Website Usability Audit: Task Completion and Error Recovery

A website usability audit checks whether people can complete defined tasks effectively and with understandable feedback. Select realistic tasks, record the test conditions and distinguish observed failures from an expert's prediction of difficulty.

What should a usability audit measure?

Record task completion, errors, assistance, detours and recovery. Time on task can be useful when the task and conditions are comparable. A faster task is not necessarily better if the participant misunderstood the result.

Define success before testing. For a service enquiry, success may include finding the scope, understanding the commitment and completing an agreed test submission. A click on the form button is only one step.

How does usability differ from UX?

Usability focuses on completing tasks. A UX audit also considers broader expectations, comprehension, satisfaction and the overall journey. Neither guarantees a commercial outcome.

Which principles guide an expert review?

Jakob Nielsen's ten usability heuristics provide a framework for inspection. Apply them to concrete tasks rather than awarding a site a score without evidence.

Show the current state and the result of an action.
Use language familiar to the audience.
Provide safe ways to cancel, go back and recover.
Keep similar controls and conventions consistent.
Prevent foreseeable input mistakes.
Make needed information visible instead of requiring memory.
Support efficient use without hiding the basic path.
Remove information that distracts from the task.
Explain errors and the action needed to correct them.
Provide relevant help when the task requires it.

Ask a participant to find something they genuinely need, without repeating the exact menu label in the task instruction. Record the first choice, mistaken routes and whether they recover.

Check the mobile menu, breadcrumbs, search results and return path. There is no universal three-click rule: a clear longer route can work better than a short ambiguous one. Judge the route against observed task success.

What makes content usable?

Content is usable when the audience can locate, understand and act on it. Define unfamiliar terms, put prerequisites before instructions and use headings that describe the information beneath them.

Test real content at narrow widths and zoom. Readability formulas can flag complex writing but cannot determine whether a specialist explanation is accurate or useful. The design audit covers typography, contrast and responsive presentation.

How should form usability be reviewed?

Test expected input, missing required input, invalid input, correction and successful completion. Check that instructions and labels remain available while typing and that useful values survive an error.

Use suitable input types and autocomplete. Explain requirements before they cause a failure. W3C's forms tutorial covers labels, instructions and feedback. Do not use colour alone to identify a problem.

Coordinate test submissions and use test data. Record confirmation and delivery separately; a success message does not prove that the business received the request.

Which testing method fits the question?

Expert inspection: identify likely barriers against a declared framework.
Task testing: observe relevant participants working through realistic goals.
Tree testing: test labels and hierarchy with less influence from visual design.
Recorded-session review: inspect permitted recordings for interactions that need investigation.
Controlled experiments: compare uncertain variants when measurement and sample size support the decision.

Participant numbers depend on the audience groups, tasks and research aim. A small qualitative round can reveal useful issues but does not guarantee a fixed percentage of all problems or a statistically representative result.

What should the usability handoff include?

Audience, tasks, conditions and success definitions.
Observed behaviour with relevant screenshots or notes.
Separate interpretation, severity and confidence.
A proposed correction and accountable owner.
Keyboard, mobile and error-recovery checks where applicable.
A retest that repeats the original task and records remaining limits.

The audit toolkit provides an implementation tracker. Use a dedicated accessibility assessment when the scope requires broader standards-based evaluation.