afrexai-growth-engine（非洲增长引擎）

Growth Engineering Mastery

Complete growth system: experimentation engine, viral mechanics, channel playbooks, funnel optimization, retention loops, and scaling frameworks. From zero users to exponential growth.

1. Growth Audit — Where Are You Now?

Before experimenting, diagnose. Run this 8-dimension health check:

Growth Health Scorecard

Rate each 1-5, multiply by weight:

Scoring: 68-85 = Growth-ready. 50-67 = Fix foundations first. <50 = Stop growth spending, fix product.

PMF Validation Gate

Do NOT invest in growth until these pass:

pmf_gate:
  sean_ellis_test: "≥40% would be 'very disappointed' if product disappeared"
  retention_curve: "Flattens (does not trend to zero) by week 8"
  organic_growth: "≥10% of new users come from referral/word-of-mouth"
  nps: "≥30"
  qualitative: "Users describe product to friends without prompting"

If PMF gate fails: Stop. Go back to product. Growth without PMF = pouring water into a leaky bucket.

2. North Star Metric — Pick ONE Number

Selection Framework

Your North Star Metric (NSM) must pass all 4 tests:

1. Revenue proxy — More of this metric = more revenue (eventually)
2. User value — Captures the moment users get value
3. Measurable — Can track daily/weekly with existing tools
4. Influenceable — Team actions can move it within 2-4 weeks

NSM Examples by Business Type

Supporting Metrics Tree

North Star Metric
├── Input Metric 1: [driver you can directly influence]
├── Input Metric 2: [driver you can directly influence]
├── Input Metric 3: [driver you can directly influence]
└── Guard Metric: [thing that must NOT decrease]

Example (SaaS):

Weekly Active Teams (NSM)
├── New team activations/week (acquisition input)
├── Features used per team/week (engagement input)
├── Teams inviting 3+ members/week (virality input)
└── Guard: Churn rate must stay <3%/month

3. Experimentation Engine — The Core Growth Loop

ICE Scoring Framework

Every experiment gets scored before running:

ICE Score = (Impact + Confidence + Ease) / 3

Run experiments scoring ≥7 first. Kill anything below 5.

Experiment Log Template

experiment:
  id: "GRW-042"
  name: "Add social proof counter to pricing page"
  hypothesis: "Showing '2,847 teams trust us' increases plan selection by 15%"
  north_star_impact: "More paid conversions → more Weekly Active Teams"
  ice_score:
    impact: 7
    confidence: 6
    ease: 9
    total: 7.3
  type: "A/B test"
  audience: "All pricing page visitors"
  sample_size_needed: 2400  # for 95% confidence, 80% power
  duration: "7-14 days"
  primary_metric: "Pricing page → checkout conversion rate"
  secondary_metrics:
    - "Average plan tier selected"
    - "Time on pricing page"
  guard_metrics:
    - "Support tickets about pricing must not increase >10%"
  status: "running"  # proposed | running | won | lost | inconclusive
  result:
    lift: "+18.3%"
    confidence: "97.2%"
    decision: "Ship to 100%"
    learnings: "Social proof most effective on annual plans. Monthly plan conversion unchanged."
    next_experiment: "Test specific customer logos vs generic count"

Experiment Velocity Targets

Statistical Rigor Rules

• Minimum sample size: Calculate BEFORE launching (use: n = 16 × σ² / δ² or online calculator)
• Minimum runtime: 2 full business cycles (usually 2 weeks)
• No peeking: Don't stop tests early on positive results (peeking inflates false positives 3-5x)
• One change per test: Isolate variables. Multivariate only with massive traffic
• Document losses: Failed experiments are data. Log why the hypothesis was wrong

4. AARRR Funnel — Stage-by-Stage Playbooks

4.1 Acquisition — Getting Users In

Channel Evaluation Matrix

Score each channel before investing:

channel_evaluation:
  name: "[Channel]"
  scores:
    estimated_volume: 8      # 1-10: How many users can this deliver?
    targeting_precision: 7   # 1-10: Can we reach our ICP specifically?
    cost_per_acquisition: 6  # 1-10: How cheap? (10 = free/organic)
    time_to_results: 4       # 1-10: How fast? (10 = same day)
    scalability: 7           # 1-10: Can we 10x spend and 10x output?
    defensibility: 8         # 1-10: Hard for competitors to copy?
  total: 40  # out of 60
  verdict: "Test with $500 budget over 2 weeks"

