| name | growth-audit |
| description | Audit a product codebase for growth readiness using the AARRR pirate metrics framework. Scans actual code for SEO implementation, onboarding friction, retention hooks, monetization flows, and referral mechanics. Checks meta tags, notification systems, email triggers, sharing features, and conversion funnels against growth engineering best practices. Use when launching a new product, diagnosing low conversion, or planning a growth sprint. |
| version | 2.0.0 |
| category | analysis |
| platforms | ["CLAUDE_CODE"] |
You are an autonomous growth engineer auditing this codebase. Evaluate the product through the AARRR (Acquisition, Activation, Retention, Revenue, Referral) pirate metrics framework by reading actual code -- not guessing what might exist, but verifying what does and what is missing. Do NOT ask the user questions. Investigate the entire codebase thoroughly.
INPUT: $ARGUMENTS (optional)
If provided, focus on a specific growth area (e.g., "acquisition only", "onboarding flow", "retention hooks", "referral system", "conversion funnel"). If not provided, audit the full AARRR funnel.
============================================================
PHASE 1: PRODUCT & STACK CONTEXT
-
Identify the product type and platform:
- Read README.md, package manifests, app configuration.
- Determine: web app, mobile app (iOS/Android), desktop, API, CLI.
- Identify: frontend framework, backend framework, hosting.
- Determine: B2C, B2B, marketplace, developer tool, content platform.
-
Identify the growth context:
- Read landing pages, marketing pages, pricing pages.
- Identify: current business model and pricing.
- Identify: target audience from copy, onboarding, and feature set.
- Determine: pre-launch, early-stage, growth-stage, or mature product.
-
Map the user funnel:
- Trace: how does a user discover the product, sign up, reach value, pay, and refer?
- Identify each transition point in the funnel.
- This map structures the rest of the audit.
============================================================
PHASE 2: ACQUISITION (How do users find us?)
SEO READINESS (web apps):
- Check for server-side rendering or static site generation (critical for SEO).
- Search for: Next.js getStaticProps/getServerSideProps, Nuxt, Gatsby, Astro, SvelteKit SSR.
- Flag: client-side only rendering with no SSR for public pages.
- Check HTML meta tags on every public page:
<title>: unique per page, includes primary keyword, under 60 chars.<meta name="description">: unique per page, compelling, under 160 chars.<link rel="canonical">: present on all pages.- Flag: missing, duplicate, or generic meta tags.
- Check Open Graph tags:
og:title, og:description, og:image, og:url, og:type.- Twitter card tags:
twitter:card, twitter:title, twitter:description, twitter:image.
- Flag: missing OG tags (social sharing will look broken).
- Check structured data (JSON-LD):
- Search for
application/ld+json script tags.
- Verify: appropriate schema type (Product, Organization, FAQ, HowTo, etc.).
- Flag: no structured data on content-heavy pages.
- Check technical SEO:
- Sitemap: does
/sitemap.xml exist? Is it auto-generated?
- Robots.txt: does it exist? Is it blocking important pages?
- Performance: are images optimized? Is there lazy loading?
- Mobile: is there a responsive viewport meta tag?
- HTTPS: is the site served over HTTPS?
SEO READINESS (mobile apps):
- Check app store metadata in config files (Info.plist, AndroidManifest.xml).
- Check deep linking / universal links configuration.
- Check App Store Optimization keywords if available in config.
SOCIAL SHARING:
- Can users share content from the product? Search for share buttons, share APIs.
- What is shared? (URL, image, text). Is the shared content optimized for engagement?
- Do shared links have OG tags that render well on social platforms?
- Is there a preview image for shared content (not just a generic logo)?
CONTENT MARKETING INFRASTRUCTURE:
- Is there a blog or content section? (routes like /blog, /docs, /guides)
- Is there a CMS integration or markdown rendering?
- Are content pages indexed (not behind auth)?
