Website UX Audit: Research, Journeys and Prioritised Findings

A website UX audit reviews how visitors understand an interface and complete their intended tasks. It combines available research, interface inspection and behavioural data into prioritised findings. The scope should name the audience, journeys and devices represented rather than claim to explain every visitor's experience.

What should a UX audit establish?

Establish which tasks work, where participants or reviewers encounter barriers, and what evidence supports a proposed correction. Distinguish an observed failure from an expert hypothesis and an untested design preference.

For example, a participant repeatedly opening the wrong menu is an observation. “The label does not match the participant's expectation” is an interpretation to investigate. A revised label needs another task test; its visual appearance alone is not verification.

How should UX findings relate to business outcomes?

Connect findings to the affected task and its business relevance without assigning invented revenue losses. An inaccessible enquiry form matters because some visitors cannot complete it; quantifying its commercial effect requires reliable measurement.

A lower mobile completion rate can reflect different audiences, intent or tracking as well as interface friction. Use a CRO audit to define the measurement and experiment plan. Do not assume every UX correction produces an immediate conversion or ranking increase.

Which evidence belongs in the review?

• Audience and task: who needs to do what, with which constraints?
• Comprehension: can visitors explain the offer, terminology and next step?
• Navigation: do labels and routes support the intended task?
• Interaction: do controls, states and errors behave predictably?
• Accessibility: can the task be completed with relevant assistive technologies and input methods?
• Trust: are claims, prices, evidence and commitments accurate and understandable?

Use Nielsen's usability heuristics as prompts for expert inspection. A heuristic finding is not the same as observing a representative participant struggle.

How should navigation be evaluated?

Give participants a realistic destination or question without naming the menu item they should choose. Record the first choice, detours, recovery and task result.

Tree testing can isolate labels and grouping from visual design. Card sorting can reveal how participants group content, but the resulting groups still need testing against real tasks. There is no universal three-click limit or ideal number of menu items.

Inspect search results, empty states, breadcrumbs and back navigation. Confirm that important destinations remain reachable on mobile and with a keyboard. The site-structure guide covers the corresponding crawl and internal-link checks.

How should forms and customer journeys be tested?

Test the route to completion and recovery from errors. Use an agreed test environment or clearly labelled test enquiry so the review does not create real orders or misleading business records.

Check persistent labels, required-field instructions, useful error messages, retained valid input and a clear confirmation. W3C's forms guidance provides an accessibility reference. A successful browser message does not establish that an enquiry reached its destination.

For detailed action-by-action inspection, use the conversion journey audit. Record the reason for each requested field rather than treating every extra field as a fixed percentage loss.

What changes in a mobile UX review?

Review the narrow-screen reading order, navigation, input methods and space taken by sticky elements. Check that overlays can be dismissed and do not cover focused controls or essential content.

Use browser emulation for repeatable widths and real devices for touch, keyboard and orientation behaviours when available. Record unavailable device evidence. Reflow, text resizing and target spacing need explicit checks; a desktop screenshot is not mobile validation.

What do UX tools tell you?

Analytics can identify recorded paths and differences between segments. Session recordings can show selected interactions. Neither reliably reveals a person's intention without additional research, and both can have consent or coverage gaps.

Automated accessibility tools identify some detectable failures. Browser performance measurements describe loading and interaction conditions. Combine them with manual review and the task-based usability workflow.

What should the UX action plan contain?

For each finding, record the task, affected page, participant or test conditions, observed barrier, severity, confidence, proposed correction and retest. Prioritise blocked tasks and broad barriers before cosmetic preferences.

Use the implementation tracker to distinguish prepared changes from verified releases. An AuditWeb HTML check can supply markup observations; it does not perform this UX review.