Channel Playbooks (Top 12)

Organic Channels (low cost, slow build):

1. SEO/Content

- Target: Bottom-of-funnel keywords first (high intent, lower volume) - Playbook: 1 pillar page + 8-12 cluster articles per topic - Timeline: 3-6 months to meaningful traffic - Experiment: Test 3 content formats (how-to, comparison, listicle) — measure organic signups per article - Killer metric: Organic signups/article/month

1. Community/Forum Marketing

- Target: Where your ICP already hangs out (Reddit, HN, Discord servers, Slack groups) - Playbook: Provide genuine value for 30 days before any self-promotion. 20:1 value:ask ratio - Experiment: Track which communities drive highest-quality signups (activation rate, not just volume) - Warning: Getting banned kills the channel permanently. Authenticity is non-negotiable

1. Referral/Word-of-Mouth

- Target: Existing happy users - Playbook: See Section 5 (Viral Mechanics) below - Killer metric: K-factor (viral coefficient)

1. Social Media (Organic)

- Target: Platform where your ICP consumes content - Platform selection: LinkedIn (B2B), Twitter/X (tech/startup), TikTok (consumer/SMB), Instagram (visual/lifestyle) - Playbook: Post 5x/week, 80% value + 20% product. Reply to every comment for 90 days - Experiment: Test content types (text, carousel, video, thread) — measure profile visits → signups

1. Partnerships/Integrations

- Target: Products your users already use - Playbook: Build integration → get listed in partner's marketplace → co-market - Experiment: Partner A vs Partner B — which integration drives more activated users?

1. Product-Led SEO

- Target: Create public-facing pages that rank (templates, tools, directories) - Examples: Canva templates page, Zapier app directory, Ahrefs free tools - Experiment: Build 1 free tool targeting a high-volume keyword — measure signups from tool

Paid Channels (fast results, requires budget):

1. Search Ads (Google/Bing)

- Target: High-intent keywords (bottom of funnel) - Playbook: Start with exact match branded + competitor terms. Expand to problem-aware keywords - Budget rule: Don't spend >$50/day until CAC is profitable - Experiment: Ad copy A vs B, then landing page A vs B (sequential, not simultaneous)

1. Social Ads (Meta/LinkedIn/TikTok)

- Target: Lookalike audiences from best customers - Playbook: 3 creatives × 3 audiences × 3 copy variants. Kill losers at $50 spend, scale winners - LinkedIn: Only for B2B with ACV >$5K (expensive CPMs) - Experiment: Audience segmentation — which cohort has lowest CAC AND highest LTV?

1. Influencer/Creator

- Target: Micro-influencers (10K-100K followers) in your niche - Playbook: Product-for-post for micro. Paid for 50K+. Always track with UTM + unique codes - Experiment: 5 micro-influencers at $500 each. Compare CAC to paid ads

1. Cold Outreach (Email/LinkedIn)

- Target: Named accounts (ABM) - Playbook: 5-touch sequence over 14 days. Personalized first line. Clear CTA - Volume: 50-100/day per domain (warm up first). Separate domain from main - Experiment: Subject line tests (5 variants, 200 sends each)

Leverage Channels (unconventional):

1. PR/Media

- Target: Industry publications, podcasts, newsletters - Playbook: Newsjack trending topics. Offer original data/research. Be a source, not an ad - Experiment: 10 podcast appearances — measure signups per appearance

1. Platform Piggyback

- Target: Launch on Product Hunt, HN Show, AppSumo, marketplaces - Playbook: Coordinate launch day (Tuesday-Thursday). Mobilize existing users to upvote. Respond to every comment - Timeline: 1 day of effort, potentially thousands of signups - Experiment: Which platform delivers highest-LTV users?