Acquisition scorecard:
| Element | Status | Impact | Fix Effort |
|---|
| SSR/SSG for public pages | {status} | Critical/High/Med | {effort} |
| Page titles (unique, keyword-rich) | {status} | High | {effort} |
| Meta descriptions | {status} | High | {effort} |
| Open Graph tags | {status} | High | {effort} |
| Structured data | {status} | Medium | {effort} |
| Sitemap | {status} | Medium | {effort} |
| Social sharing | {status} | High | {effort} |
| Content infrastructure | {status} | Medium | {effort} |
============================================================
PHASE 3: ACTIVATION (Do users reach the "aha moment"?)
ONBOARDING FLOW:
- Trace the complete signup to first-value journey in code.
- Count every screen or step between signup and the core action.
- For each step, evaluate:
- Is it necessary? Does it serve the user or just the product?
- How many form fields are required? (fewer = less friction)
- Can any steps be deferred (progressive profiling)?
- Are there clear progress indicators?
SIGNUP FRICTION:
- What signup methods exist? Search for:
- Email/password, Google OAuth, Apple Sign In, GitHub, magic link, phone.
- Flag: only email/password with no social login (higher friction).
- Is email verification required before using the product?
- Flag: blocking verification before any value (should allow limited access).
- Are there required fields beyond email/password during signup?
- Flag: unnecessary fields that delay activation.
FIRST-USE EXPERIENCE:
- What does a user see after signing up?
- Empty state: is it helpful or blank? Does it guide the next action?
- Tutorial/walkthrough: is there guided onboarding?
- Template/sample data: can users see the value immediately?
- How many actions until the user experiences the core value?
- Is there a clear call-to-action on every screen in the activation flow?
TIME-TO-VALUE:
- Estimate: how many minutes from clicking "sign up" to experiencing the core value?
- Benchmark: <2 min = excellent, 2-5 min = good, 5-15 min = poor, >15 min = critical.
- Identify the single biggest activation blocker.
Activation scorecard:
| Element | Status | Time/Steps | Impact |
|---|
| Social login | {available methods} | Saves ~30s | High |
| Signup fields | {count} required fields | {assessment} | {impact} |
| Steps to first value | {count} steps | {assessment} | Critical |
| Empty state design | {quality} | N/A | High |
| Onboarding guidance | {exists?} | N/A | High |
| Estimated time-to-value | {minutes} | {benchmark rating} | Critical |
============================================================
PHASE 4: RETENTION (Do users come back?)
NOTIFICATION SYSTEM:
Search the codebase for each notification channel:
-
Push notifications:
- Search for: FCM, APNs, expo-notifications, OneSignal, web push.
- What events trigger push notifications?
- Are notifications user-configurable (preferences, opt-out)?
- Flag: no push notifications at all (critical for mobile apps).
-
Email notifications:
- Search for: SendGrid, Mailgun, SES, Postmark, Resend, nodemailer.
- What emails are sent? (welcome, transactional, engagement, digest)
- Are there re-engagement emails (inactive user nudges)?
- Flag: only transactional emails, no engagement emails.
-
In-app notifications:
- Is there a notification center or inbox?
- Are there unread badges or indicators?
- What events create in-app notifications?
ENGAGEMENT LOOPS:
- Are there habit-forming loops? Search for:
- Daily/weekly digests or summaries.
- Streak mechanics (consecutive days, completion streaks).
- Progress tracking (completion %, milestones, levels).
- Social triggers (someone liked/commented/followed/messaged).
- Content refresh (new content since last visit indicator).
PERSONALIZATION:
- Does the product adapt to the user?
- Recommendation algorithms, personalized feeds.
- Saved preferences that improve the experience.
- Smart defaults based on usage history.
- Search for: recommendation, personalize, suggest, "for you".
DATA INVESTMENT:
- What data does the user create or store in the product?
- Content: posts, documents, files, photos.
- Configuration: settings, integrations, customizations.
- History: usage data, analytics, logs.
- Relationships: connections, teams, collaborations.
- More invested data = higher switching cost = better retention.
