Expert Usability Audits

What an audit is good for

An audit is the quickest way to get an informed outside view of a site. It works well before a redesign, after a launch, when a site has grown for years without a review, or when you want to clean up obvious problems before paying for research with participants.

Because the reviewer works alone, an audit costs less and moves faster than a study with recruits. The tradeoff is that its findings are predictions. An experienced reviewer can spot many problems reliably, but some issues only appear when real people with real goals use the site. The audit report labels every finding as an expert judgment so that nobody mistakes it for observed behavior.

How we run the audit

We begin by agreeing on the scope: the pages, the tasks and the audiences that matter most. A starting scope covers the home page, two or three core journeys and the forms along them, on both desktop and phone.

The reviewer then walks each task as a new visitor would, noting every point of confusion, effort or doubt. Each issue is checked against Jakob Nielsen's ten usability heuristics, which our heuristic evaluation guide explains in detail, along with basic accessibility checks from WCAG 2.2 and the Core Web Vitals thresholds for loading and stability ¹.

Different reviewers notice different problems. If you want wider coverage, someone on your team can review the same scope independently with our checklist, and we merge both sets of findings and remove duplicates.

How findings are rated

Every finding receives a severity rating so that your team can see at a glance what to fix first.

Severity reflects three things together: how many people a problem is likely to affect, how badly it disrupts the task and whether people can work around it. A frequent annoyance can outrank a rare but dramatic error.

What you receive

The audit report opens with a one-page summary of the most important issues. The body lists every finding with a screenshot, the page and task where it occurs, the principle it breaks, the severity rating and a concrete recommendation. A spreadsheet version of the findings lets your team import them into a backlog. We finish with a video readout where we walk through the top issues and answer questions.

You can see how a report is laid out, with fictional content, on our sample audit report page.

Timeline and inputs

We plan for one to three weeks, depending on the number of pages and flows. We need access to the site, any logged-in areas in scope through a test account, and a short brief describing the audience and the business goal of each flow. If you have analytics, read-only access helps us focus on the pages that carry the most traffic.

When to follow up with research

An audit can raise questions it cannot settle on its own, such as whether a confusing label really confuses your audience or only the reviewer. Those are good candidates for a short round of research with participants. Testing the top audit findings with five people is an efficient way to confirm which problems are real and how serious they are. If the audit points to a menu that people cannot navigate, tree testing and card sorting are the better follow-up.