Skill: growth-cro
Conversion rate optimization (CRO) and funnel analysis. Trigger when the user wants to optimize conversions, map/analyze a conversion funnel, improve a signup form, a checkout, a landing page, or an onboarding/activation flow.
Configuration
| Property | Value |
|---|---|
| Context | fork |
| Allowed tools | Read, Glob, Grep, Edit, Write |
| Keywords | growth, cro, how, connect your github repo, create your first project, start from a template, create account |
Detailed description
Conversion Rate Optimization (CRO)
Goal
Identify and fix friction points in user journeys to maximize the conversion rate.
CRO Domains
┌──────────────────────────────────────────────────────────────────┐
│ CRO FRAMEWORK │
├──────────────────────────────────────────────────────────────────┤
│ │
│ PAGE CRO Optimize marketing/landing pages │
│ ═════════ │
│ │
│ SIGNUP FLOW CRO Improve signup and account creation │
│ ══════════════ │
│ │
│ ONBOARDING CRO Reduce time-to-value post-signup │
│ ═════════════ │
│ │
│ FORM CRO Optimize capture forms │
│ ════════ │
│ │
│ POPUP CRO Improve popups, modals, overlays │
│ ═════════ │
│ │
│ PAYWALL CRO Optimize paywalls and upsells │
│ ═══════════ │
│ │
└──────────────────────────────────────────────────────────────────┘
1. Page CRO
Landing page checklist
| # | Element | Best practices |
|---|---|---|
| 1 | Headline | Clear benefit in < 10 words, no jargon |
| 2 | Sub-headline | Explain the "how" in 1 sentence |
| 3 | Primary CTA | Specific action, visible above-the-fold |
| 4 | Social proof | Testimonials, customer logos, numbers |
| 5 | Objections | FAQ or sections answering doubts |
| 6 | Urgency/Scarcity | Timer, limited spots (if authentic) |
| 7 | Hero visual | Product screenshot or demo video |
| 8 | Navigation | Minimal (no full menu on landing) |
Conversion patterns
Hero section (benefit + CTA)
↓
Social proof (logos, testimonials)
↓
Features/Benefits (3-5 max)
↓
How it works (3 steps)
↓
Pricing (if applicable)
↓
FAQ (common objections)
↓
Final CTA (same action as the hero)
2. Signup Flow CRO
Golden rules
| # | Rule | Impact |
|---|---|---|
| 1 | Minimum fields (email only to start) | +20-30% signups |
| 2 | Social login (Google, GitHub) | +15-25% signups |
| 3 | No blocking email confirmation | -40% drop-off |
| 4 | Progress indicator if multi-step | +10% completion |
| 5 | Value proposition visible next to the form | +15% signups |
| 6 | Password strength indicator | +5% completion |
| 7 | Inline error, not at submit | +20% completion |
Anti-patterns to avoid
- Asking for too much info at signup (name, phone, address)
- CAPTCHA visible for all users
- Confirmation email before product access
- Redirect to login page after signup
- No feedback after form submission
3. Onboarding CRO
Time-to-Value framework
Signup → [Activation] → [Aha Moment] → [Habit Formation]
| | |
v v v
First setup First value Regular usage
(< 2 min) (< 5 min) (Day 7+)
Effective patterns
| Pattern | When | Example |
|---|---|---|
| Checklist | 3-5 activation steps | "Complete your profile: 3/5" |
| Wizard | Technical setup required | "Connect your GitHub repo" |
| Empty state | First visit, empty page | "Create your first project" |
| Template | Complex product | "Start from a template" |
| Tour guide | Complex interface | Discovery tooltips |
4. Form CRO
Form optimization
| # | Technique | Detail |
|---|---|---|
| 1 | One field per line | No multi-column layout on mobile |
| 2 | Labels above fields | No floating labels |
| 3 | Correct input type | email, tel, number for adapted keyboard |
| 4 | HTML autocomplete | autocomplete="email", "given-name" |
| 5 | Font size >= 16px | Avoids iOS zoom |
| 6 | Descriptive submit button | "Create account" not "Submit" |
| 7 | Immediate feedback | Validate on blur, not at submit |
| 8 | Easy error recovery | Message + focus on the field in error |
5. Popup/Modal CRO
Rules
| # | Rule | Detail |
|---|---|---|
| 1 | Timing: not before 30s or 50% scroll | Let users discover the content |
| 2 | Exit-intent > time-based | Less intrusive |
| 3 | Easy to close | Visible X, click outside, Escape |
| 4 | One popup at a time | No stack of modals |
| 5 | Limited frequency | Max 1x per session or 1x per week |
| 6 | Clear value proposition | Not just "Subscribe" |
| 7 | Mobile: bottom sheet > centered modal | Better mobile UX |
6. Paywall/Upgrade CRO
Strategies
| Strategy | Detail | When |
|---|---|---|
| Feature gate | Show the feature, block access | Premium feature requested |
| Usage limit | "3/5 projects used" | Free tier limit approached |
| Trial expiration | Countdown + demonstrated value | End of trial |
| Upgrade prompt | Contextual suggestion | Premium action attempted |
| Social proof | "Join 10,000+ teams" | Pricing page |
Pricing page patterns
[ Free ] [ Pro ★ ] [ Enterprise ]
$0 $29/mo Custom
3 projects Unlimited Unlimited
1 user 10 users Unlimited
Basic support Priority Dedicated
[Start trial]
- Highlight the recommended plan (badge, color)
- Monthly/annual toggle (show the savings)
- Feature comparison table below
- FAQ on billing
7. Funnel mapping & analysis
Before optimizing a single step, map and measure the whole funnel.
