---
name: usability-testing
description: Use when asked to run a usability test — observing real users attempting real tasks to find friction points — the core user-research method within UI/UX testing (see ui-ux-testing) and user-centred design (see user-centred-design).
---

# Usability Testing

Usability testing evaluates how usable and user-friendly a product,
system, or interface actually is by observing real users interacting
with it — identifying usability issues, obstacles, and improvement areas
directly from observed behavior rather than from assumption.

## The process

1. **Define objectives** — what specifically is being evaluated.
2. **Plan scenarios** — realistic tasks representative of actual use.
3. **Recruit participants** — representative of the real target user
   base, not a convenience sample (see [[user-centred-design]]'s
   identical warning about recruiting representative research
   participants).
4. **Conduct tests** — observe participants attempting the tasks.
5. **Analyze findings** — identify patterns across observed sessions,
   not just isolated one-off comments.
6. **Create recommendations** — concrete, prioritized design changes.
7. **Iterate** — usability testing is a repeated cycle, not a one-time
   event; see [[user-centred-design]]'s iterative design/test/
   refine loop.

## Why it's worth the effort

- **Enhanced user experience** — directly addressing observed friction
  improves satisfaction, task completion, and error rates.
- **Issue identification** — surfaces problems invisible during
  development, because developers already know how their own interface
  is "supposed" to work in a way real first-time users don't.
- **Data-driven decisions** — replaces design arguments settled by
  seniority or opinion with empirical evidence of how real users actually
  behave.
- **Cost and time savings** — catching usability issues early avoids
  expensive redesigns and updates after launch.

## Common pitfalls

- **Recruiting convenient, unrepresentative participants** (colleagues,
  friends) — produces confident-sounding but misleading results, since
  they don't share the target users' actual context or naivety about the
  product.
- **Leading participants during the session** — a moderator hinting at
  the "correct" path defeats the point of observing genuine, unprompted
  behavior.
- **Testing once and never again** — usability testing validates one
  specific design snapshot; it doesn't automatically validate whatever
  changes come after it.
- **Skipping the "create recommendations" and "iterate" steps** —
  observing issues without acting on them wastes the research effort
  entirely.

## Learn more

- [[user-centred-design]] for the broader iterative design methodology usability testing is a core part of.
- [[ui-ux-testing]] for the wider set of complementary evaluation techniques (surveys, heatmaps, eye-tracking) alongside usability testing.