Five users is enough to change your roadmap

The most common reason teams skip usability testing is that they picture it as an expensive, formal exercise: a lab, a recruiting agency, a two-week turnaround, a report nobody reads. So it gets pushed to next quarter, permanently.

The version that actually works is much smaller. Five participants, a clickable prototype, three tasks, half an hour each. You can run it inside a single week without pausing the build.

What to do in the session

Give the participant a task, not a tour. 'You want to send £50 to a friend' — then stop talking. The instinct to help is overwhelming and it destroys the data. Every time you rescue someone, you delete the finding you came for.

• Write three task scenarios, phrased as goals rather than instructions.
• Ask what they expect to happen before they tap, then let them tap.
• Count hesitations — they mark the same friction as failures, more quietly.
• Record the session so you can quote it later instead of paraphrasing.
• Invite one stakeholder to watch live. It persuades more than any report.

Turning sessions into decisions

Afterwards, list every problem observed and rate it on two axes: how many participants hit it, and how badly it blocked them. Something that stopped four of five people is not feedback, it is a defect. Something one person disliked is a preference — note it and move on.

Then write the findings as changes, not observations. Not 'users were confused by the filter panel' but 'move the three most-used filters onto the results screen and collapse the rest'. A finding a team cannot act on is a finding that gets archived.