Run a live website UX audit without leaving Figma

Website audit in Figma

Review the current website state next to the design response

Design files are good at showing intent and weak at preserving the accidental state of production. Cookie banners, campaign bars, embedded widgets, stale labels, and mobile stacking behavior may never appear in the source file. Capturing the live page gives critique a shared starting point: this is what a visitor could actually see when the audit ran.

Keeping that capture in Figma shortens the distance between evidence and response. A designer can place exploration frames beside the audited state, refer back to the exact pinned problem, and preserve the reason for a change. The canvas becomes a working review artifact instead of a collection of screenshots with detached comments.

Workflow

Capture, review by lens, then apply product context

Start by pasting the public URL in the plugin and choosing desktop, mobile, or both. If the page needs a public interaction before capture — opening a menu, dismissing a banner, or changing a pricing toggle — describe the desired state in plain language when configuring AI assistance. Snap Site captures the rendered result and organizes it on the canvas.

Review the first-pass findings one lens at a time: usability heuristics, selected accessibility evidence, hierarchy and clarity, then responsive differences. Treat every observation as a candidate. Confirm the pin, edit the explanation, change the severity, reject what lacks context, and add what only someone who knows the product and its users can identify.

• Paste a public URL and choose the target screen sizes
• Use separate heuristic, accessibility, hierarchy, clarity, and responsive passes
• Confirm each observation against its pinned evidence
• Keep redesign options and resolved decisions beside the original capture

Heuristic evaluation

Use AI for breadth without handing it the final decision

The first pass applies consistent lenses across a long page: Nielsen Norman Group's usability heuristics, evidence-supported WCAG 2.2 A/AA checks, and curated UX patterns for content, forms, navigation, conversion cues, and responsive behavior. It is useful precisely because it does not lose patience halfway down the page.

Consistency is not context. The plugin cannot know which audience matters most, what research already disproved, why a trade-off was made, or which metric the team is protecting. The reviewer remains responsible for deciding whether the evidence is accurate, whether the finding matters, and whether the next step is a design change, experiment, user test, or no action.

When to use it

Best when the audit is going to become design work

Use the Figma plugin for heuristic evaluations, redesign discovery, responsive critiques, competitor reviews, and pre-launch checks where the output needs to sit beside design explorations. Use the browser workspace when the review is feeding a cross-functional meeting, a shared link, or a PDF handoff. Both surfaces draw from the same audit history.

The plugin reviews captured public pages and uploaded visual evidence; it does not turn Figma into a browser test runner. Private authenticated journeys, exact performance behavior, screen-reader output, and representative user behavior still need their appropriate tools and people.