The discussion about how many users is enough for a usability test has been going on for years. Research by Jakob Nielsen and Tom Landauer showing that tests with five users will reveal an average of 85% usability problems has been seen as a proof that five is enough.

According to Laura Faulker, Nielsen and Laudauer's prediction is right. Five users will reveal 85% usability problems - on average. In a study, she found that problems found with five users range from nearly 100% down to only 55%. Thus, relying on a single set of five users, we run the risk that nearly half the problems could be missed.

Dr. Eric Schaffer concludes that "…for a routine usability test run 12 people for each segment. For an important one where the stakes are high run 30. If resources are really tight, you can drop to five-six per segment, but this is bad."

Links:

• Enough is enough... but five probably isn't
• Faulkners paper Beyond the five-user assumption
• Jakob Nielsen's article Why You Only Need to Test With 5 Users

Henrik Olsen - July 01, 2004

See also: Usability testing (34)