What is a UX audit?
A UX audit (also called a usability audit or UX review) is an expert inspection of an interface. Instead of asking users, the auditor walks through the key journeys and compares every screen with established principles: how people perceive, decide, remember and move a pointer or a thumb.
It answers three questions: where do visitors struggle, why do they struggle, and what should change first? The output is not a redesign; it is a list of evidence-backed problems and fixes that a designer or developer can act on.
A UX audit is different from usability testing. Testing watches real people use the product; an audit applies known rules. The two complement each other — an audit is faster and cheaper, so it usually comes first and removes the obvious problems before you spend time testing with users.
When do you need a UX audit?
- Before a redesign — so you keep what works and fix what doesn't, instead of starting from opinions.
- After a drop in conversions — to find which change or which page is losing people.
- Before handing a website to a client — to deliver proof that it follows usability and accessibility standards.
- After launch, on a schedule — because every content update or deploy can quietly break something.
- Before investing in ads — paid traffic is wasted on a page that confuses visitors.
What a UX audit covers
A complete website UX audit looks at more than visual design. These are the areas to review and the reference standard for each one:
| Area | What to check | Reference |
|---|---|---|
| Navigation & structure | Clear menu, findable pages, logo links home, breadcrumbs on deep pages | Hick's Law, Jakob's Law |
| Interaction | Size and spacing of buttons and links, primary action placement | Fitts's Law, WCAG 2.5.8 |
| Content & readability | Scannable headings, short paragraphs, plain language | Miller's Law, reading level |
| Visual hierarchy | One clear primary action per screen, contrast, grouping | Gestalt principles, Von Restorff effect |
| Forms | Few fields, inline validation, helpful error messages | Nielsen heuristics 5 and 9 |
| Accessibility | Contrast, alt text, heading order, keyboard access, skip links | WCAG 2.2 AA |
| Performance | Server response, loading speed, interaction delay | Doherty Threshold, Core Web Vitals |
| Trust & persuasion | Honest pricing, no dark patterns, proof and guarantees | Dark-pattern research |
How to do a UX audit in 7 steps
Define goals and key journeys
Write down what the site must achieve (sales, sign-ups, enquiries) and the 3–5 journeys that lead there, such as “landing page → pricing → sign-up”. You audit those journeys first.
Collect behaviour data
Check analytics for pages with high exit rates, funnel steps where people drop, slow pages and device splits. Support tickets and sales questions point to confusing content.
Run a heuristic review
Walk each journey and compare every screen with Nielsen's 10 usability heuristics and the laws of UX. Log each problem with a screenshot and the rule it breaks.
Check accessibility
Test text contrast with a contrast checker, tab through the page with the keyboard, check headings, alt text, form labels and touch-target size against WCAG 2.2 AA.
Measure performance
Look at Core Web Vitals — Largest Contentful Paint, Interaction to Next Paint, Cumulative Layout Shift — and server response time. Slow answers break the user's flow (see the Doherty Threshold).
Prioritise the findings
Rate every issue by severity (how badly it blocks the user), reach (how many users meet it) and effort to fix. Fix high-severity, high-reach, low-effort issues first.
Report and re-test
Write each finding as: problem, evidence, rule, severity, recommendation. After the fixes ship, re-run the audit to confirm they worked and nothing else broke.
UX audit checklist
Use this checklist on every key page. Each line is a yes/no question; every “no” becomes a finding.
Navigation
- Does the logo link to the home page?
- Does the main menu have 7 items or fewer?
- Is the current page highlighted in the navigation?
- Do deep pages show breadcrumbs?
- Is search available on large sites?
Interaction
- Are buttons and links at least 44 × 44 px on touch screens?
- Is there one clear primary action per screen?
- Do buttons react on hover, focus and press?
- Is the primary action within thumb reach on mobile?
Content
- Can you understand the page in five seconds?
- Do headings describe what follows?
- Are paragraphs short and lists used for steps?
- Is the language plain and free of jargon?
Forms
- Is every field really needed?
- Are fields validated as the user types?
- Do error messages say how to fix the problem?
- Are input types right (email, tel, url)?
Accessibility
- Does text reach a 4.5:1 contrast ratio?
- Do images have meaningful alt text?
- Do headings follow h1 → h2 → h3 without gaps?
- Is there a “skip to content” link?
- Is the page language declared?
Performance & trust
- Does the server answer in under 400 ms?
- Do images use modern formats and lazy loading?
- Are prices and conditions clear before checkout?
- Are there no pre-checked add-ons or guilt-trip buttons?
How to prioritise UX audit findings
An audit that lists 80 problems with no order is ignored. Score each finding on three axes and sort by the result:
| Score | Severity | Reach | Effort |
|---|---|---|---|
| 3 | Blocks the task (user can't continue) | Most visitors, key journey | Minutes |
| 2 | Slows the user down or causes errors | A common segment, e.g. mobile | Hours |
| 1 | Cosmetic or minor friction | Rare edge cases | Days or more |
Fix the issues with high severity and reach first, especially when the effort is low. Nielsen's severity scale (0 = not a problem, 4 = usability catastrophe) is a good shared vocabulary between designers and developers.
Manual vs. automated UX audits
Many audit rules are measurable: a button's size in pixels, a contrast ratio, the number of menu items, the server response time, a missing alt attribute. A tool can check those on every page in minutes and repeat the check after each change. Judging whether a headline is persuasive or an image fits the brand still needs a person.
| Manual audit | Automated audit (Lensux) | |
|---|---|---|
| Speed | About a day per site | Minutes per site |
| Coverage | Key pages, sampled | Every checked page, every rule |
| Repeatability | Depends on the reviewer | Same rules, same scoring |
| Context and copy | Strong | AI review plus your judgement |
| After launch | Rarely repeated | Scheduled monitoring |
The most efficient approach combines both: let the tool clear the measurable issues, then spend expert time on content, flows and user testing.
Lensux runs 43 checks across UX laws, accessibility, typography, layout, colors, assets, design psychology and trends. Each finding names the element, the rule, the severity and the fix, and monitoring re-checks your sites after every change.
What to include in a UX audit report
Keep the report short and actionable. For each finding include:
- Problem — what goes wrong, in one sentence.
- Evidence — screenshot, measurement or analytics number.
- Rule — the heuristic, law or WCAG criterion it breaks.
- Severity — from cosmetic to blocking.
- Recommendation — the concrete change, and who owns it.
Start the report with a one-page summary: overall score, the five most important fixes and the expected impact. Decision-makers read that page; the team uses the rest.
Frequently asked questions
How long does a UX audit take?
A manual audit of a small website usually takes one to several days, depending on the number of pages and journeys. An automated audit with a tool like Lensux takes minutes per site, which makes it possible to repeat it after every change.
What is the difference between a UX audit and usability testing?
A UX audit is an expert review against known principles; usability testing observes real users trying to complete tasks. Audits are faster and cheaper and catch known problems; testing reveals problems nobody predicted. Most teams audit first and test afterwards.
Who should do a UX audit?
Ideally someone who did not design the site, because fresh eyes spot more problems. Nielsen's research suggests that three to five evaluators together find most usability problems that a single reviewer would miss.
Can a UX audit be done for free?
Yes. You can follow the checklist on this page by hand, or run the free Lensux scan, which checks 24 rules on any public page in seconds without an account.