UX audit

A UX audit looks at an existing product or site and answers three questions: where do people get stuck, why, and what should be fixed first. The expert part usually uses Jakob Nielsen’s 10 usability heuristics, first developed with Rolf Molich in 1990 and refined in 1994, such as visibility of system status, consistency and standards, and error prevention. Nielsen Norman Group recommends three to five evaluators reviewing independently, because one reviewer misses issues others catch.

Good audits add evidence on top: funnel analytics, heatmaps and scroll depth, session recordings, device and browser splits, support tickets, sales call notes and an accessibility check against WCAG 2.2. The output is a prioritised list of issues, each tied to evidence and a recommended fix.

Why it matters for a SaaS website

Redesigns fail when they fix what the team dislikes instead of what users struggle with. An audit grounds the work in data, separates quick wins from structural problems, and gives you a sequence: fix the trial sign-up before polishing the blog. It is also the low-risk way to start with a new design partner, because the result is useful even if you do the work in-house.

How DesignMe handles it

On product design projects, week one is a product audit: the current product, users, flows, competitors, business goals, analytics where available, and the places people get stuck. It ends with flow notes, UX issues and sprint priorities. Website projects start the same way with a website audit.

The Seamless.AI engagement began with a three-day audit sprint on the landing page and brand, with one concrete recommendation: a refreshed hero. A deeper audit followed with their CRO and growth team, covering heatmaps, click maps, device and browser breakdowns, GA events and page-by-page analysis of the home, pricing and demo pages. It showed only about 5% of homepage visitors gave an email and clicked the main CTA, and around a quarter scrolled past the fold.

Example

Say a B2B SaaS has healthy traffic but few trial sign-ups. The audit finds the pricing page hides the free trial below three plan tables, the sign-up form asks for company size and phone number before showing the product, and mobile users drop off at a date field that opens the wrong keyboard. Each issue gets evidence, a severity and a fix, and the top three ship in the first sprint.

Common mistakes

• Opinions without evidence. “The hero feels weak” is not a finding. Tie each issue to data or a heuristic.
• A 90-slide report. If nothing is ranked, nothing gets fixed.
• Auditing screens, not flows. Most problems sit between screens, in the user flow.
• Skipping accessibility. Contrast, focus states and form labels affect every user, not only those using assistive tech.