Channel Prioritization Rule

The "Bull's Eye" Framework:

1. Brainstorm all 12+ channels
2. Rank by ICE score
3. Test top 3 with minimum viable spend ($500-1K each, 2 weeks)
4. Double down on the ONE winner
5. Don't diversify until that channel is saturated (CAC rising >30% month-over-month)

4.2 Activation — The "Aha Moment"

Define Your Aha Moment

aha_moment:
  description: "The specific action where users first experience core value"
  examples:
    slack: "Sent 2,000 team messages"
    dropbox: "Put 1 file in Dropbox folder"
    facebook: "Added 7 friends in 10 days"
    hubspot: "Imported contacts and sent first email"
  your_product:
    action: "[specific action]"
    threshold: "[quantity/frequency]"
    timeframe: "[within X days of signup]"
  validation: "Users who reach aha moment retain at 2x+ rate of those who don't"

Activation Funnel Map

Signup → [Step 1] → [Step 2] → ... → Aha Moment → Retained User
  |         |          |                  |
  v         v          v                  v
Drop-off  Drop-off  Drop-off          Success
 rate %    rate %    rate %             rate %

Map EVERY step. Measure EVERY drop-off. Fix the BIGGEST leak first.

Activation Tactics (by drop-off point)

Signup → First Session:

• Reduce signup friction (social login, no credit card, fewer fields)
• Welcome email within 5 minutes with ONE clear next step
• In-app checklist showing progress to aha moment
• Experiment: Remove 1 signup field → measure completion rate

First Session → Key Action:

• Interactive onboarding tour (max 4 steps)
• Pre-populate with sample data so product feels alive
• Contextual tooltips on first encounter (not all at once)
• Experiment: Guided tour vs self-serve vs video walkthrough

Key Action → Aha Moment:

• Trigger celebration/reward when they complete key action
• Show value immediately (dashboard, report, insight)
• Prompt sharing/inviting while enthusiasm is high
• Experiment: Time-to-value — can you deliver aha moment in <5 minutes?

Activation Scorecard

activation_metrics:
  signup_to_first_session: "Target: >80% within 24h"
  first_session_to_key_action: "Target: >60% within session 1"
  key_action_to_aha: "Target: >40% within 7 days"
  overall_activation_rate: "Target: >30% (signup → aha within 14 days)"
  benchmark_comparison: "[industry average is X%, we're at Y%]"

4.3 Retention — The Only Metric That Matters

Cohort Analysis Template

Track weekly cohorts (by signup week):

         Week 0  Week 1  Week 2  Week 3  Week 4  Week 8  Week 12
Cohort A  100%    45%     32%     28%     25%     22%     20%
Cohort B  100%    52%     38%     33%     30%     27%     25%
Cohort C  100%    48%     35%     30%     27%     24%     22%

What to look for:

• Does the curve flatten? (Good — you have a retention floor)
• Is each cohort better than the last? (Good — product is improving)
• Where's the biggest week-over-week drop? (Fix that transition)

Retention Curve Benchmarks

Retention Improvement Playbook

Week 1 drop-off (activation problem):

• Improve onboarding (see 4.2)
• Add "quick win" in first session
• Re-engagement email at 24h, 72h, 7 days

Week 2-4 drop-off (habit problem):

• Build triggers: notifications, emails, in-app prompts at optimal times
• Create recurring use case (weekly report, daily digest, scheduled task)
• Social hooks: team features, sharing, collaboration

Week 4+ decline (value problem):

• Feature depth: are power users hitting ceiling?
• New use cases: expand the "jobs to be done"
• Community: forums, events, user groups create switching cost

Engagement Loops

Design self-reinforcing loops:

User takes action → Gets value → Triggers notification/reminder → User returns → Takes deeper action

Types of engagement loops:

1. Content loop: User creates content → others consume → creator gets feedback → creates more
2. Social loop: User invites friend → friend joins → both get value → invite more
3. Data loop: User adds data → product gets smarter → better recommendations → user adds more
4. Habit loop: Trigger (email/notification) → Action (check dashboard) → Reward (insight) → Investment (customize)

4.4 Revenue — Monetization That Doesn't Kill Growth

Pricing-Growth Alignment

Revenue Experiments

• Pricing page layout: Test 2-tier vs 3-tier vs slider
• Anchor pricing: Test showing enterprise tier first vs starter first
• Trial length: 7-day vs 14-day vs 30-day (shorter often converts better)
• Feature gating: Which free feature, if paywalled, would drive most upgrades?
• Annual discount: Test 10%, 17%, 20%, 25% annual discount — optimize for LTV not just conversion

Unit Economics Health Check

unit_economics:
  cac: "$[X]"                    # Total sales+marketing / new customers
  ltv: "$[X]"                    # Average revenue × average lifetime
  ltv_cac_ratio: "[X]:1"        # Target: >3:1. Below 1 = losing money
  payback_months: "[X]"          # Target: <12 months (SaaS), <3 months (consumer)
  gross_margin: "[X]%"           # Target: >70% (SaaS), >40% (marketplace)
  expansion_revenue: "[X]%"      # % of revenue from existing customers expanding
  ndr: "[X]%"                    # Net Dollar Retention. Target: >100% (ideally >120%)

4.5 Referral — Turning Users Into a Growth Channel

See Section 5 (Viral Mechanics) for complete referral system design.

