sinapse-ai 9.3.0 → 9.4.0

package/squads/squad-growth/knowledge-base/growth-master-reference.md

@@ -0,0 +1,1168 @@
+# Growth Master Reference
+
+> **Source:** MS-004 Growth Master System Research (42+ sources, verified 2026-04-07)
+> **Squad:** squad-growth | **Orchestrator:** Catalyst
+> **Scope:** Frameworks, SEO, analytics, experimentation, retention, viral, CRO, Brazilian context
+
+---
+
+## Table of Contents
+
+1. [Growth Frameworks](#1-growth-frameworks)
+2. [North Star Metric](#2-north-star-metric)
+3. [Growth Loops](#3-growth-loops)
+4. [Product-Led Growth (PLG)](#4-product-led-growth-plg)
+5. [SEO & Organic Acquisition](#5-seo--organic-acquisition)
+6. [Content Marketing Flywheel](#6-content-marketing-flywheel)
+7. [Analytics Stack](#7-analytics-stack)
+8. [Experimentation & A/B Testing](#8-experimentation--ab-testing)
+9. [Retention & Hook Model](#9-retention--hook-model)
+10. [Viral & Referral Mechanics](#10-viral--referral-mechanics)
+11. [CRO — Conversion Rate Optimization](#11-cro--conversion-rate-optimization)
+12. [Brazilian Growth Context](#12-brazilian-growth-context)
+13. [Prioritization Frameworks](#13-prioritization-frameworks)
+14. [Data-Driven Decision Making](#14-data-driven-decision-making)
+15. [Email & Lifecycle Marketing](#15-email--lifecycle-marketing)
+16. [AI & Growth](#16-ai--growth)
+17. [Key People & Bibliography](#17-key-people--bibliography)
+18. [Actionable Checklists](#18-actionable-checklists)
+
+---
+
+## 1. Growth Frameworks
+
+### 1.1 AARRR — Pirate Metrics (Dave McClure, 2007)
+
+The five stages of the user lifecycle:
+
+| Stage | Metric | Key Question | Examples |
+|-------|--------|--------------|----------|
+| **Acquisition** | New users/visitors | How do users find you? | SEO, ads, referral, PR |
+| **Activation** | % reaching "aha moment" | Do users have a positive first experience? | Complete onboarding, first value |
+| **Retention** | % returning | Do users come back? | DAU/MAU, recurring login |
+| **Revenue** | Revenue per user | Do users pay? | Trial-to-paid, ARPU |
+| **Referral** | % who refer | Do users bring others? | K-factor, NPS, invites |
+
+### 1.2 RARRA — Retention-First Reordering (Brian Balfour / Reforge)
+
+Prioritization order should be **RARRA**, not AARRR:
+
+```
+Retention → Activation → Revenue → Referral → Acquisition
+```
+
+**Rationale:** Acquiring users who don't stay is a leaky bucket. Fix retention first, then optimize upstream.
+
+### 1.3 Reforge Growth Model — Four Loops
+
+1. **Acquisition Loop** — How new users arrive (paid, viral, content, sales)
+2. **Engagement Loop** — How users keep using (Trigger → Action → Variable Reward → Investment)
+3. **Monetization Loop** — How value converts to revenue (freemium → upgrade → expansion)
+4. **Defensibility** — How advantages accumulate (network effects, data moats, brand)
+
+### 1.4 Growth Eras
+
+| Era | Period | Focus |
+|-----|--------|-------|
+| Growth Hacking | 2010-2015 | Tactical hacks, viral tricks, underused channels |
+| Growth as Discipline | 2015-2020 | Reforge, systematic frameworks, dedicated teams |
+| Growth Engineering | 2020-present | Product-integrated, retention-first, PLG, AI automation |
+
+### 1.5 Sustainable Growth Requires
+
+1. **Strong retention** — Product solves a real problem and creates habit
+2. **Compound loops** — Existing user activity generates new users
+3. **Channel diversification** — Portfolio of channels at different maturities
+4. **Data moats** — Advantages that accumulate with scale (network effects, proprietary data)
+
+> **Law of Shitty Clickthroughs** (Andrew Chen): Every acquisition channel degrades over time. Banner CTR was 78% in 1994, now below 0.1%.
