UX Audits: How to Find and Fix the Usability Problems Costing You Users

Products accumulate usability debt. Features are added, flows change, content grows and new devices appear. Users adapt, complain or leave. A UX audit is a structured way to find these problems, understand their impact and decide what to fix first.

This guide sets out a practical audit process that teams can run themselves or with a partner.

When to run a UX audit

• Conversion, activation or retention metrics are falling or below expectations.
• Support tickets reveal repeated confusion about the same tasks.
• Before a redesign, to know what to keep and what to change.
• After major feature additions or a platform migration.
• Before entering new markets or serving new user groups.
• When accessibility compliance is required or questioned.

What a UX audit covers

A step-by-step process

1. Define goals and journeys. Agree which business metrics matter and which journeys support them.
2. Review analytics. Funnels, drop-off points, time on task, error events, search queries with no results and device mix.
3. Review qualitative data. Support tickets, reviews, session recordings and survey comments.
4. Heuristic evaluation. Two or three reviewers independently assess journeys against Nielsen’s 10 usability heuristics — first published in revised form in 1994 — then consolidate findings.
5. Accessibility review. Automated checks plus manual keyboard and screen-reader testing against WCAG 2.2, which became a W3C Recommendation on 5 October 2023 and added nine success criteria.
6. Usability testing. Observe five to eight representative users attempting key tasks; note where they hesitate, fail or misunderstand.
7. Synthesise. Group findings, link them to evidence and estimate their impact.
8. Prioritise and plan. Agree fixes, owners and the metrics that should improve.

Prioritising findings

A simple scoring approach multiplies severity by the share of users affected and the value of the journey. Quick wins — high impact, low effort — go first; larger redesigns are planned into the roadmap.

Turning findings into results

• Write each finding as a problem with evidence, not a solution: “40% of users abandon at the address step; testing shows confusion between billing and delivery addresses.”
• Propose solutions and test them with prototypes or A/B experiments where impact is uncertain.
• Track metrics before and after each fix.
• Feed patterns back into the design system and guidelines so the same problems do not recur.

Worked example: a B2B onboarding flow

A hypothetical SaaS product sees only a minority of trial users completing setup. The audit finds:

• Analytics: most drop-offs occur at the “connect data source” step.
• Heuristics: technical jargon, no progress indicator, error messages showing raw codes.
• Testing: five of seven users could not find the API key they needed in their other system.
• Accessibility: the setup wizard traps keyboard focus in a modal.

Fixes include plain-language instructions with screenshots, a progress indicator, a “send instructions to a colleague” option, clear error messages and an accessible modal. Completion of setup is measured weekly before and after release.

UX audit checklist for a key journey

• Can a first-time user understand what to do on every screen without help?
• Are required fields minimal, clearly labelled and validated with helpful messages?
• Does the interface show progress, loading and success states clearly?
• Can users recover from errors without losing work?
• Are labels and terminology consistent across the journey?
• Does the journey work with keyboard only and with a screen reader?
• Are text contrast, size and touch targets adequate on mobile?
• Is performance acceptable on a mid-range phone and slow network?
• Do analytics events capture each step so drop-off can be measured?
• Are help and contact options visible where users get stuck?

Combining UX and QA

UX audits and QA work well together. QA teams already test journeys systematically; adding heuristic and accessibility checks to regression testing catches usability regressions before release. Conversely, audit findings become test cases so fixed problems stay fixed. Our DesignOps guide describes how design QA fits into the definition of done.

What a good audit report contains

How often to audit

• Full audit: annually, or before and after major redesigns.
• Quarterly checks: funnel review, accessibility scan and a small usability test on one journey.
• Continuous signals: dashboards for drop-off, errors and support themes.

Frequently asked questions

How long does a UX audit take?

A focused audit of a few key journeys typically takes two to four weeks, depending on access to data and users.

Can automated tools do a UX audit?

They help with accessibility, performance and analytics, but usability problems need human review and observation of real users.

Who should conduct the audit?

People with UX research and design expertise, ideally including someone who did not design the current product, together with product and QA team members.

How is a UX audit different from usability testing?

Usability testing is one input. An audit combines it with analytics, expert review, accessibility checks and prioritised recommendations.