7 Usability Evaluation Methods

What usability evaluation measures

Usability evaluation examines how well people can use an interface to achieve goals in a particular context. Usability has three dimensions: effectiveness, efficiency, and satisfaction. The appropriate evaluation method depends on what a team needs to learn, the maturity of the design, and the time and resources available.

Heuristic evaluation

In a heuristic evaluation, one or more evaluators inspect an interface and compare it with usability principles, or heuristics. Common heuristics include showing system status, using familiar language, supporting undo and error recovery, and keeping the interface consistent. Evaluators can document each issue, its location, the affected task, and a proposed improvement.

This method is relatively quick and can be used with early prototypes or finished products. It helps identify likely design problems before user sessions, but it reflects evaluator judgment rather than direct evidence of how users behave. It should complement, not replace, evaluation with users.

Usability testing

Usability testing observes people from the intended user group as they attempt representative tasks. A facilitator presents scenarios, watches participants, and records successes, errors, delays, questions, and signs of confusion. Participants may be invited to think aloud, but this can affect task performance and should be considered when interpreting timing data.

Testing can be moderated or unmoderated, conducted in person or remotely, and use paper prototypes, interactive prototypes, or a working product. It provides behavioral evidence about what participants could accomplish, where they struggled, and what they said about the experience. Combining observations with performance measures gives a fuller picture than relying on opinions alone.

Task-based assessment

Task-based assessment evaluates performance against specific user goals. For each task, define what counts as success, which errors or detours matter, and what measures to collect. It may be conducted as a usability test; the term emphasizes measuring performance on defined tasks.

Useful measures include:

• Completion: whether the task was completed independently, completed with help, or not completed.

• Accuracy: errors, incorrect selections, or incomplete results.

• Efficiency: time, number of steps, or requests for assistance.

• Experience: reported ease, confidence, or satisfaction.

For example, an assessment of a library website could ask participants to find a particular book and determine whether it is available at their local branch. Record whether they find the correct item, how they locate availability, and any confusion along the way.

Planning a study

Begin by setting a decision-focused research question that specifies what the team needs to learn. One example is whether first-time users can submit an application without assistance.

Then define the context and participants. Identify intended users, relevant experience, devices, and accessibility needs. Recruit people who reasonably represent the user group, including people with disabilities when relevant to the product and study goals.

Choose an interface and method that fit the question and design stage: inspect a design, observe task performance, or combine methods. Select important tasks and write realistic, goal-oriented scenarios without naming the interface controls participants should use. For example, “You need to reschedule your appointment for next week” is more neutral than “Open Settings and select Reschedule.” Pilot instructions to catch ambiguity.

Prepare a consistent procedure covering consent, setup, prompts, task order, note-taking, recording, and post-task questions. Facilitate neutrally: help participants continue only as the protocol allows, and distinguish assistance from unaided success. Choose measures in advance, defining completion criteria and how errors, time, assistance, and subjective feedback will be recorded. When comparing versions, use comparable tasks and conditions.

Evaluating accessibility

Accessibility evaluation involves more than an automated or standards checklist. Conformance evaluation can assess a digital product against WCAG, while involving people with disabilities can reveal practical barriers in real use.

WCAG-EM recommends defining the evaluation scope, examining the product, selecting representative pages or views, evaluating them, and reporting the findings.

Interpreting and reporting results

Organize observations by task and issue. Describe each problem concretely, including what happened, the context, and the effect on the user’s goal. Assess severity by considering how often the problem occurred, how much it obstructed task completion, and whether it could cause serious consequences. Prioritize consequential barriers over cosmetic preferences.

Use quantitative and qualitative evidence together. A low completion rate indicates a performance problem; observations and participant comments may help explain why. A fast completion time is not automatically a success if participants made mistakes or felt uncertain.

Report the participant group, tasks, study conditions, measures, notable limitations, and recommendations. Small exploratory studies can uncover issues, but should not be treated as precise estimates of how common those issues are in the entire user population. After changes, retest important tasks to check whether performance improved and whether new problems appeared.