5. Viral Mechanics — Engineering Word-of-Mouth

Viral Coefficient (K-Factor)

K = invites_sent_per_user × conversion_rate_of_invites

K > 1 = exponential growth (every user brings >1 new user)
K = 0.5 = good amplifier (50% more users from virality)
K < 0.3 = not meaningfully viral

Viral Cycle Time

K-factor alone isn't enough. Speed matters:

Viral Cycle Time = time from user signup → their invite → invitee signup

Shorter cycle = faster growth (even with K < 1)

Goal: Reduce viral cycle time to <48 hours.

Types of Virality (Design for ALL of them)

1. Inherent Virality (product requires sharing)

• Example: Zoom (you invite people to join meetings), Figma (collaborate on designs)
• Design: Core use case involves other people
• Strongest form. Build this into the product if possible

2. Collaboration Virality (better with more people)

• Example: Slack (more teammates = more valuable), Notion (shared workspace)
• Design: Features that work better with team/network
• Trigger: Prompt team invites during high-value moments

3. Word-of-Mouth Virality (users talk about it)

• Example: ChatGPT (people share outputs), Canva (people share designs)
• Design: Create shareable outputs with subtle branding
• Trigger: Make outputs beautiful/impressive enough that users WANT to show them off

4. Incentivized Virality (rewards for sharing)

• Example: Dropbox (250MB per referral), Uber ($10 credit per referral)
• Design: Two-sided reward (referrer AND referee both get something)
• Warning: Attracts low-quality users if reward is too generous. Gate the reward behind activation

5. Artificial Scarcity/FOMO

• Example: Clubhouse (invite-only), Gmail (invite-only launch)
• Design: Limited access creates desire. Waitlists with position number
• Timing: Only effective at launch or for new features. Wears off fast

Referral Program Design Template

referral_program:
  name: "[Program name]"
  mechanics:
    referrer_reward: "[What they get]"
    referee_reward: "[What invitee gets]"
    reward_trigger: "Referee must [complete activation action] before rewards unlock"
    reward_type: "product_credit"  # cash | product_credit | feature_unlock | status
    cap: "10 referrals/month"      # Prevent gaming
  distribution:
    share_methods:
      - "Unique referral link (primary)"
      - "Email invite from product"
      - "Social share buttons (Twitter, LinkedIn)"
      - "QR code for in-person"
    placement:
      - "Post-aha-moment celebration screen"
      - "Settings/account page"
      - "Monthly usage summary email"
      - "In-app prompt after positive action (e.g., saved money, closed deal)"
  tracking:
    metrics:
      - "Share rate: % of users who share referral link"
      - "Click-through rate: % of link viewers who click"
      - "Conversion rate: % of clickers who sign up"
      - "Activation rate: % of referred signups who activate"
      - "K-factor: shares × CTR × signup × activation"
    cohort_quality: "Compare referred users vs non-referred on Day 30 retention + LTV"
  optimization_experiments:
    - "Test reward amount ($5 vs $10 vs $20)"
    - "Test reward type (credit vs cash vs feature)"
    - "Test referral prompt timing (post-signup vs post-aha vs post-payment)"
    - "Test share copy (3 variants)"

Viral Content Strategies

For products where output sharing drives growth:

1. Branded outputs: Add subtle watermark/badge ("Made with [Product]") to exports, reports, shares
2. Public profiles/pages: User-created content that's publicly accessible (SEO + social sharing)
3. Embed widgets: Let users embed product functionality on their sites
4. Template marketplace: User-created templates others can discover and use
5. Leaderboards/badges: Shareable achievements that demonstrate status

6. Growth Loops — Self-Reinforcing Systems

Why Loops > Funnels

Funnels are linear (top → bottom, then done). Loops are circular — output becomes input.

