Usability Testing
Watching real users attempt tasks to find where a design breaks down.
Usability testing puts a product (or prototype) in front of real users and observes them attempting realistic tasks. It surfaces confusion, dead ends, and friction that the team — too close to the work — can no longer see.
Even five users will reveal the majority of major usability problems, which is why lightweight, frequent testing beats one large study at the end.
Key characteristics
- Watches real users attempt real tasks while thinking aloud.
- Can be moderated (live, with a facilitator) or unmoderated (recorded remotely).
- Reveals what people actually do, which often differs from what they say in surveys.
- Around five participants per round catches most major issues.
Example
Testing a checkout, three of five users miss the promo-code field because it looks like body text — a quick visual fix recovered measurable revenue.
Frequently asked questions
How is usability testing different from A/B testing?
Usability testing is qualitative — you observe a few users to learn why something is hard. A/B testing is quantitative — you serve variants to many users to learn which performs better. They answer different questions and work well together.
Do you need a finished product to run a usability test?
No. You can test sketches, wireframes or prototypes. Testing earlier and cheaper is usually better because changes cost less before code.
Related terms