+
+---
+
+## 2. North Star Metric
+
+### Definition
+
+The single metric that best captures the core value a product delivers to users. Serves as organizational alignment — all teams optimize toward the same guiding star.
+
+### Criteria for a Good NSM
+
+1. Reflects value delivery to the user (not just revenue)
+2. Is a leading indicator of future revenue
+3. Is measurable and actionable
+4. Can be influenced by multiple teams
+
+### Classic Examples
+
+| Company | NSM | Logic |
+|---------|-----|-------|
+| Facebook | DAU | Daily engagement = value for users and advertisers |
+| Airbnb | Nights booked | Core value: completed stays |
+| Spotify | Listening time | More listening = more value = more retention |
+| Slack | Messages sent per team | Adoption within the organization |
+| HubSpot | Weekly Active Teams | Recurring team engagement |
+| Dropbox | Files synced | Core value: active storage |
+
+### Guardrails
+
+NSM does NOT replace guardrail metrics. If Spotify optimized only for listening time, it could degrade quality (infinite autoplay). Guardrails like NPS, churn, and revenue prevent perverse optimization.
+
+---
+
+## 3. Growth Loops
+
+### Why Loops > Funnels
+
+| Aspect | Funnel (Linear) | Loop (Compound) |
+|--------|-----------------|-----------------|
+| Source | Constant external input (budget) | Output becomes input |
+| Marginal cost | Constant or increasing | Decreasing |
+| Curve | Flat (or decelerating) | Exponential (J-curve) |
+| Sustainability | Depends on budget | Self-sustaining |
+| Examples | Ads, outbound sales | Viral, UGC, network effects |
+
+### Acquisition Loops
+
+**1. Viral Loop:**
+```
+User → Uses product → Invites/shares → New user → Uses product → ...
+```
+Examples: Slack, Zoom, Calendly
+
+**2. UGC Loop:**
+```
+User → Creates content → Content indexed by Google → New user finds it → Creates content → ...
+```
+Examples: Pinterest, Reddit, Quora, Stack Overflow
+
+**3. Paid Loop:**
+```
+Revenue → Invest in ads → New users → Revenue → Invest more → ...
+```
+Works if LTV > CAC with margin to reinvest.
+
+**4. Sales Loop:**
+```
+Revenue → Hire salespeople → New clients → Revenue → Hire more → ...
+```
+
+### Engagement Loops
+
+| Loop Type | Mechanism | Examples |
+|-----------|-----------|----------|
+| Personal Utility | User stores data → data becomes valuable → returns | Notion, Google Drive |
+| Social | User posts → receives feedback → motivated → posts more | Instagram, LinkedIn |
+| Notification | Event → push/email → user returns → generates activity | WhatsApp, Slack |
+
+### Monetization Loops
+
+**Expansion Revenue:**
+```
+User → Uses more → Hits limits → Upgrades plan → Uses more → ...
+```
+
+**Marketplace Liquidity:**
+```
+More sellers → More options → More buyers → More demand → More sellers → ...
+```
+
+> **Goal:** Every company should identify and optimize at least one compound loop. Linear loops (paid, outbound) bootstrap and complement but shouldn't be the only growth source.
