Most redesigns start with a problem the team can feel but can’t precisely locate. Conversion is low. Users aren’t activating. The product looks dated. So a brief goes out, new layouts get sketched, and months later the team discovers the new version has different problems — because nobody diagnosed the original ones.
This checklist prevents that. It’s organized around the 6-Layer UX Audit Framework we use at Masterly — each layer surfaces a different category of failure, and each category connects to different business outcomes. Run the full checklist before a redesign, or use individual layers as targeted diagnostics when something specific isn’t working.
For each question, classify what you find using the Masterly Impact Tiers:
- Strength — working well, worth explicitly preserving
- Friction — usable but creating unnecessary drag on engagement or conversion
- Blocker — active failure costing you users, conversions, or revenue right now
- Opportunity — not broken, but a meaningful improvement is available
Layer 1: Design Review (Q1–Q8)
Design review evaluates screens against usability principles — hierarchy, clarity, consistency, feedback. Most audits start here. Most audits also stop here, which is why they miss most of what’s actually affecting business outcomes. Use it as a foundation, not a conclusion.
| # |
Question |
How to check |
Typical tier |
| Q1 |
Does the homepage communicate what the product is and who it's for within 5 seconds? |
Show to someone unfamiliar — ask them to describe what it does without prompting |
Blocker |
| Q2 |
Is there a clear visual hierarchy on each key screen? |
Screenshot each key screen and squint — one element should dominate visually |
Friction |
| Q3 |
Are interactive elements visually distinct from non-interactive content? |
Tab through the page using keyboard only — note anything ambiguous |
Blocker |
| Q4 |
Does the navigation reflect how users think, not how the team is organized? |
Card sorting with 5 users unfamiliar with the product |
Friction |
| Q5 |
Are error, empty, and loading states designed? |
Trigger each state manually — check for generic text or blank screens |
Friction |
| Q6 |
Is the typography system applied consistently across all page templates? |
Compare 5 different page types side by side — heading sizes, line heights, spacing |
Friction |
| Q7 |
Is the color system semantically consistent? |
List every use of the primary CTA color — anything that isn't a CTA is a failure |
Friction |
| Q8 |
Are primary actions visually stronger than surrounding promotional content? |
Assess visual weight of main CTA vs nearest ad or upsell block |
Blocker |
Real audit finding: In one product, a sponsor’s promotional block had stronger visual weight than the site’s own navigation. The first thing users saw on load was an advertisement — not the content they came for.
Layer 2: Information Architecture (Q9–Q15)
IA problems are almost never visible in individual screens. They emerge when you map the full structure: every page type, every navigation path, every dead end. The tools here are crawlers and user research, not screen review.
| # |
Question |
How to check |
Typical tier |
| Q9 |
Can a visitor predict where each navigation item leads before clicking? |
5-second test with new users — ask them where each nav item goes |
Friction |
| Q10 |
Are category and topic pages curated hubs — or just filtered lists? |
Check your 3 highest-traffic category pages for editorial curation and sub-topic paths |
Opportunity |
| Q11 |
Does every page have a designed next step? |
Read the last line of your 10 most-visited pages — what's the designed exit? |
Blocker |
| Q12 |
Is the footer a navigation aid or a dead end? |
Check if footer offers meaningfully different paths than the main nav |
Opportunity |
| Q13 |
What percentage of core content is buried deeper than 3 clicks from the homepage? |
Screaming Frog crawl → filter by crawl depth |
Friction |
| Q14 |
Are there orphaned pages with no inbound internal links? |
Screaming Frog → Reports → Orphaned Pages |
Friction |
| Q15 |
Does the IA support the business model — do the most-visited pages connect to conversion? |
Map your 10 highest-traffic pages to the nearest conversion point |
Blocker |
Real audit finding: A media platform with 100,000 monthly visitors had no structured path from article completion to a related article. Users who finished reading had one designed option: close the tab. Over 17,000 pages were buried deeper than 3 clicks from the homepage.
Layer 3: Monetization and Conversion (Q16–Q22)
This is the layer most design teams skip — which is precisely why it contains the highest-value findings in most audits. Every item here connects directly to revenue or activation.
| # |
Question |
How to check |
Typical tier |
| Q16 |
Is the primary CTA visible above the fold on mobile without scrolling? |
Test on a real phone — note whether the CTA is visible before any scroll |
Blocker |
| Q17 |
Is advertising clearly labeled and visually separated from editorial content? |
Screenshot a content page — assess in 2 seconds what's ad vs editorial |
Blocker |
| Q18 |
Is there a designed conversion moment in the user journey? |
Map the user flow — identify the exact point where an upgrade or CTA triggers |
Friction |
| Q19 |
Are upgrade or subscription prompts triggered after value delivery? |
Walk through the trial flow — note where the first paywall or upgrade prompt appears |
Friction |
| Q20 |
Does the pricing or upgrade page reduce anxiety? |
Check for: tier comparison, payment terms, refund policy, social proof |
Friction |
| Q21 |
What happens to a user who completes the first value action but doesn't convert? |
Walk through trial to completion — note the next designed prompt, if any |
Blocker |
| Q22 |
Is there a recovery mechanism for users who abandon mid-flow? |
Drop off mid-checkout or mid-signup — check for email, session persistence, or saved state |
Opportunity |
Principle: Ask users to pay before they’ve experienced value and you train them to distrust the product. The sequence that converts: first action — value delivery — upgrade prompt. Products that gate core features before first use systematically lose the users most likely to pay.
