Purpose
When a product is built but people get stuck on tasks, usability testing shows the friction at concrete points. Observed use instead of self-report makes visible where orientation, understanding, or interaction breaks down.
UX Research Research method
Usability Testing
Observes how people complete concrete tasks with a product.
Usability testing examines how people complete a specific flow in a realistic use situation. Observed actions and obstacles identify opportunities for improvement. Findings apply to the use situation examined.
Can relevant users complete their important tasks with the current design on their own?
- Participants
- Based on research question
- Time
- Depends on project
The phase bar groups preparation, sessions, and analysis.
Frame
Learning goal and flow
Choose a concrete task and the relevant interface.
Prepare
Tasks and setup
Prepare scenarios, recruitment, session flow, and an observation grid.
Observe
Run sessions
Let the participant work through tasks and intervene only when needed.
Synthesize
Synthesize issues
Compare evidence across tasks and sessions, then derive testable changes.
Run a bounded test round, cluster issues, change the design, and check important changes with new or appropriately selected participants.
A good fit when
Choose this method when an existing or prototyped flow contains testable tasks and you need to see where people hesitate, fail, or take detours.
Ideal for
- Prototype review
- Pre-launch validation
- Task-flow debugging
Not good for
- Opportunity exploration
- Market sizing
Outputs
Observed behavior
Task-level actions and obstacles.
Checked changes
A justified next design and focused follow-up tasks.
Test observation
Fictional teaching example
Keep task, behavior, evidence, and follow-up decision distinct.
| Observation row | Task | Observation | Evidence or context | Follow-up or change |
|---|---|---|---|---|
| Task observationRecord one observation from a concrete task. | Fictional task: Find a cordless drill for Saturday. | T1 compares the tool first, then separately searches for pickup time. | Fictional prototype version 0.4; availability and location appear on separate views. | Show location and pickup time together at reservation. |
| Review a findingSeparate observation, impact, and possible change. | Fictional task: Reserve a tool for a repair appointment. | T2 stops because no confirmed pickup window is visible. | One fictional individual case, not yet a pattern. | Check comprehension with more relevant participants. |
1. Task
- Task observation
- Fictional task: Find a cordless drill for Saturday.
- Review a finding
- Fictional task: Reserve a tool for a repair appointment.
2. Observation
- Task observation
- T1 compares the tool first, then separately searches for pickup time.
- Review a finding
- T2 stops because no confirmed pickup window is visible.
3. Evidence or context
- Task observation
- Fictional prototype version 0.4; availability and location appear on separate views.
- Review a finding
- One fictional individual case, not yet a pattern.
4. Follow-up or change
- Task observation
- Show location and pickup time together at reservation.
- Review a finding
- Check comprehension with more relevant participants.
Keep evidence and interpretation distinguishable.
First step
Write a learning question, bound the flow, and describe a real use goal.
Good to do first
Deep dive
How it works
Tasks are observed in a realistic interface while wrong turns, hesitation, and detours are recorded. The patterns of use produce a clear picture of which hurdles are structural and which changes have the biggest effect on success and speed.
In detail
Usability testing works because observed behavior in real tasks is far more reliable than self-reported ease or clarity. The method becomes strong when tasks are realistic, success criteria are unambiguous, and observations are cleanly separated from interpretation. Hesitation, wrong paths, self-corrections, and verbal cues carry the weight, so concrete friction points can be translated into prioritizable improvements.
Facilitation
Prepare a realistic interface, credible tasks, and a clear observation structure so the session stays calm and focused. The decisive moment lies in quiet irritation, when people don't fail but visibly struggle; important friction often hides there. The work closes well with a brief pattern analysis, separating obvious quick wins from deeper design problems.
The method guide walks execution through 4 phases, timing, 4 pitfalls, and clear stop criteria.
Open method guideYour perspective
Does this method help with your work?
Your feedback helps shape the next improvements to methodatlas.
