UX Audit

UX Audit

When to Use

Use this skill when:

• The user asks to audit, evaluate, or review the usability of an existing interface -- a web application, mobile app, dashboard, checkout flow, onboarding sequence, settings screen, or any interactive product surface
• The user wants a structured heuristic evaluation grounded in Nielsen's 10 usability heuristics, producing severity-rated findings and a prioritized remediation plan
• The user needs to identify why users are struggling with a product -- support tickets mention confusion, drop-off analytics show abandonment at a specific step, or user testing revealed friction but not root causes
• The user wants to compare a redesign against an existing interface to evaluate whether usability has improved, regressed, or stayed the same
• The user wants to benchmark a product's usability quality before commissioning user research or a formal redesign, establishing a baseline of known issues
• The user wants a quick expert evaluation (heuristic evaluation is 75% as effective as full usability testing at finding serious problems, per Nielsen Norman Group research) before investing in full usability studies
• The user is a product manager, designer, or developer who needs a structured list of UX debt to feed into sprint planning or backlog grooming

Do NOT use when:

• The user needs an accessibility audit measured against WCAG 2.1 or 2.2 success criteria -- color contrast ratios, ARIA landmark roles, focus order, keyboard navigation compliance (use accessibility-review)
• The user wants a visual design critique focused on typography hierarchy, color palette harmony, grid alignment, or brand consistency (use visual-hierarchy-review)
• The user wants to map the end-to-end user journey, identify where flows diverge, detect dead ends, or audit navigation architecture (use user-flow-mapping)
• The user wants to design a new interface from scratch and needs wireframe specs or layout guidance (use wireframe-specification)
• The user wants to analyze quantitative data -- funnel drop-off rates, heatmaps, session recordings, or A/B test results without an interface to evaluate (use analytics-interpretation)
• The user wants to conduct usability testing with real participants -- recruiting, scripting, moderating (use usability-test-planning)
• The request is purely about content strategy, microcopy tone, or brand voice rather than interface usability (use content-strategy)

Process

Step 1: Gather Context Before Beginning

Never begin an audit without sufficient context. A heuristic evaluation without user context produces findings that are technically correct but strategically useless.

• Identify the product surface: Ask for the specific screen, flow, or product being audited. Accept URLs, annotated screenshots, written descriptions of the interface, or a combination. If the user provides a URL, describe what you observe before evaluating -- don't assume.
• Establish the target user profile: Who are the primary users? Novice vs. expert distinction is critical -- a violation of H7 (flexibility and efficiency) matters far more for a tool used by power users 40 hours a week than for a consumer app used once a month. Ask about technical literacy, domain expertise, and usage frequency.
• Define the critical task set: Identify 3-5 core tasks users must accomplish. These become your severity multipliers -- a violation that blocks a core task scores higher than an identical violation in a rarely used secondary feature.
• Collect known signals: Ask about existing evidence: support tickets, NPS comments, user research findings, drop-off rates at specific steps, or help documentation requests. These validate and prioritize findings rather than drive them.
• Agree on audit scope: Clarify whether this is a full product audit, a focused flow audit (e.g., only the checkout), a single-screen deep dive, or a comparison audit (current vs. proposed). Different scopes require different output depth. A full product audit should not attempt to evaluate 50 screens in equal depth -- identify the 5-10 highest-traffic or highest-stakes screens.

Step 2: Apply Nielsen's 10 Usability Heuristics Systematically

Evaluate every heuristic in sequence. Do not skip heuristics that seem obviously fine -- passing heuristics should be marked as Pass with a brief rationale. Experienced evaluators find more violations when they work through each heuristic in isolation rather than doing a general walkthrough.