- Map the steps — events, pages, actions from entry to conversion.
- Measure — conversion per step and the global rate.
| Step | Users | Conv. | Drop-off | Trend |
|---|---|---|---|---|
| ... | ... | ... | ... | ... |
- Segment — device, source, country, cohort (drop-offs often hide in a segment).
- Diagnose drop-offs — friction (effort), anxiety (trust), clarity (confusion) — ask why, not just how much.
- Prioritize — score opportunities with ICE (Impact × Confidence × Ease); optimize one step at a time.
- Monitor — continuous tracking + alerts on step regressions.
Reliable tracking is a prerequisite — never analyze a funnel on unreliable data.
Metrics to track
| Metric | Formula | Target |
|---|---|---|
| Conversion rate | Conversions / Visitors | Depends on the domain |
| Drop-off rate | Drop-offs per funnel step | < 20% per step |
| Time to convert | Visit duration → conversion | Reduce |
| Bounce rate | Bounces / Sessions | < 40% landing pages |
| Activation rate | Activated users / Signups | > 40% |
| Trial-to-paid | Payments / Trials | > 15% |
Expected output
## CRO Audit: [Page/Flow]
### Current estimated conversion rate: X%
### Improvement potential: +Y%
### Quick wins (immediate impact)
1. [Action] - Estimated impact: +X%
2. [Action] - Estimated impact: +X%
### Structural improvements
1. [Action] - Detail and implementation
### Recommended A/B tests
1. [Hypothesis] - Variant A vs B
Rules
- Always base recommendations on data or proven best practices
- Propose quick wins AND structural changes
- Do not sacrifice UX for conversion (dark patterns forbidden)
- Suggest A/B tests to validate important changes
Automatic triggering
This skill is automatically activated when:
- The matching keywords are detected in the conversation
- The task context matches the skill's domain
Triggering examples
- "I want to growth..."
- "I want to cro..."
- "I want to how..."
Context fork
Fork means the skill runs in an isolated context:
- Does not pollute the main conversation
- Results are returned cleanly
- Ideal for autonomous tasks
Practical examples
1. Example: Landing Page A/B Test Optimization
Example: Landing Page A/B Test Optimization
Scenario
A SaaS product landing page has a 1.8% signup conversion rate. Goal: reach 3%+ through systematic A/B testing.
Current State Analysis
| Metric | Value | Benchmark |
|---|---|---|
| Visitors/month | 25,000 | - |
| Signup rate | 1.8% | 3-5% (SaaS avg) |
| Bounce rate | 72% | < 50% |
| Avg time on page | 18s | > 45s |
| CTA clicks | 3.2% | > 8% |
Identified Issues
- Headline: Feature-focused ("AI-powered analytics platform") instead of benefit-focused
- CTA: "Get Started" is vague, below the fold on mobile
- Social proof: None visible above the fold
- Form friction: 6-field signup form requiring credit card
A/B Test Plan
Test 1: Headline (highest impact)
Control (A): "AI-Powered Analytics Platform"
Variant (B): "Cut Your Reporting Time by 75%"
Variant (C): "Stop Wasting 10 Hours/Week on Reports"
Metric: Signup rate
Traffic split: 33/33/33
Min sample: 3,000 per variant (95% confidence, 80% power)
Duration: ~12 days at current traffic
Test 2: CTA Copy + Placement
Control (A): "Get Started" (below fold)
Variant (B): "Start Free Trial - No Credit Card" (above fold)
Metric: CTA click rate
Min sample: 2,500 per variant
Duration: ~6 days
Test 3: Social Proof
Control (A): No social proof above fold
Variant (B): Logo bar "Trusted by 500+ companies" + 3 logos
Variant (C): Testimonial quote + photo + company name
Metric: Signup rate
Duration: ~12 days
Test 4: Form Simplification
Control (A): 6 fields + credit card
Variant (B): Email + password only (2 fields)
Variant (C): "Sign up with Google" single button
Metric: Form completion rate
Duration: ~10 days
Expected Impact Model
Test 1 (Headline): +0.3-0.5% conversion lift
Test 2 (CTA): +0.2-0.4% conversion lift
Test 3 (Social proof): +0.1-0.3% conversion lift
Test 4 (Form): +0.4-0.8% conversion lift
Combined (estimated): 1.8% -> 3.0-3.5%
Implementation Checklist
- Set up analytics events:
page_view,cta_click,form_start,signup_complete - Configure A/B tool (PostHog/LaunchDarkly) with feature flags
- Run tests sequentially (not simultaneously) to avoid interaction effects
- Wait for statistical significance before calling winner (p < 0.05)
- Monitor for novelty effect: check results stable after 2 full weeks
- Document winning variants and roll forward
Key Decisions
- Sequential testing: Avoids confounding variables from simultaneous changes
- Benefit-first headlines: Address pain point, not feature description
- Reduce friction first: Removing credit card requirement often gives largest lift
- Statistical rigor: Minimum sample sizes calculated upfront, no peeking
- PostHog over Google Optimize: Self-hosted, GDPR-friendly, integrates with product analytics