The five-user rule belongs to formative usability testing, not surveys or inference
Aliases: five-user rule · small-sample usability testing · discovery-rate model
What it is
The five-user rule is a simplified heuristic for iterative problem discovery in formative usability testing, not a universal research minimum. It cannot estimate proportions, mean differences, associations, or population preference. Surveys and statistical tests require planning from effect size, precision, variance, and the intended model.
Why it happens
Small rounds create value through fix-and-retest cycles. Common, easily triggered problems may recur among early participants, so marginal discovery declines. But detecting a problem is not estimating its prevalence, and discovery varies with tasks, heterogeneity, moderation, and severity. A fixed five assumes stable independent detection probabilities that rarely hold.
Studying it
For formative tests, report tasks, segments, issue definitions, and the curve of new serious findings by round. For estimation, use power or precision analysis with attrition, clustering, and multiplicity where relevant. Comparisons of five versus larger samples should separate cumulative discovery, severity coverage, and prevalence error.
Where it stops holding
Five people may reveal a critical barrier or reveal none; the number is neither a ceiling nor certification. Distinct user groups, tasks, and devices require coverage. Small exploratory samples can guide iteration but do not become adequate for surveys or hypothesis tests merely because all are called user research.
Applying it
- Identify whether the decision requires discovery, frequency estimation, or condition comparison.
- Run formative work in small iterations and continue according to new findings and decision risk.
- Size inferential studies from a meaningful effect or target interval width, not the five-user heuristic.
- Report uncovered users, tasks, and contexts; no new issue in one round is not proof of no issue.