Full site audit — performance, SEO, UX, conversion. Actionable fixes.
You know your site has problems but you don't know which ones cost you money. Generic Lighthouse runs flag 80 issues — most don't matter, a few are killing conversions. This tool runs a structured conversion + UX + performance audit on your live site and prioritizes findings by likely revenue impact, not technical severity. The AI looks at your homepage and key landing pages and flags concrete issues: above-the-fold copy that doesn't match search intent, missing trust signals on the pricing page, CTAs that compete, forms with friction, mobile layout problems, Core Web Vitals issues that actually affect ranking, and accessibility gaps that block real users. Every finding includes a 'why this matters' and a specific fix you can hand to a dev, not vague advice like 'improve user experience'.
Brief used: “Site: acmehq.com (made up). B2B SaaS, project management for construction crews. $99/mo plan, free trial. Main traffic source: Google Ads + SEO. Conversion rate from /pricing to trial signup is 1.2%, want to know why. We have a homepage, /features, /pricing, /integrations, and 30 blog posts.”
## Website Audit — Construction PM SaaS **Pages reviewed:** /, /features, /pricing, /integrations **Audit framework:** conversion + UX + Core Web Vitals + accessibility ### CRITICAL (fix this week) **1. Homepage H1 doesn't say what you do** — page: / Current H1: 'Built for builders'. A visitor lands cold from a Google Ad and has no idea this is software, what it does, or who it's for. Time-to-clarity is ~6 seconds based on scroll-depth heuristics, which is past the 3-second bounce threshold for paid traffic. *Fix:* H1 should name the category (project management software) and the audience (construction crews) explicitly. Sub-headline can carry the personality. *Effort:* 1 hour **2. Pricing page CTA collision** — page: /pricing Three CTAs in the same viewport: 'Start free trial', 'Talk to sales', 'Compare plans'. They have equal visual weight. Visitors freeze. *Fix:* Promote 'Start free trial' to primary (filled button, brand color). Demote 'Talk to sales' to a text link below. Move 'Compare plans' to a tab. *Effort:* 2 hours **3. LCP = 4.1s on mobile** — page: / Hero background image is 1.8MB unoptimized PNG. Largest Contentful Paint fails Core Web Vitals threshold (>2.5s). This affects ranking AND bounce rate from Google Ads (Quality Score penalty). *Fix:* Convert to WebP, serve responsive sizes, preload. Target LCP <2.0s. *Effort:* Half day ### HIGH **4. No trust signals above the fold on /pricing** No logos, no testimonials, no security badges, no guarantee. For a $99/mo commitment from a skeptical construction buyer, this is a friction point. *Fix:* Add 'No credit card required' under the trial CTA, plus 3-4 customer logos or a single high-credibility testimonial. *Effort:* 1 day **5. Trial signup form asks for company size on step 1** Forcing a dropdown choice before email increases drop-off. Move to step 2 or onboarding. *Effort:* 2 hours *(8 more HIGH findings, 12 MEDIUM, 6 nice-to-have)* ### Top 5 to Fix This Week 1. Homepage H1 rewrite 2. Pricing CTA hierarchy 3. Hero image optimization 4. Pricing page trust signals 5. Trial form field reduction
Static example — your run uses Claude live on your specific brief.
SaaS / DTC / agency / portfolio sites doing 1k-100k visitors/month who suspect their site under-converts but can't afford a $5k audit from a CRO consultant. Especially useful pre-launch (catch the obvious stuff before paid ads) or after a redesign (validate the new site actually works). Not for sites with <100 visitors/mo — you need data to know what to fix. Not a substitute for proper user testing on a high-revenue product — but a sharp first pass that catches 80% of common issues.
A prioritized findings report sorted by impact (critical / high / medium / nice-to-have). Each finding includes: the page and element affected, what's wrong, why it hurts conversion or rank, the recommended fix with implementation notes, and an estimated effort level (1 hour / half day / sprint). Sections covered: above-the-fold messaging clarity, value-prop hierarchy, CTA placement and copy, trust signals (social proof, guarantees, security), pricing page friction, mobile experience, Core Web Vitals interpretation (LCP / CLS / INP), accessibility blockers (contrast, alt text, keyboard nav), and broken / dead-end pages. Closes with a 'top 5 to fix this week' list.
You're 2 weeks from a public launch and want to catch the embarrassing stuff. The audit flags the broken signup form on Safari mobile, the pricing page that doesn't load fonts, and the hero copy that doesn't say what you do — before Product Hunt sees it.
You're spending $3k/mo on Google Ads with a 0.8% conversion rate. The audit looks at the landing page and flags message mismatch with ad copy, a CTA below the fold, no trust signals, and a form asking for 8 fields when 3 would do.
Your site evolved through 6 months of feature launches and now has 4 competing CTAs in the hero. The audit identifies the accumulated UX debt and ranks what to fix first based on revenue impact, not aesthetic preference.
The AI analyzes the public HTML, computed styles, and visible content of the pages you specify. It does not run a full crawler like Screaming Frog — for that, use a dedicated crawler. The value here is qualitative judgment about conversion and messaging, which crawlers don't do.
Lighthouse tells you your LCP is 4.1s and your contrast ratio fails on 3 elements. It does not tell you that your H1 is meaningless or that your pricing page has 3 competing CTAs. This audit covers conversion and messaging issues Lighthouse can't see, and prioritizes the technical issues by actual user impact.
No. It catches the common high-impact problems — the ones a sharp consultant would flag in the first hour. It won't catch issues that require user research, A/B testing, or session recording analysis. Treat it as a starting point, not the final word.
Yes — point it at any public URL. Useful for understanding why a competitor converts better than you do, or for spotting their weak pages before you compete on the same keywords.
Yes. You get an anonymous preview instantly with no signup. Drop your email and you unlock 3 full-length runs per month for Website Audit — no credit card. Unlimited runs are $79 one-time, or $19/mo for every tool.
Paid ($79 one-time) unlocks unlimited runs for Website Audit, longer outputs from Claude Sonnet, full exports, and priority generation. $19/mo unlocks every tool on JustNeeda.
Free runs render in-browser and can be copy-pasted. Paid unlocks copy-to-clipboard, Markdown, and plain-text exports — and history of every run tied to your account.
No. Every run hits Claude live with your specific input. We don't reuse outputs across users. Your input stays private to your session and account.