+
+---
+
+## 4. Product-Led Growth (PLG)
+
+### Core Principles
+
+1. **Product is the channel** — User experiences value before talking to sales
+2. **Self-serve first** — Frictionless onboarding, no mandatory demo
+3. **Minimum time-to-value** — User hits "aha moment" ASAP
+4. **Bottom-up adoption** — Individual users adopt, then organization buys
+5. **Data-driven expansion** — Upsell based on behavior, not sales pitch
+
+### PLG Examples
+
+| Company | PLG Model | Mechanism |
+|---------|-----------|-----------|
+| Slack | Freemium + viral | Teams adopt, then buy |
+| Zoom | Freemium + viral | Host uses, guests experience |
+| Figma | Freemium + collaboration | Designers share, teams adopt |
+| Notion | Freemium + templates | Templates attract, workflows retain |
+| Calendly | Freemium + viral | Every scheduling link = marketing |
+| Loom | Freemium + viral | Every video sent = marketing |
+
+### Pricing Models
+
+| Model | Free Duration | Typical Conversion | Best For |
+|-------|---------------|-------------------|----------|
+| **Freemium** | Permanent (limited) | 2-5% | High volume, low touch |
+| **Free Trial** | 7-30 days (full access) | 15-25% (great); 8-12% (good) | Complex products needing time |
+| **Reverse Trial** | Starts premium, downgrades | 10-15% | Products where premium value isn't immediate |
+| **Hybrid (PLG + Sales)** | Self-serve + sales-assisted | Varies | SMB self-serve, Enterprise sales |
+
+### Product Qualified Leads (PQLs)
+
+Users who demonstrate purchase intent through product behavior:
+
+| Signal | Weight | Threshold |
+|--------|--------|-----------|
+| Users in workspace | 5 | >= 5 |
+| Premium features used | 4 | >= 3 |
+| Pricing page visits | 3 | >= 2 in 7 days |
+| Weekly active time | 3 | >= 3 hours |
+| Active integrations | 2 | >= 2 |
+
+### Onboarding Framework
+
+1. **Sign-up** — Minimum fields, social login, no credit card (freemium)
+2. **Welcome Survey** — 2-3 questions to personalize (JTBD, role, goal)
+3. **Setup Checklist** — Clear steps with visual progress
+4. **Quick Win** — First value in <5 minutes
+5. **Celebrate** — Positive reinforcement at milestones
+6. **Ongoing Education** — Contextual tooltips, educational emails
+
+**Onboarding Metrics:**
+- Time-to-First-Value (TTFV)
+- Activation Rate (% completing critical setup)
+- D1 / D7 / D30 Retention by onboarding cohort
+- Setup Completion Rate
+
+### Network Effects
+
+| Type | Description | Example |
+|------|-------------|---------|
+| **Direct** | More users = more value per user | WhatsApp, phone |
+| **Indirect (cross-side)** | More users on one side = more value for the other | Uber (riders/drivers) |
+| **Data network** | More usage = better product (via data) | Waze, Google Search |
+| **Marketplace** | Liquidity attracts both sides | Airbnb, Amazon Marketplace |
+
+---
+
+## 5. SEO & Organic Acquisition
+
+### Three Pillars
+
+1. **Technical SEO** — Infrastructure enabling crawling and indexation
+2. **Content SEO** — Relevant content matching user search intent
+3. **Off-page SEO** — Domain authority via backlinks and mentions
+
+> Organic Google traffic = **53.3%** of all web traffic (BrightEdge 2025). AI search is growing but still <1% of referral traffic.
+
+### 5.1 Technical SEO
+
+| Element | Description | Impact |
+|---------|-------------|--------|
+| **Core Web Vitals** | LCP, INP, CLS | Ranking factor since 2021 |
+| **Mobile-first indexing** | Google indexes mobile version first | Ranking factor |
+| **Site speed** | Load time | Directly correlated with bounce rate |
+| **Structured data** | Schema.org markup | Rich snippets, knowledge panels |
+| **XML Sitemap** | URL map for crawlers | Crawling efficiency |
+| **Robots.txt** | Crawler directives | Crawl budget control |
+| **Canonical tags** | Preferred URL indicators | Prevent duplicate content |
+| **Hreflang** | Language/region version | International SEO |
+| **HTTPS** | SSL certificate | Ranking factor |
+| **Internal linking** | Links between site pages | Distributes PageRank, helps crawling |
+
+#### Core Web Vitals (Updated 2025-2026)
+
+INP replaced FID as Core Web Vital in March 2024. INP measures end-to-end responsiveness.
+
+| Metric | Measures | Good | Needs Improvement | Poor |
+|--------|----------|------|-------------------|------|
+| **LCP** | Loading speed | <= 2.5s (gold: <2.0s) | <= 4.0s | > 4.0s |
+| **INP** | Responsiveness | <= 200ms | <= 500ms | > 500ms |
+| **CLS** | Visual stability | <= 0.1 | <= 0.25 | > 0.25 |
+
+**Key data:** Only ~47% of sites meet CWV thresholds in 2026. INP is the most failed metric (43% fail). Sites with INP >300ms reported ranking drops up to 31% (especially mobile). Google Dec 2025 Core Update (volatility 8.7/10) affected 40-60% of sites globally.