Layer 4: Technical Performance (Q23–Q29)
A UX audit that ignores performance is diagnosing half the patient. Slow pages create user experience failures that no design improvement can compensate for — you cannot redesign your way out of a 22-second load time.
| # |
Question |
How to check |
Typical tier |
| Q23 |
What is the Largest Contentful Paint (LCP) on mobile? |
Google PageSpeed Insights → Mobile tab. Good: <2.5s. Poor: >4s |
Blocker if >4s |
| Q24 |
What is the Cumulative Layout Shift (CLS) score? |
PageSpeed Insights → CLS. Good: <0.1. Poor: >0.25 |
Blocker if >0.25 |
| Q25 |
Are images served in modern formats (WebP/AVIF) and sized correctly? |
PageSpeed Insights → Opportunities → "Serve images in modern formats" |
Friction |
| Q26 |
Are render-blocking CSS and JavaScript files deferred? |
PageSpeed Insights → "Eliminate render-blocking resources" |
Friction |
| Q27 |
Is full-page caching configured? |
GTmetrix → Waterfall → check TTFB. Should be under 200ms |
Friction |
| Q28 |
Are there 5xx server errors reachable through normal navigation? |
Screaming Frog → Response Codes → 5xx filter |
Blocker |
| Q29 |
Is mobile performance tested at realistic network conditions? |
Lighthouse → Mobile mode → Slow 4G throttling |
Friction |
Real audit finding: Mobile LCP of 22.2 seconds. CLS of 0.405 — elements jumping while the page loaded, causing users to tap the wrong targets. 11,831 internal pages returning 5xx errors, reachable through ordinary navigation. No design change addresses any of these.
Layer 5: SEO Health (Q30–Q36)
For content and SaaS products, SEO health directly affects organic traffic volume, which affects every other metric downstream. A product with strong UX and poor SEO health is invisible to the audience it’s designed for.
| # |
Question |
How to check |
Typical tier |
| Q30 |
Does every important page have a unique title tag and meta description? |
Screaming Frog → Page Titles → filter duplicates and missing |
Friction |
| Q31 |
Are there duplicate pages from URL parameters, pagination, or auto-generated archives? |
Google Search Console → Coverage → Excluded → Duplicate without canonical |
Blocker |
| Q32 |
Are there HTTP URLs or mixed-content issues on an HTTPS site? |
Screaming Frog → Response Codes → Protocol filter |
Blocker |
| Q33 |
Is structured data implemented on pages that qualify? |
Google Rich Results Test → check for Article, FAQ, Breadcrumb schema |
Opportunity |
| Q34 |
Do internal links use descriptive anchor text? |
Screaming Frog → Bulk Export → All Inlinks → filter "click here" and "read more" |
Friction |
| Q35 |
Are the most important pages reachable within 3 clicks from the homepage? |
Screaming Frog → crawl depth report → flag anything beyond depth 3 |
Friction |
| Q36 |
Is the XML sitemap current, submitted to Search Console, and free of errors? |
Google Search Console → Sitemaps → check for errors or excluded URLs |
Friction |
Real audit finding: A Screaming Frog crawl at 52% completion had already discovered 326,000+ URLs — far above what the site’s content volume warranted. 25,008 mixed-content issues and 12,087 permanent redirects, each fragmenting link authority and consuming crawl budget.
Layer 6: Analytics (Q37–Q42)
Analytics validates or challenges everything the other five layers surface. Without it, you’re working from observation and inference. With it, you’re working from evidence — and you can prioritize by actual impact, not by assumption.
| # |
Question |
How to check |
Typical tier |
| Q37 |
Are conversion events configured to track real business goals — not just pageviews? |
GA4 → Events → verify sign-up, activation, and payment events exist |
Blocker |
| Q38 |
What is pages-per-session — and is it appropriate for the product type? |
GA4 → Engagement → Pages & Screens → avg pages per session by user type |
Friction |
| Q39 |
What is the mobile-to-desktop split — does it match category benchmarks? |
GA4 → Tech → Tech Overview → Device Category |
Friction |
| Q40 |
Are there pages with above-average bounce rates that don't warrant it by content type? |
GA4 → Pages → segment bounce rate by page type |
Friction |
| Q41 |
Is there funnel tracking on every conversion flow? |
GA4 → Explore → Funnel Exploration → check for gaps in step-level data |
Blocker |
| Q42 |
Can you attribute revenue or signups to specific traffic sources and landing pages? |
GA4 → Acquisition → Traffic Acquisition → conversion column by source/medium |
Friction |
Real audit finding: A media platform showed 78.3% desktop / 21% mobile — an unusual split for a content site where the industry benchmark is 60–70% mobile. Mobile users weren’t returning. The likely cause: mobile performance failures suppressing repeat visits before analytics could even record them.
How to Use Your Results
Once you’ve worked through the checklist, you have classified findings across six layers. The prioritization logic is straightforward: fix Blockers first, regardless of implementation cost. These are active failures — users are being lost, misled, or frustrated right now.
After Blockers, address high-impact Frictions starting with the ones that are cheapest to fix. Quick wins build momentum and demonstrate audit ROI before the larger structural work begins.
Structural problems — IA redesign, performance overhaul, SEO cleanup — belong in a parallel workstream, not competing with feature work. A platform with a 47/100 mobile performance score can’t absorb new features without making the underlying problem worse.
One final principle from real audit work: document your Strengths with the same rigor you apply to failures. Products that redesign without an explicit list of what’s working often accidentally remove the things quietly responsible for their best metrics. Clean reading experience, no intrusive modals, fast checkout — these are competitive advantages. Easy to miss. Easy to remove.
If you've worked through this checklist and found more Blockers and Frictions than you expected — that's the point.
A Masterly UX audit gives you a prioritized roadmap across all six layers, with findings classified by urgency and mapped to the business outcomes they affect.
Talk to us about a UX audit →
Related: How to Conduct a UX Audit (And What to Do with the Findings) · The Complete Guide to SaaS Onboarding UX · Fintech Dashboard Design: Why Most Look the Same