What a UX Audit Actually Finds

A UX audit sounds like an opinion with an invoice attached. Done properly it is closer to an inspection: a list of specific places where the product costs you money, each with evidence and an estimate of what fixing it is worth.

It is not a redesign proposal

The most common misunderstanding is that an audit ends with new screens. It should not. An audit that recommends a redesign has usually skipped the hard part, which is working out which of the hundred things wrong with a product actually matter.

What comes back instead is a ranked list. Ranked by cost, not by how much the reviewer disliked it.

Where the money usually hides

Forms that ask too early. Asking for a phone number before showing a price is the single most reliable way to lose people in a booking flow. It is also the easiest to fix, which makes it a good first finding.

States nobody designed. Empty, loading, error, and "you already did this" are where products fall apart, because they were never drawn. Users meet them constantly.

Navigation that mirrors the org chart. If your menu is named after your departments, visitors have to learn your company structure before they can find anything.

Copy that describes the feature rather than the outcome. "Configure your workspace" tells someone what the software does. "Add your team" tells them what they get.

What evidence looks like

Every finding should carry one of three things: a recording of a real user hitting it, an analytics drop-off at that exact step, or a heuristic breach specific enough to argue with. "Feels cluttered" is not a finding. "Seven of ten users scrolled past the primary action because it sits below a decorative banner" is.

What to expect from the report

We hand over the ranked list, the evidence for each item, and a rough effort estimate, because a finding you cannot schedule is just criticism. Roughly a third of items are usually copy or ordering changes that a developer can ship in an afternoon. Those are worth doing before anyone opens a design tool.

The remainder splits into things worth designing properly and things worth leaving alone. Saying which is which is most of the value.

When an audit is the wrong tool

If you have not launched, an audit has nothing to inspect, you want research instead. If you already know what is wrong and just need it fixed, an audit is an expensive way to confirm it. Audits earn their fee when a product is live, the numbers are disappointing, and nobody can agree on why.