40 · Researching
The Five-User Rule
Testing with five people finds most of a product's usability problems. Most, and sometimes far fewer.
Evidence: Holds, with limitsWhat we do with it
- 01
Five people a round is a good start. Several small rounds beat one big one, because each fix gets tested too.
- 02
Five isn't enough when problems are rare, when users differ a lot from each other, or when you need a number to report.
- 03
Test each distinct group of users separately. Five buyers and five admins are two studies.
Examples
Getting it wrong, getting it right.
Don't
One test with five people, then launch, with “we found 80% of the problems” in the report.
Do
Three rounds of five, fixing between rounds, and a plain note on what the study can't tell you.
The evidence
What the research says.
Laura Faulkner tested it in 2003 with 60 users, drawing random sets from the whole. Sets of five found 85% of the problems on average, but some found only 55% and others 99%. Every set of ten found at least 80%, and every set of twenty at least 95%. Wonil Hwang and Gavriel Salvendy's 2010 review of published studies put the number needed to find 80% of the problems at 10, give or take 2.
How our view changed
We plan rounds of five to find problems, and bigger groups when a decision rests on a number.
Sources
- Virzi, R. A. (1992). Refining the test phase of usability evaluation: How many subjects is enough? Human Factors, 34(4).
- Nielsen, J., & Landauer, T. K. (1993). A mathematical model of the finding of usability problems. INTERCHI '93.
- Faulkner, L. (2003). Beyond the five-user assumption: Benefits of increased sample sizes in usability testing. Behavior Research Methods, Instruments, & Computers, 35(3).
- Hwang, W., & Salvendy, G. (2010). Number of people required for usability evaluation: The 10±2 rule. Communications of the ACM, 53(5).
Last reviewed November 7, 2026. How we keep this current