+
+### 5.2 Content SEO
+
+#### Search Intent Types
+
+| Intent | Example | Ideal Format |
+|--------|---------|--------------|
+| **Informational** | "how to do SEO" | Blog post, guide, tutorial |
+| **Navigational** | "Ahrefs login" | Product/login page |
+| **Commercial** | "best SEO tool" | Comparison, review |
+| **Transactional** | "subscribe Ahrefs" | Landing page, pricing |
+
+#### E-E-A-T
+
+- **Experience** — Does the author have direct experience?
+- **Expertise** — Does the author have demonstrable expertise?
+- **Authoritativeness** — Is the site/author a recognized authority?
+- **Trustworthiness** — Is the content reliable? Is the site secure?
+
+Critical for YMYL (Your Money or Your Life) content but impacts all niches.
+
+#### Topical Authority — Pillar-Cluster Model
+
+```
+Pillar Page: "Complete Guide to SEO" (3000+ words)
+  +-- Cluster: "Technical SEO" (1500 words)
+  +-- Cluster: "Keyword Research" (1500 words)
+  +-- Cluster: "Link Building" (1500 words)
+  +-- Cluster: "Content SEO" (1500 words)
+  +-- Cluster: "Local SEO" (1500 words)
+```
+
+Each cluster links to the pillar and vice-versa, creating a thematic silo signaling authority.
+
+#### Programmatic SEO
+
+Create thousands of optimized pages using templates + structured data.
+
+**Components:** Head term + Modifier + Template + Data + Unique Value
+
+**Examples:** Zapier (25K+ integration pages), TripAdvisor (destination pages), Wise (currency conversion pages)
+
+**Risks:** Thin/duplicate content penalties, pages without unique value won't rank, wasted crawl budget.
+
+### 5.3 Link Building Strategies
+
+| Strategy | Description | Scalability |
+|----------|-------------|-------------|
+| Guest posting | Write for other sites | Medium |
+| Digital PR | Create stories press covers | High |
+| Broken link building | Find broken links, offer alternative | Medium |
+| Skyscraper technique | Create better content, request substitution | Low |
+| Data/research | Publish original citable data | High |
+| Tools/calculators | Create useful tools that attract links | High |
+| HARO/Connectively | Respond to journalists seeking sources | Medium |
+
+### 5.4 AI Search Impact (SGE / AI Overviews)
+
+| Data Point | Value | Source |
+|------------|-------|--------|
+| AI Overviews appearance | ~25.8% of US searches (up to 48-60% by some methodologies) | Semrush, BrightEdge, Ahrefs (Jan 2026) |
+| Organic CTR drop with AI Overview | -61% (from 1.76% to 0.61%) | Seer Interactive 2025 |
+| Zero-click searches | ~58.5-60% of all Google searches | SparkToro/Datos 2024-2025 |
+| Zero-click with AI Overviews active | ~83% | SparkToro |
+| Mobile zero-click | ~77% vs ~47% desktop | SparkToro |
+| Sites cited IN AI Overviews | CTR can increase up to 35% | Multiple sources |
+
+**Adaptation Strategies:**
+1. Optimize for AI citation — well-structured content, factual data, authority
+2. Focus on transactional searches — less impacted by AI Overviews
+3. Create content AI can't replicate — original experience, proprietary data, expert opinion
+4. Diversify beyond Google — YouTube, TikTok, Reddit, Perplexity
+5. Brand building — branded searches are not impacted
+
+---
+
+## 6. Content Marketing Flywheel
+
+### Flywheel vs. Calendar
+
+The traditional content calendar is linear: plan, produce, publish, promote. The **Content Flywheel** is compound:
+
+```
+Research → Long-form content → Atomize into smaller pieces → Distribute across channels
+   ^                                                                    |
+   +-------- Performance data feeds next research cycle ----------------+
+```
+
+### Pillar-Cluster Architecture
+
+1. **Pillar Page** — Comprehensive guide (2000-5000 words)
+2. **Cluster Pages** — Focused subtopic articles (1000-2000 words)
+3. **Internal links** — Each cluster links to pillar and vice-versa
+4. **Semantic coverage** — Together, pages cover the topic exhaustively
+
+### Content Scoring Framework
+
+| Dimension | Weight | Metrics |
+|-----------|--------|---------|
+| Organic traffic | 30% | Sessions, impressions, CTR |
+| Engagement | 25% | Time on page, scroll depth, bounce rate |
+| Conversion | 25% | Leads generated, trial signups, PQLs |
