. This analogy works as well in product, UI or service design: the first seven to eleven people we ask to give feedback during a test are most valuable. Quite likely the last three people we interview will merely repeat and reconfirm what has been said already or maybe give one more new insight if we are lucky.

* note: provided the fish pond consists of a homogeneous group of fish


The number of participants required during a test to get a certain percentage of usability problems occurring when using a product has been studied in more detail by H. Kanis How many participants - Departement of Industrial Design, Delft, Netherlands. The article questioned the initial rule-of-thumb of using only five people for trials, leading to the conclusion that the rule-of-thumb of five might have been too optimistic. Based on practical verification it can be concluded that using seven to eleven people covers around 80% of all usability problems. The big benefit of verifying our design like this, is that it only requires a few days to a week to test our product, UI or service and gives a lot of valuable data.