← Community

Run a usability test

Watch five people who actually do the job try real tasks, and change the one thing that stalled them — instead of interviewing them about taste. Use when a flow is new, before a redesign ships, or when opinions about a screen will not converge.

---
name: run-a-usability-test
category: design
description: >-
  Watch five people who actually do the job try real tasks, and change the
  one thing that stalled them — instead of interviewing them about taste. Use
  when a flow is new, before a redesign ships, or when opinions about a
  screen will not converge.
---

# Run a Usability Test

A critique tells you what designers think. An analytics dashboard tells you
where people left. Neither tells you *why* they stalled. That takes watching
someone who did not build it try to do the job.

Five people, real tasks, you keep your mouth shut. That is the whole method.
Ten more people will mostly repeat the first five; a survey of fifty will
tell you they "would use it" and teach you nothing.

## Recruit the people who do the job

Not your teammates, not your mum unless she is the user, not a Slack channel
of power users who already know the workarounds.

The filter is: **have they done this job in the last month, without you
standing over them?** If the product is for first-time admins, test first-time
admins. A veteran who "gets it" will complete the task and leave you with a
false green.

Five is enough to find the stall that matters. After three, you usually know
the shape; four and five confirm it is not one person's quirk. If those five
disagree with each other, you have two user groups, not a sample-size problem.

Pay them. Unpaid tests fill with people who like you, which is a different
population from the people who have to use this on a Thursday.

Shared publicly by the author and open for any use. See the Terms.