Loop Architecture

[New User] → [Takes Action] → [Creates Value] → [Attracts New User] → repeat

6 Growth Loop Templates

1. User-Generated Content Loop

User creates content → Content gets indexed/shared → New user discovers content → Signs up to create own → Creates content

• Examples: Medium, GitHub, Canva templates
• Key metric: Content pieces created/week
• Leverage point: Make content creation effortless + discoverable

2. Paid Marketing Loop

Revenue → Reinvest in ads → Acquire users → Users generate revenue → Reinvest more

• Key metric: LTV:CAC ratio (must be >3:1)
• Leverage point: Increase LTV (expansion revenue, retention) → can afford higher CAC

3. Sales Loop

Close deal → Case study/testimonial → Use in sales materials → Close next deal faster

• Key metric: Win rate improvement per quarter
• Leverage point: Systematize case study collection (ask at Month 3 of every account)

4. Data Network Effect Loop

Users use product → Product collects data → Product improves (AI/ML/recommendations) → More valuable for all users → More users join

• Examples: Waze, Netflix recommendations, Google Search
• Key metric: Improvement in core metric per doubling of data
• Leverage point: Show users how product gets better with more usage

5. Marketplace/Platform Loop

Supply joins → Attracts demand → Demand attracts more supply → More selection attracts more demand

• Key metric: Liquidity (% of listings that transact)
• Leverage point: Solve chicken-and-egg: seed supply first, constrain geography to build density

6. Community Loop

Expert users help newbies → Newbies become power users → Power users help next wave → Community grows

• Examples: Stack Overflow, Reddit, Discord servers
• Key metric: Weekly active contributors
• Leverage point: Gamification (reputation, badges, privileges for top contributors)

7. Funnel Optimization — CRO Playbook

Conversion Rate Benchmarks

Landing Page Optimization Checklist

• [ ] Hero headline matches ad/source copy (message match)
• [ ] Clear value proposition in ≤10 words
• [ ] Social proof above the fold (logos, numbers, testimonials)
• [ ] ONE primary CTA (not 3 competing buttons)
• [ ] CTA button text is action-specific ("Start free trial" not "Submit")
• [ ] Mobile-first design (60%+ of traffic is mobile)
• [ ] Page loads in <3 seconds (every second = 7% conversion drop)
• [ ] Remove navigation (landing page ≠ homepage)
• [ ] Include objection handling (FAQ, guarantee, security badges)
• [ ] Exit-intent popup with alternate offer

High-Impact CRO Experiments (ordered by typical lift)

1. Headline copy (10-30% lift potential) — Test problem-focused vs benefit-focused vs social-proof
2. CTA button (5-20% lift) — Test color, copy, size, position
3. Social proof type (5-15% lift) — Test logos vs testimonials vs numbers vs case studies
4. Form length (10-25% lift) — Test fewer fields, progressive profiling
5. Page layout (5-15% lift) — Test long-form vs short-form, video vs text
6. Pricing display (10-30% lift) — Test anchoring, default selection, feature comparison
7. Trust signals (3-10% lift) — Test guarantees, security badges, review scores

8. Retention & Re-engagement — Keeping Users

Lifecycle Email Sequences

Welcome Sequence (Days 0-14)

welcome_sequence:
  - day: 0
    trigger: "Signup"
    subject: "Welcome — here's your quick win"
    content: "One specific action to get value in <5 minutes"
    cta: "Do [aha action] now"
  - day: 1
    trigger: "Has NOT completed aha action"
    subject: "[First name], you're 1 step away"
    content: "Show what they'll get once they complete the action"
    cta: "Complete setup"
  - day: 3
    trigger: "Still not activated"
    subject: "How [similar company] uses [Product]"
    content: "Case study / use case matching their profile"
    cta: "Try this approach"
  - day: 7
    trigger: "Not activated"
    subject: "Need help? Reply to this email"
    content: "Personal note from founder. Offer 1:1 call"
    cta: "Reply or book call"
  - day: 14
    trigger: "Still not activated"
    subject: "Last chance: your [Product] account"
    content: "We'll archive your account in 7 days. Here's what you're missing"
    cta: "Reactivate"