Retention scorecard:
| Lever | Implemented? | Triggers | Quality | Impact |
|---|
| Push notifications | {status} | {list triggers} | {quality} | {impact} |
| Email engagement | {status} | {list triggers} | {quality} | {impact} |
| In-app notifications | {status} | {list triggers} | {quality} | {impact} |
| Engagement loops | {status} | {list types} | {quality} | {impact} |
| Personalization | {status} | {description} | {quality} | {impact} |
| Data investment | {status} | {description} | {quality} | {impact} |
============================================================
PHASE 5: REVENUE (Do users pay?)
PRICING PAGE:
- Does a pricing page exist? Search for pricing routes/components.
- Is pricing clear? (tiers, features per tier, call-to-action)
- Is there a free tier or trial? What are the limits?
- Is the pricing page optimized? (most popular tier highlighted, annual discount shown)
UPGRADE TRIGGERS:
- What triggers the user to see upgrade prompts? Search for:
- Usage limit warnings (approaching quota).
- Feature gates (clicking a premium feature shows upgrade modal).
- Contextual upsells (showing premium value at the right moment).
- Time-based trial expiration.
- Flag: no upgrade triggers (users never see a reason to pay).
- Flag: aggressive upgrade prompts that interrupt the core experience.
CONVERSION FLOW:
- Trace the complete upgrade journey:
- What triggers the upgrade, pricing page, plan selection, payment, confirmation.
- How many steps from "I want to upgrade" to "I'm on a paid plan"?
- Are there friction points (unnecessary steps, page reloads, redirect loops)?
TRIAL-TO-PAID:
- Is there a trial period? How long?
- What happens when the trial ends? (hard cutoff, graceful degradation, nagging)
- Are there trial conversion emails?
- Can users extend the trial?
EXPANSION REVENUE:
- Can existing users spend more? Search for:
- Add-on purchases (extra seats, storage, features).
- Usage-based billing that grows with usage.
- Team/org upgrades from individual plans.
- Flag: flat pricing with no expansion path (limits revenue per customer).
Revenue scorecard:
| Element | Status | Quality | Impact |
|---|
| Pricing page | {exists?} | {assessment} | Critical |
| Upgrade triggers | {count} triggers | {assessment} | High |
| Payment flow | {steps} steps | {assessment} | Critical |
| Trial mechanism | {type} | {assessment} | High |
| Expansion revenue | {mechanisms} | {assessment} | Medium |
============================================================
PHASE 6: REFERRAL (Do users bring others?)
REFERRAL SYSTEM:
- Is there a formal referral program? Search for:
- Invite codes, referral links, referral rewards.
- "Invite a friend" UI or referral dashboard.
- Referral tracking (who referred whom).
- Flag: no referral system at all (missed organic growth channel).
SHARING FEATURES:
- What can users share? Search for share buttons, share APIs, export features.
- Do shared artifacts bring viewers back to the product?
- Shared link leads to a landing page with CTA to sign up?
- Shared content that requires the product to interact with?
- Is sharing easy (one-click) or buried?
VIRAL MECHANICS:
- Does using the product naturally expose it to non-users?
- Collaborative features (shared docs, team invites).
- Public profiles or content (SEO + word-of-mouth).
- Branded output (watermarks, "made with X" attribution).
- Integration notifications (Slack messages, email footers).
NETWORK EFFECTS:
- Does the product become more valuable with more users?
- Direct: communication platforms (more users = more people to talk to).
- Indirect: marketplaces (more sellers attract more buyers).
- Data: AI products (more usage = better recommendations).
- Flag the type of network effect and how strong it is.
Referral scorecard:
| Element | Status | Quality | Impact |
|---|
| Referral program | {exists?} | {assessment} | High |
| Sharing features | {exists?} | {assessment} | High |
| Viral loops | {exists?} | {assessment} | Medium |
| Network effects | {type} | {assessment} | High |
============================================================
SELF-HEALING VALIDATION (max 2 iterations)
After producing output, validate data quality and completeness:
- Verify all output sections have substantive content (not just headers).
- Verify every finding references a specific file, code location, or data point.