Sources and topics
Topics
Sources
Usability (User) Testing 101 - NN/Gnngroup.comTask-based observation with relevant participants, realistic tasks, neutral facilitation, and evidence-led analysis.Turn User Goals into Task Scenarios for Usability Testing - NN/Gnngroup.comScenarios should invite realistic action without revealing the interface path or desired answer.User research in discovery - GOV.UK Service Manualgov.ukInclude a broad range of users and access needs; research helps identify user needs and barriers.
Artifact templates
Usability Testing: WorksheetPlan tasks, individual sessions, and shared analysis.markdown
Usability Testing: WorksheetPlan tasks, individual sessions, and shared analysis.
markdownusability-testing-working-template.md
Plan tasks, individual sessions, and shared analysis.
Usability Testing: Worksheet
Question
To be confirmed
Desired outcome
To be confirmed
Scope
One focused test round Time: Multi-session research initiative; scope recruitment, sessions, and analysis to the learning question and knowledge gaps.
Complete this worksheet on paper or in your own document during the work.
Preparation for this scope
Method setup
- Materials: Learning question, neutral task scenarios, observable prototype or product, consent information, and observation grid.
- Roles: A moderator guides tasks; an observer records behavior. Participants fit the use situation under study.
- Advance information: Set the learning question, flows, product version, recruitment criteria, and decision context.
- Overall time needed: About 45–90 minutes per session; plan preparation, recruitment, and analysis separately.
- Setup: Prepare the test object, clarify privacy and consent, check technology, and pilot moderation.
Prepare for the work steps
Frame: Learning goal and flow
- Write a learning question, bound the flow, and describe a real use goal.
Prepare: Tasks and setup
- Bounded test objective and task-level success criteria.
Observe: Run sessions
- Piloted test script and observation grid.
Synthesize: Synthesize issues
- Contextual observation notes with session code and test version.
Test observation
| Observation row | Task | Observation | Evidence or context | Follow-up or change |
|---|---|---|---|---|
| Task observation | ||||
| Review a finding |
- Task observation: Record one observation from a concrete task.
- Review a finding: Separate observation, impact, and possible change.
Blank worksheet: add your own tasks and observations.
Frame: Learning goal and flow
Expected artifact: Bounded test objective and task-level success criteria.
Bounded test objective and task-level success criteria.
Entry:
...
- The question names observable behavior; the test version supports the task without artificial help.
Prepare: Tasks and setup
Expected artifact: Piloted test script and observation grid.
Piloted test script and observation grid.
Entry:
...
- Scenarios reveal neither route nor desired answer; participation and recording are transparent.
Observe: Run sessions
Expected artifact: Contextual observation notes with session code and test version.
Contextual observation notes with session code and test version.
Entry:
...
- The moderator does not hint at a solution; help, stops, and technical failures are marked.
Synthesize: Synthesize issues
Expected artifact: Issue list with evidence, severity rationale, and retest decisions.
Issue list with evidence, severity rationale, and retest decisions.
Entry:
...
- Severity is justified; the qualitative sample is not reported as a representative rate.
Open questions and next steps
...
Method guide: https://methodatlas.meierhoff-systems.de/en/methods/usability-testing/run-sheet
Alternatives and neighbors
Fitting alternatives with decision guidance, followed by other related methods.
UX Research
First-click Testing
Instead of Usability Testing, when you want to use First-click Testing to isolate the first navigation decision from the rest of the flow.
UX Research
Tree Testing
Tests search paths in a navigation and shows where findability drops and labels or structure need sharpening.
UX Research
Card Sorting
Shows how people group and name content so content groups, labels, and IA hypotheses become grounded.
UX Research
Surveys
Collects answers in a scalable format and condenses them into results, charts, and segment insights.
Product Discovery
Solution Interview
Clarifies customer problems, solution ideas, and evidence by preparing a solution concept, recruiting target customers, and deriving risks and next tests.
Innovation
Design Sprint
Brings problem space, prototype, and test findings together in a few days, enabling a well-founded decision.
Constraints and origin
MediumWorkshop + async1-2 h je Session
Profile
- Method form
- Research method
- Participants
- Based on research question
- Time
- 1-2 h je Session
- Maturity
- Canonical
- Origin
- UX Research Community
Next up: Affinity Diagramming