# Usability Testing

Observing representative participants attempt realistic tasks to identify where an interface supports or obstructs successful use.

- Type: method
- Domain: Research & behavior
- Level: Foundational
- Last reviewed: 2026-09-21
- Original: [Soroush’s Product Wiki](<https://wiki.soroush.uk/topic/usability-testing/>)

## What it means

Usability testing evaluates a product or prototype by asking representative participants to attempt realistic tasks while researchers observe. Sessions may be moderated or unmoderated, remote or in person, and qualitative or quantitative. The central evidence is behavior: where people hesitate, misunderstand, recover, fail, or succeed—not whether they say they like the design.

## Why it matters

A screen can look polished and still make essential tasks difficult. Usability testing exposes breakdowns before they become support tickets, abandonment, or safety risks. It gives teams concrete evidence for prioritizing changes and helps distinguish an interface problem from a training, content, policy, or expectation problem.

## When to use it

Use it throughout design: on sketches to compare interaction approaches, on prototypes to repair flows before development, and on live products to investigate persistent friction. It is especially valuable before launching high-frequency, high-risk, or business-critical tasks.

## How it works

1. Identify target users, critical tasks, research questions, and the fidelity needed to answer them.
2. Recruit participants whose experience and constraints match the intended audience.
3. Write neutral task scenarios with plausible goals, success criteria, and no hidden instructions.
4. Observe task behavior, ask participants to think aloud when appropriate, and avoid rescuing them too quickly.
5. Synthesize issues by evidence, severity, frequency, and impact; revise the design and test again.

## A practical example

A banking team tests a prototype for adding a transfer recipient. Four of six participants search under “Payments” while the feature sits under “Contacts,” and two abandon the task after a security warning appears to erase progress. The team moves the entry point, preserves form data through verification, and retests the revised flow before engineering handoff.

## Where it breaks

A test session is an artificial setting, and think-aloud narration can change behavior. Small qualitative studies find important issues but do not estimate their population frequency reliably. Prototype gaps may create false problems, facilitator behavior can bias outcomes, and testing cannot by itself establish product desirability, accessibility conformance, or long-term adoption.

## Common mistakes

- Testing vague opinions instead of giving participants realistic tasks.
- Explaining the interface or leading participants as soon as they hesitate.
- Reporting every stumble equally without judging task impact and evidence strength.

## Visualization

[View the visualization on the website](<https://wiki.soroush.uk/topic/usability-testing/#visualization-usability-testing>)

## Sources

1. [Usability \(User\) Testing 101](<https://www.nngroup.com/articles/usability-testing-101/>) — Nielsen Norman Group · Professional guidance
2. [Using moderated usability testing](<https://www.gov.uk/service-manual/user-research/using-moderated-usability-testing>) — UK Government Service Manual · Government guidance

## Related topics

- [First-Click Testing](<https://wiki.soroush.uk/topic/first-click-testing/>)
- [Cognitive Walkthrough](<https://wiki.soroush.uk/topic/cognitive-walkthrough/>)
- [Heuristic Evaluation](<https://wiki.soroush.uk/topic/heuristic-evaluation/>)