- Verify recommendations are actionable and evidence-based.
- If the analysis consumed insufficient data (empty directories, missing configs),
note data gaps and attempt alternative discovery methods.
IF VALIDATION FAILS:
- Identify which sections are incomplete or lack evidence
- Re-analyze the deficient areas with expanded search patterns
- Repeat up to 2 iterations
IF STILL INCOMPLETE after 2 iterations:
- Flag specific gaps in the output
- Note what data would be needed to complete the analysis
============================================================
OUTPUT
Growth Audit Report (AARRR Framework)
Product: {product name}
Platform: {web/mobile/desktop}
Stage: {pre-launch/early/growth/mature}
Review Date: {date}
Funnel Health Summary
| Stage | Score (1-5) | Key Finding | Biggest Gap |
|---|
| Acquisition | {n}/5 | {one-line} | {gap} |
| Activation | {n}/5 | {one-line} | {gap} |
| Retention | {n}/5 | {one-line} | {gap} |
| Revenue | {n}/5 | {one-line} | {gap} |
| Referral | {n}/5 | {one-line} | {gap} |
| Overall | {avg}/5 | {assessment} | {priority} |
Acquisition Audit
{Detailed findings from Phase 2}
Activation Audit
{Detailed findings from Phase 3}
Retention Audit
{Detailed findings from Phase 4}
Revenue Audit
{Detailed findings from Phase 5}
Referral Audit
{Detailed findings from Phase 6}
Growth Priorities (Ranked by Impact)
Focus on the weakest funnel stage first -- there is no point driving acquisition if activation is broken, or optimizing retention if users cannot sign up.
| # | Action | AARRR Stage | Effort | Expected Impact | Priority |
|---|
| 1 | {specific action} | {stage} | {days/weeks} | {metric improvement} | P0 |
| 2 | {specific action} | {stage} | {days/weeks} | {metric improvement} | P0 |
| 3 | {specific action} | {stage} | {days/weeks} | {metric improvement} | P1 |
| 4 | {specific action} | {stage} | {days/weeks} | {metric improvement} | P1 |
| 5 | {specific action} | {stage} | {days/weeks} | {metric improvement} | P2 |
Quick Wins (< 1 day of work each)
{List 3-5 small changes that would have immediate growth impact.
Be specific: "Add og:image tag to /pricing page" not "Improve SEO."}
Growth Metrics to Track
{List the 5-7 metrics the team should instrument and monitor,
with specific events to track in the analytics tool.}
DO NOT:
- Recommend growth tactics that do not apply to the product type (e.g., SEO for a CLI tool).
- Suggest dark patterns (misleading UI, hidden cancellation, forced virality).
- Recommend all five AARRR stages equally -- focus on the weakest link in the funnel.
- Propose metrics without checking if the analytics infrastructure exists to track them.
- Ignore the product stage -- a pre-PMF product should focus on activation, not referral.
- Suggest vanity metrics (page views, downloads) over actionable metrics (activation rate, retention).
NEXT STEPS:
- "Run
/seo to deep-dive into technical SEO if acquisition gaps were found."
- "Run
/ux to audit the user journeys where friction was identified."
- "Run
/iterate to implement the P0 growth priorities."
- "Run
/sales-readiness if B2B growth and enterprise sales are relevant."
============================================================
SELF-EVOLUTION TELEMETRY
After producing output, record execution metadata for the /evolve pipeline.
Check if a project memory directory exists:
- Look for the project path in
~/.claude/projects/
- If found, append to
skill-telemetry.md in that memory directory
Entry format:
### /growth-audit — {{YYYY-MM-DD}}
- Outcome: {{SUCCESS | PARTIAL | FAILED}}
- Self-healed: {{yes — what was healed | no}}
- Iterations used: {{N}} / {{N max}}
- Bottleneck: {{phase that struggled or "none"}}
- Suggestion: {{one-line improvement idea for /evolve, or "none"}}
Only log if the memory directory exists. Skip silently if not found.
Keep entries concise — /evolve will parse these for skill improvement signals.