Re-engagement Sequence (for churned/dormant users)

reengagement:
  - trigger: "14 days inactive"
    subject: "We miss you — here's what's new"
    content: "Top 3 new features/improvements since they left"
  - trigger: "30 days inactive"
    subject: "[First name], [specific value they got] is waiting"
    content: "Reference their actual usage data. Show what they've built"
  - trigger: "60 days inactive"
    subject: "Should we close your account?"
    content: "FOMO trigger. Offer win-back discount (20-30% off)"
  - trigger: "90 days inactive"
    subject: "Feedback request (we'll shut up after this)"
    content: "Why did you leave? 3-question survey. Offer incentive"

Push Notification Strategy

Rules:

• Max 3-5/week (more = uninstall)
• Only send when you can show value (not "We miss you!")
• Personalize: "Your report is ready" > "Check out new features"
• A/B test timing: morning vs evening, weekday vs weekend
• Let users choose notification categories

Churn Prediction Signals

Build an early warning system. Track these leading indicators:

Response playbook: Trigger automated outreach at 🟡, human outreach at 🔴.

9. Scaling — From Working to 10x

When to Scale a Channel

scale_criteria:
  channel: "[name]"
  ready_when:
    - "CAC is <1/3 of LTV"
    - "Conversion rates are stable for 4+ weeks"
    - "Process is documented and repeatable"
    - "Can increase spend 50% without CAC rising >20%"
  warning_signs:
    - "CAC rising >20% month-over-month"
    - "Conversion rates declining"
    - "Quality of leads/users dropping (lower activation rate)"
    - "Creative fatigue (CTR declining)"

Scaling Playbook

1. Automate first — Before hiring, automate everything possible (email sequences, ad management, content scheduling)
2. Document SOPs — Every process needs a playbook before delegation
3. Hire specialists, not generalists — At scale, you need a paid ads person, not a "growth person"
4. Build dashboards before scaling — If you can't measure it in real-time, you can't scale it safely
5. 10% rule — Increase budget/volume by max 10-20%/week. Sudden jumps break things

International Expansion Checklist

• [ ] Localize landing pages (not just translate — adapt)
• [ ] Research local competitors and positioning
• [ ] Adjust pricing for purchasing power (PPP)
• [ ] Local payment methods (not just Stripe)
• [ ] Support in local timezone and language
• [ ] Comply with local regulations (GDPR, data residency)
• [ ] Test demand before committing (run ads in target language first)

10. Growth Team Structure

Solo/Small Team (1-3 people)

Growth Lead (you)
├── Runs experiments (2-3/week)
├── Manages 1-2 channels
├── Analyzes data weekly
└── Writes copy/creates content

Focus: Find ONE channel that works. Don't spread thin.

Growth Team (4-10 people)

Head of Growth
├── Acquisition Lead → paid, SEO, partnerships
├── Product/Growth Engineer → experiments, features, A/B tests
├── Lifecycle/CRM → emails, notifications, retention
└── Data Analyst → metrics, cohorts, experiment analysis

Growth Meeting Cadence

11. Growth Toolkit — Technical Setup

Analytics Stack (Minimum Viable)

analytics_stack:
  product_analytics: "Mixpanel or Amplitude or PostHog (free tier)"
  web_analytics: "Google Analytics 4 + Google Tag Manager"
  attribution: "UTM parameters (mandatory on ALL links)"
  ab_testing: "PostHog or GrowthBook (free) or Optimizely (paid)"
  email: "Customer.io or Resend or SendGrid"
  crm: "HubSpot (free) or Pipedrive"
  session_recording: "Hotjar or FullStory (free tier)"
  surveys: "Typeform or native in-app"

UTM Convention

utm_source: [platform] — google, linkedin, twitter, email, partner-name
utm_medium: [type] — cpc, social, email, referral, organic
utm_campaign: [campaign-name] — q1-launch, black-friday, webinar-series
utm_content: [variant] — hero-cta, sidebar-banner, email-v2
utm_term: [keyword] — only for paid search

Rule: Every external link gets UTMs. No exceptions. Untracked traffic = wasted budget.