+| Quality | 10% | E-E-A-T score, backlinks earned |
+| Freshness | 10% | Last update, temporal relevance |
+
+### Distribution Channels
+
+| Channel | Type | When to Use |
+|---------|------|-------------|
+| SEO/Google | Owned | Evergreen content, clear search intent |
+| Email newsletter | Owned | Existing audience, nurturing |
+| Social media (organic) | Borrowed | Brand awareness, engagement |
+| YouTube | Owned/Borrowed | Tutorials, thought leadership |
+| Podcast | Owned | Thought leadership, audience building |
+| Communities (Reddit, Discord) | Borrowed | Specific niches, feedback |
+| LinkedIn (personal) | Borrowed | B2B, employer branding |
+| Partnerships/Guest | Earned | Reach new audiences |
+
+### Content Repurposing Model
+
+```
+Webinar (60min)
+  +-- Blog post (3000 words)
+  +-- YouTube video (edited, 15min)
+  +-- 10 short clips (TikTok, Reels, Shorts)
+  +-- Podcast episode (audio)
+  +-- Infographic
+  +-- Twitter/X thread
+  +-- LinkedIn carousel
+  +-- Email series (3 emails)
+  +-- Newsletter digest
+```
+
+**Rule of thumb:** 80% distribution, 20% creation.
+
+---
+
+## 7. Analytics Stack
+
+### Modern Analytics Ecosystem
+
+| Category | Tools | Use Case |
+|----------|-------|----------|
+| **Web Analytics** | GA4, Plausible, Fathom | Traffic, sessions, web conversions |
+| **Product Analytics** | Mixpanel, Amplitude, Heap, PostHog | In-product behavior |
+| **Session Recording** | Hotjar, FullStory, LogRocket | Individual sessions, heatmaps |
+| **Attribution** | Triple Whale, Rockerbox, Northbeam | Multi-touch attribution |
+| **Data Warehouse** | BigQuery, Snowflake, Databricks | Centralized storage |
+| **BI/Visualization** | Looker, Metabase, Tableau | Dashboards, exploration |
+| **CDP** | Segment, RudderStack, Jitsu | Data collection and routing |
+| **ETL/ELT** | Fivetran, Airbyte, dbt | Data transformation |
+| **Experimentation** | Statsig, Optimizely, LaunchDarkly | A/B testing, feature flags |
+
+### GA4 vs Universal Analytics
+
+| Aspect | Universal Analytics | GA4 |
+|--------|-------------------|-----|
+| Data model | Sessions-based | Events-based |
+| Tracking | Page views + events | Everything is an event |
+| User identity | Client ID | User ID + Google Signals |
+| ML/AI | Basic | Predictive metrics, anomaly detection |
+| Privacy | Cookie-dependent | Consent mode, modeling |
+| Reporting | Pre-built reports | Explorations (custom) |
+
+### Product Analytics Comparison
+
+| Tool | Strength | Best For |
+|------|----------|----------|
+| **Mixpanel** | Event-based, ad hoc queries, funnels | Self-serve analytics |
+| **Amplitude** | Behavioral cohorts, experimentation, CDP | Enterprise |
+| **Heap** | Autocapture, retroactive analysis | Teams with low eng bandwidth |
+| **PostHog** | Open-source, all-in-one (analytics + flags + testing) | Developer-first teams |
+
+### Attribution Models
+
+| Model | Logic | Best When |
+|-------|-------|-----------|
+| Last-click | Full credit to last click | Simple but misleading |
+| First-click | Full credit to first contact | Valuing awareness |
+| Linear | Equal credit distribution | All touchpoints matter |
+| Time-decay | More credit to recent touchpoints | Long sales cycles |
+| Position-based (U) | 40% first + 40% last + 20% middle | Balanced approach |
+| Data-driven | ML determines credit | Requires data volume |
+| Incrementality testing | Controlled experiments | Gold standard (most accurate) |
+
+### Unit Economics
+
+| Metric | Formula | SaaS Benchmark |
+|--------|---------|----------------|
+| **CAC** | Total acquisition spend / New customers | Varies by sector |
+| **LTV** | ARPU x Gross Margin x (1/Churn Rate) | LTV/CAC >= 3:1 |
+| **LTV/CAC** | LTV / CAC | >= 3:1 healthy |
+| **Payback Period** | CAC / (ARPU x Gross Margin) | < 12 months |
+| **ARPU** | Total revenue / Users | Depends on model |
+| **NRR** | (Starting MRR - churn + expansion) / Starting MRR | >100% = growth without acquisition |
+| **DAU/MAU** | Daily active / Monthly active | >25% good for SaaS |
+
+### Cohort Analysis
+
+Group users by shared characteristic (usually signup date) and analyze behavior over time.
