Usability test

In a usability test, real people try to complete tasks on your website or app while you observe where they hesitate, get confused or give up. A task could be "find the price of the basic plan" or "book an appointment for next Tuesday".

How it works

A typical usability test has a handful of participants, a few realistic tasks and an observer. Participants are often asked to think aloud, so you hear what they expect and why they click where they click. Sessions can be moderated, with a researcher present, or unmoderated, where participants record themselves.

How many participants?

In a well-known article from 2000, Jakob Nielsen argued that testing with five users finds about 85% of the usability problems, and that several small rounds beat one large study (Nielsen, 2000). He adds a condition: this holds for comparable users. For different user groups, you test with a few people from each group.

Small numbers work because a usability test looks for problems: once three people get stuck on the same button, you don't need a fourth. Measuring a score, such as how much people trust a page, needs more people, because you want an average. See how many votes you need.

Usability test or first-impression test?

• A usability test asks: can people do what they came to do? It needs time and a working site, and gives detailed insight.
• A first-impression test asks: does the first screen convince people to stay? It takes seconds per person and gives comparable scores.

The first-impression test comes first: if visitors leave within seconds, they never get to the tasks. A five-second test is a related quick method.