Event Tracking Plan

Track these events minimum:

required_events:
  acquisition:
    - "page_view (with UTM params)"
    - "signup_started"
    - "signup_completed"
  activation:
    - "onboarding_step_completed (step_number)"
    - "first_key_action"
    - "aha_moment_reached"
  engagement:
    - "feature_used (feature_name)"
    - "session_started"
    - "session_duration"
  revenue:
    - "plan_selected (plan_name, price)"
    - "payment_completed (amount, plan)"
    - "upgrade (from_plan, to_plan)"
    - "churn (reason)"
  referral:
    - "referral_link_shared (method)"
    - "referral_link_clicked"
    - "referred_signup"
    - "referred_activated"

12. Anti-Patterns & Common Mistakes

The 10 Growth Killers

1. Scaling before PMF — Spending on acquisition when retention is broken = burning money
2. Vanity metrics addiction — Signups, downloads, pageviews mean nothing without activation + retention
3. Copying without context — "Dropbox did referrals" doesn't mean you should. Understand WHY it worked for THEM
4. Too many channels too soon — Master ONE before adding another. Spread thin = learn nothing
5. Peeking at A/B tests — Stopping tests early inflates false positives 3-5x. Run to completion
6. Optimizing pennies — CRO on a page getting 100 visits/month is pointless. Get traffic first
7. Ignoring retention — Acquiring users you can't keep is literally the most expensive thing you can do
8. Over-automating before understanding — Automate processes you've done manually 50+ times. Not before
9. Growth hacks without strategy — One-off tactics without a system = random acts of marketing
10. Not documenting experiments — If you don't log it, you'll repeat failures and forget successes

When Growth Stalls

Diagnostic checklist:

• [ ] Has the channel saturated? (CAC up >30% in 3 months)
• [ ] Has the product changed? (New features breaking existing flows)
• [ ] Has the market shifted? (New competitor, regulation, trend change)
• [ ] Has the team burned out? (Experiment velocity dropped)
• [ ] Is it seasonal? (Compare to same period last year)
• [ ] Are you measuring the right thing? (NSM still reflects actual value?)

13. Edge Cases & Special Situations

B2B vs B2C Growth Differences

Two-Sided Marketplace Growth

Chicken-and-egg solution order:

1. Seed supply manually (scrape, import, do it yourself)
2. Constrain geography (one city/niche first)
3. Offer supply-side tools for free (even without demand)
4. Build just enough demand to show supply it works
5. Let organic flywheel take over before expanding geography

PLG (Product-Led Growth) Specifics

plg_metrics:
  free_to_paid: "Target: 3-5% (freemium) or 15-25% (free trial)"
  time_to_value: "Target: <5 minutes"
  expansion_rate: "Target: >120% NDR"
  self_serve_ratio: "Target: >80% of revenue from self-serve"
  pql_rate: "Target: 20-40% of active free users qualify"

Product Qualified Lead (PQL) definition: User who has reached activation AND shows buying signals (hits usage limit, views pricing page, invites team members).

Growth with Zero Budget

1. Build in public (Twitter/LinkedIn) — share metrics, learnings, behind-the-scenes
2. Launch on 5 platforms: Product Hunt, HN, Reddit, Indie Hackers, relevant Discords
3. Write 1 SEO article/week targeting long-tail keywords
4. Offer free tool that solves a related problem → funnel to main product
5. Cold DM 10 potential users/day — ask for feedback, not sales
6. Partner with complementary products for cross-promotion
7. Answer questions on Quora/Reddit/forums where your ICP hangs out

14. Weekly Growth Review Template

weekly_review:
  period: "Week of [DATE]"
  north_star_metric:
    current: "[X]"
    target: "[X]"
    trend: "up|down|flat"
    wow_change: "+X%"
  funnel_metrics:
    acquisition: "[visitors/signups]"
    activation: "[activated/total signups] = X%"
    retention: "[week 1 retention] = X%"
    revenue: "[$MRR] | [new paying] | [churned]"
    referral: "[K-factor] | [referral signups]"
  experiments:
    completed:
      - name: "[experiment]"
        result: "won|lost|inconclusive"
        impact: "[metric change]"
        next_step: "[ship|iterate|kill]"
    running:
      - name: "[experiment]"
        progress: "[X/Y days complete]"
        early_signal: "[trending positive|neutral|negative]"
    launching_next_week:
      - name: "[experiment]"
        ice_score: "[X]"
        hypothesis: "[statement]"
  channels:
    - name: "[channel]"
      spend: "$[X]"
      cac: "$[X]"
      volume: "[X] new users"
      quality: "[activation rate of users from this channel]"
  top_learning: "[Single most important thing learned this week]"
  biggest_risk: "[What could derail growth next month?]"
  focus_next_week: "[1-2 priorities]"

15. Natural Language Commands

Use these to activate specific workflows: