/mktg:cro-auditCRO Audit
Charge la page avec Chrome DevTools MCP (pour voir le vrai rendu, pas juste le HTML), évalue 8 dimensions avec un scoring /10 par axe. Cite les passages concernés, propose des fixes concrets rangés par priorité P0/P1/P2. Ne devine pas : toutes les affirmations sont sourcées sur un élément visible de la page. Sort un rapport de ~500 mots actionnable dans la demi-journée.
Marketers ou PMs qui doivent défendre une refonte de landing, benchmarker un concurrent, ou comprendre pourquoi la conversion stagne. Nettement plus précis avec le MCP Chrome DevTools branché : sans lui, l'audit se limite au HTML et rate ce qui se passe au rendu (CLS, interactions, mobile réel).
Configuration
mktg:cro-auditReadWriteWebFetchChrome DevTools<URL to audit>Installation
$ mkdir -p ~/.claude/skills/mktg:cro-audit && \
curl -sSL https://www.cedricrittie.com/api/skills/mktg-cro-audit/raw \
-o ~/.claude/skills/mktg:cro-audit/SKILL.mdPuis redémarre Claude Code. Test avec `/mktg:cro-audit`.
Besoin d'installer Claude Code d'abord ? Voir la fiche Claude Code · Télécharger le .md brut
Exemples d'utilisation
/mktg:cro-audit cedricrittie.comAudit complet : 8 scores, 3-5 fixes P0, 5-8 fixes P1, captures écran annotées si Chrome DevTools MCP est installé.
Audit CRO — cedricrittie.com Global : 72/100 Headline 8/10 — promesse claire CTA 6/10 — CTA unique, libellé flou Preuve sociale 4/10 — aucune preuve above fold Urgence 3/10 — aucun signal de rareté/temps Confiance 7/10 — bio auteur présente Friction 9/10 — pas de formulaire, juste du scroll Mobile 8/10 — lisible, tailles de police OK Vitesse 9/10 — LCP 1.2s Fixes P0 (bloquants) 1. Preuve sociale — ajoute un one-liner avec une métrique (abonnés, articles lus, témoignage) above fold 2. CTA — "Explorer" est flou. Teste "Lire le dernier article →" Fixes P1 [...] Fixes P2 [...]
/mktg:cro-audit https://competitor.comBenchmark concurrent. Utile pour défendre un choix de design ou identifier ce qu'ils font mieux que toi.
Audit concurrent — competitor.com Global : 81/100 [...] Écarts clés vs cedricrittie.com - Ils te battent sur la preuve sociale (9 vs 4) - Tu les bats sur la vitesse (9 vs 6) et la friction (9 vs 5)
Le Skill en entier
CRO Audit — Landing Page Conversion Scoring
Scores a web page on 8 conversion dimensions. Produces a grade (A+ to F), prioritized fixes, and benchmark comparisons.
Input
- A URL (required)
- Optional: industry context for benchmark comparison (SaaS, ecommerce, media, consumer app)
- If no URL provided, ask for one
Step 1: Capture the Page
Use the available tools to analyze the page:
- Navigate to the URL and take a screenshot (desktop + mobile viewport)
- Fetch the HTML content via WebFetch for text analysis
- If Chrome DevTools MCP is available:
- Take screenshots at desktop (1440px) and mobile (390px) widths
- Run a Lighthouse audit for performance/accessibility data
- Evaluate DOM for CTA count, form fields, trust badges, social proof elements
If Chrome DevTools is not available, fall back to WebFetch + HTML analysis only.
Step 2: Score 8 Dimensions
Each dimension is scored 0-100. The weighted aggregate gives the overall score.
1. Headline Clarity (weight: 15%)
- Is the value proposition clear in under 5 seconds?
- Does the headline answer "what is this and why should I care?"
- Is it specific (numbers, outcomes) or vague ("the best solution")?
- 90+: Specific outcome + target audience in headline. "Read 100 feeds in one place. No algorithms."
- 70-89: Clear but generic. "The modern RSS reader."
- 50-69: Vague or feature-focused. "A powerful feed aggregation platform."
- <50: No clear value prop, or headline is the product name only.
2. CTA Visibility (weight: 20%)
- Is there a primary CTA above the fold?
- Is the CTA text action-oriented and specific? ("Start reading free" > "Sign up")
- Is there only ONE primary CTA? (multiple = confusion)
- Is the CTA visually distinct (contrast, size, whitespace)?
- 90+: Single clear CTA above fold, action-specific text, high contrast, repeated at scroll points
- 70-89: CTA present but generic text or low contrast
- 50-69: CTA below fold or competing with secondary CTAs
- <50: No clear CTA or CTA hidden
3. Social Proof (weight: 15%)
- User count, testimonials, logos, ratings, press mentions?
- Are testimonials specific (name, role, company) or anonymous?
- Are numbers real and verifiable?
- 90+: Named testimonials with photos + specific metrics + recognizable logos
- 70-89: Some social proof but generic ("loved by thousands")
- 50-69: Weak social proof or buried below fold
- <50: No social proof at all
4. Urgency (weight: 5%)
- Is there a reason to act now? (limited spots, launch pricing, beta access)
- Is the urgency real or manufactured?
- NOTE: Not all pages need urgency. Score 60 (neutral) if urgency doesn't apply. Only penalize if urgency would clearly help but is missing.
- 90+: Real, specific urgency (countdown, limited spots with number)
- 70-89: Soft urgency ("Early access", "Beta")
- 50-69: No urgency but not needed
- <50: Urgency would help but is completely absent
5. Trust Signals (weight: 10%)
- Security badges, privacy policy link, HTTPS, team/company info?
- Is there a real human/company behind this? (about page, team photos, address)
- Free trial / money-back guarantee / no credit card needed?
- 90+: Multiple trust layers (SSL + privacy + real team + guarantee + social proof)
- 70-89: Basic trust (SSL, privacy link)
- 50-69: Minimal trust signals
- <50: Feels sketchy or anonymous
6. Form Friction (weight: 15%)
- How many fields in the signup/contact form?
- Can you start with just an email? Or Google/social login?
- Is there progressive disclosure (don't ask everything upfront)?
- 90+: One-click signup (Google/Apple) or single email field
- 70-89: 2-3 fields, clear labels
- 50-69: 4+ fields or confusing labels
- <50: Long form, required phone number, or unclear what happens after submit
7. Mobile Responsiveness (weight: 10%)
- Does the layout work on mobile (390px)?
- Is the CTA tappable (44px+ touch target)?
- Is text readable without zooming?
- Do images/videos resize properly?
- 90+: Flawless mobile experience, CTA thumb-reachable
- 70-89: Works but some elements feel cramped
- 50-69: Layout breaks or CTA hard to find on mobile
- <50: Not mobile-friendly
8. Page Speed Indicators (weight: 10%)
- Does the page feel fast? (no heavy animations, lazy-loaded images, minimal JS)
- Lighthouse performance score if available
- Large hero images or videos that block rendering?
- 90+: Lighthouse 90+, instant load feel, no layout shift
- 70-89: Loads in 2-3 seconds, minor layout shifts
- 50-69: Noticeable lag, heavy assets
- <50: Slow, blocking resources, poor mobile performance
Step 3: Calculate Overall Score
Overall = (Headline x 0.15) + (CTA x 0.20) + (Social Proof x 0.15) +
(Urgency x 0.05) + (Trust x 0.10) + (Friction x 0.15) +
(Mobile x 0.10) + (Speed x 0.10)
Grade scale:
- A+ (95-100): Exceptional. Ship and monitor.
- A (90-94): Strong. Minor polish only.
- B+ (85-89): Good. 1-2 improvements would make a real difference.
- B (80-84): Solid foundation with clear gaps.
- C (70-79): Significant issues. Fix before major traffic push.
- D (60-69): Major problems. Needs redesign of weak areas.
- F (<60): Fundamental issues. Don't spend on traffic until fixed.
Step 4: Industry Benchmarks
Compare against typical scores for the industry:
| Industry | Avg Score | Top 10% |
|---|---|---|
| SaaS | 72 | 88+ |
| E-commerce | 68 | 85+ |
| Consumer App | 65 | 82+ |
| Media/Content | 60 | 78+ |
| Agency/Services | 70 | 86+ |
Step 5: Output
---
## CRO Audit: [URL]
**Date**: [YYYY-MM-DD]
**Industry**: [detected or specified]
**Overall Score**: XX/100 (Grade: X)
**Benchmark**: [vs industry avg and top 10%]
### Dimension Scores
| # | Dimension | Score | Weight | Weighted |
|---|-----------|-------|--------|----------|
| 1 | Headline Clarity | XX/100 | 15% | XX |
| 2 | CTA Visibility | XX/100 | 20% | XX |
| 3 | Social Proof | XX/100 | 15% | XX |
| 4 | Urgency | XX/100 | 5% | XX |
| 5 | Trust Signals | XX/100 | 10% | XX |
| 6 | Form Friction | XX/100 | 15% | XX |
| 7 | Mobile | XX/100 | 10% | XX |
| 8 | Page Speed | XX/100 | 10% | XX |
### Top 3 Priority Fixes
1. **[Dimension]** (current: XX -> target: XX)
- What's wrong: [specific issue]
- Fix: [specific actionable recommendation]
- Impact: [expected improvement]
2. ...
3. ...
### Dimension Details
[For each dimension: what was observed, what's good, what needs work. 2-3 lines each max.]
### Screenshots
[Reference desktop and mobile screenshots if taken]
---
Rules
- Be specific in feedback. "Your CTA says 'Submit'" is useful. "Improve your CTA" is not.
- Reference what you actually see on the page, not generic advice.
- If you can't assess a dimension (e.g., no mobile screenshot available), score it as "N/A" and exclude from the weighted average.
- Don't fabricate Lighthouse scores. If you can't run it, say so and assess speed qualitatively.
- Urgency is the one dimension where a neutral score (60) is fine. Not every page needs urgency.
Version publique de ce Skill. 185 lignes. Copie-colle dans ~/.claude/skills/mktg:cro-audit/SKILL.md pour l'installer.
Skills liés
/mktg:expert-panelScore n'importe quel contenu avec un panel de 7-10 experts simulés. Itère jusqu'à un score de 90/100 (max 3 rounds). Le Humanizer est toujours inclus à 1.5x pour détecter le slop.
/mktg:trend-scoutScanne Hacker News, Reddit, Google Trends et X pour trouver les tendances pertinentes sur un domaine. Score chaque tendance et suggère des angles de contenu.
/content:planOrchestre la stratégie de contenu. Scanne les articles, les clippings, les tendances et X pour décider quoi publier, quand et pourquoi. Génère des calendriers éditoriaux.