What the audit reviews
| Area | Questions asked |
|---|
| Navigation and structure | Can users find features and content? Do labels match their language? |
|---|
| Key task flows | Can users complete core tasks without confusion, dead ends or repeated steps? |
|---|
| Feedback and status | Does the product show what's happening, what's next and who needs to act? |
|---|
| Forms and inputs | Are fields clear, validated helpfully and forgiving of different formats? |
|---|
| Errors and recovery | Do errors explain what happened and how to fix it, without losing work? |
|---|
| Content and terminology | Is language clear, consistent and appropriate for the audience? |
|---|
| Visual hierarchy | Does the most important information stand out? Are actions obvious? |
|---|
| Consistency | Do the same things look and behave the same way across the product? |
|---|
| Accessibility | Contrast, keyboard access, labels, focus and structure against WCAG 2.1 AA |
|---|
| Performance perception | Do loading and waiting states keep users informed? |
|---|
The review draws on Jakob Nielsen's 10 usability heuristics - principles first published in 1994 and still unchanged in substance - applied to your specific users and context.
Evidence behind every finding
Each finding in the audit report includes:
- Where - the page, screen or step, with a screenshot.
- What - a plain description of the problem.
- Why it matters - the principle it breaks and the user impact.
- Evidence - analytics, support themes, test observations or benchmark comparisons.
- Severity - how badly it affects users and the business.
- Recommendation - a specific fix, with an example where useful.
- Effort - a rough estimate agreed with your team where possible.
Where evidence is thin, the recommendation says so - and suggests a quick test rather than a confident change.
Severity and prioritisation
| Severity | Definition | Example |
|---|
| Critical | Prevents task completion or creates risk | A required form can't be submitted by keyboard; a status is misleading |
|---|
| Major | Causes frequent errors, delays or abandonment | Users repeatedly enter the same data; errors lose progress |
|---|
| Moderate | Slows users or causes confusion they recover from | Ambiguous labels; inconsistent button placement |
|---|
| Minor | Cosmetic or low-frequency | Spacing inconsistencies; typos |
|---|
Combining severity with effort gives the fix-first list: high-severity, low-effort issues at the top. In regulated contexts such as healthcare, a third factor - compliance risk - is added. See UX audits for healthcare platforms.
Validating with users
An expert review finds a lot, but users confirm what matters. Where time allows, I run a small round of moderated sessions on the biggest suspected issues. Testing with about five users per round uncovers roughly 85% of usability problems, according to Jakob Nielsen's analysis - enough to confirm or rule out the top findings quickly.
