What is a usability test? Definition for product research
A usability test is a session where a researcher watches a real person attempt a task on the product, recording where they hesitate, get confused, or fail.
A usability test is a session where a researcher watches a real person attempt a task on the product, recording where they hesitate, get confused, or fail.
The motivation is irreplaceable. Designers and PMs are too close to the product to see it as a first-time user does; the team's understanding of the obvious is not the same as the new user's understanding of the obvious. A short test against five real users almost always surfaces things the team missed and would have shipped.
What the output looks like
The raw output is recordings of each session. The useful output is a synthesis: where did multiple people get stuck, what assumption did the design rely on that users did not share, which task did most people complete and which did most fail.
A short visual-feedback session is often a useful intermediate format. The researcher captures the moments that stood out from each test, adds dictated notes, and ships the result as a single document. The PM, designer, and engineer all read the same artifact and converge on the same understanding of where the product needs to change.
Frequently asked questions
How many participants does a usability test need?
Five is the most-cited number for catching the obvious problems on a flow. Add more if the audience is highly segmented or if quantitative measures are involved.
Moderated or unmoderated?
Moderated tests catch more — the researcher can probe at the moments that look interesting. Unmoderated tests scale further and are cheaper per session. Most teams use both.
Capture your first review.
About a minute from open tab to a shareable URL your agent can ingest.
Start capturing