UX audit: find what stops visitors and users, ranked by impact

What does a UX audit include?#

A UX audit is a fixed-scope diagnosis of the tasks that matter, ending in findings your team can act on. The proposal names the tasks, user groups, devices and methods in scope:

• An agreed list of the tasks each user group must be able to complete
• An expert review of each task by reviewers working separately, then compared
• Funnels, session recordings and heatmaps read in your GA4 or PostHog
• Moderated usability tests, about five participants per round
• Keyboard, screen reader and contrast checks on the audited tasks
• A ranked findings list, annotated screens and a readout call

Is a UX audit the right starting point?#

Choose a UX audit when you know something isn't working but not where or why: sign-ups stall at one step, support hears the same question every week, or a redesign is planned and nobody agrees on what to keep.

Some briefs fit another scope. If the problems are known and traffic is steady, conversion rate optimization ships changes and measures them. For WCAG 2.2 AA conformance before a tender or after a complaint, a web accessibility audit tests every template against each success criterion. Pages that are simply slow need website speed optimization. If the site no longer describes the business, plan a website redesign and run the audit inside its discovery.

Which methods we use, and what each one finds#

No single method finds everything, so each audit combines an expert review, behavioral data and tests with real people. The review uses Jakob Nielsen's 10 usability heuristics, which Nielsen Norman Group (NN/g) calls broad rules of thumb, unchanged since 1994. NN/g also says a heuristic evaluation cannot replace user research, so the review never stands alone.

PostHog heatmaps show clicks, dead clicks, rage clicks and scroll depth; we run PostHog on our own site, with an opt-out in the footer. The W3C notes that evaluation tools can assist but not determine accessibility, so we test each task with a keyboard and a screen reader. If tracking is missing or wrong, the audit says so, and analytics setup fixes it before the next review.

How the audit runs, from scope to readout#

Each audit tests named tasks, not screens at random. Scoping lists the tasks per user group, such as finding pricing, booking a demo or exporting a report, and gathers what you already have: analytics, support tickets, the questions sales hears and past research.

Three of us then review the tasks separately, the minimum NN/g recommends, because each evaluator misses problems another one catches. Findings that rest on an assumption become questions for the tests.

Tests are moderated remote sessions in English or Spanish, in which participants attempt each task and think aloud. Nielsen's model estimates that five participants uncover about 85% of a design's usability problems, and he recommends three rounds of five over one round of fifteen, so where the budget allows we test, fix and test again. Participants come from your customers or a recruiting panel, with incentives listed as a third-party cost. Bogotá is on UTC−5 all year, so sessions fit US working hours.

How findings are ranked: severity, reach and effort#

Each finding gets a severity, a reach and an effort estimate, so the list sorts itself into what to fix now and what can wait. Severity uses NN/g's 0–4 scale, which weighs how often a problem occurs, how hard it is to overcome and whether it keeps bothering people: 4 is a catastrophe to fix before release, 1 is cosmetic. Reach is the share of sessions or accounts that get to that step; effort is a design and code estimate.

Every entry has the same parts, so it can become a ticket without another meeting:

• What happens, where, and on which device
• The evidence: timestamped clips, replays, funnel numbers or the heuristic it breaks
• Who it affects, and how many people reach that step
• The recommended change, shown on an annotated screen

Auditing a website versus a product#

Website visitors and product users need different audits. Website visitors usually arrive anonymous, often on a phone, and decide quickly whether they're in the right place, so a website audit tests whether the offer is clear, whether navigation uses their words and whether forms and the mobile path to an inquiry hold up.

Product users come back, logged in, to get work done. A product audit covers onboarding, how fast an experienced user finishes the core task, empty, error and permission states, and recovery from mistakes. New and experienced users are tested as separate groups, as NN/g advises for distinct user groups. When the answer is a new flow rather than a fix, product design takes over.

What a UX audit can't tell you#

A UX audit shows where and why people struggle; it does not measure what a fix is worth. These questions need other methods:

• Demand: a usability test observes use, not whether people would pay
• How often a problem occurs across all visitors: NN/g puts quantitative studies at 20 participants or more
• The effect of a change: that needs an A/B test or a clean before-and-after comparison
• Full WCAG conformance across every template and success criterion

What moves the cost and timeline?#

There's no published starting price for a UX audit; the quote follows the scope. The audit stands alone, so you can fix the issues in-house. These factors move the number:

• How many tasks, flows and devices are tested
• User groups, since each one needs its own participants
• Test rounds, participant recruiting and incentives
• A product behind a login, with roles, test accounts and seeded data
• Sessions and findings in English, Spanish or both
• Whether Dardo also designs, builds and retests the fixes

The proposal sets the dates. To start, we need read access to your analytics and recordings, test accounts, any support or sales notes and one person who approves the task list. Our published prices and budget ranges show how other Dardo services are priced.

Questions before you choose#

From the blog#

• PostHog vs Google Analytics: which one a marketing site or SaaS product needs, and when to run both

  GA4 links to Google Ads and Search Console; PostHog adds replay, feature flags and experiments. Which one you need, what each costs and how to run both.