+
+| Cohort Type | Grouped By | Use |
+|-------------|-----------|-----|
+| Acquisition | Signup date | Retention trends |
+| Behavioral | Action taken (e.g., "used feature X in week 1") | Feature impact |
+| Feature | Feature set used | Upgrade propensity |
+
+**Reading cohort tables:** If newer cohorts retain better, product changes are working. If the retention curve flattens, there's product-market fit. If it goes to zero, no amount of growth fixes it.
+
+---
+
+## 8. Experimentation & A/B Testing
+
+### Culture Principles
+
+1. **Speed matters** — Experiment velocity correlates with growth
+2. **Most fail** — 70-80% of experiments have no positive result (normal)
+3. **Hypothesis first** — Every experiment needs a falsifiable hypothesis
+4. **Data beats opinions** — Evidence over HiPPO (Highest Paid Person's Opinion)
+5. **Learning > Winning** — Experiments that teach are valuable even without positive results
+
+### Hypothesis Framework
+
+```
+If [change], then [expected outcome], because [logic/insight].
+We'll measure [primary metric] and expect [X% change] in [period].
+Guardrails: [metrics that should not degrade].
+```
+
+### Frequentist vs. Bayesian
+
+| Aspect | Frequentist | Bayesian |
+|--------|-------------|----------|
+| Question | "Probability of seeing this data if H0 is true?" | "Probability that A is better than B?" |
+| Output | p-value + confidence interval | Posterior probability |
+| Sample size | Fixed (calculated upfront) | Can stop early |
+| Interpretation | Technical (often misinterpreted) | Intuitive ("90% chance A is better") |
+| Peeking | Forbidden (inflates false positives) | Allowed (with care) |
+| Tools | VWO | Optimizely (STATS Engine), Statsig |
+
+### Statistical Significance
+
+- **Alpha** = 0.05 (5% false positive chance, industry standard)
+- **Power** (1-Beta) = 0.80 (80% chance of detecting real effect)
+- **MDE** = Minimum Detectable Effect (smallest effect worth detecting)
+
+### Multi-Armed Bandit (MAB)
+
+Alternative to classic A/B test that optimizes during the experiment. Progressively allocates more traffic to the better-performing variant.
+
+| Algorithm | Approach |
+|-----------|----------|
+| Epsilon-greedy | Explore X% traffic, exploit (1-X)% |
+| UCB | Balance uncertainty and performance |
+| Thompson Sampling | Sample from posterior distribution (Bayesian) |
+
+**Pros:** Minimizes regret, adapts automatically, ideal for continuous optimization.
+**Cons:** Less statistically rigorous, no clear p-value, may converge prematurely.
+
+### Experimentation Tools
+
+| Tool | Type | Differentiator |
+|------|------|---------------|
+| **Statsig** | Full-stack | Feature flags + experiments + analytics |
+| **Optimizely** | Enterprise | STATS Engine (Bayesian sequential) |
+| **VWO** | CRO-focused | Visual editor, heatmaps, testing |
+| **LaunchDarkly** | Feature flags | Feature management first |
+| **PostHog** | Open-source | Feature flags + experiments integrated |
+| **GrowthBook** | Open-source | Bayesian statistics, warehouse-native |
+| **AB Tasty** | Enterprise | Strong in Europe and Brazil |
+
+### Velocity Benchmarks
+
+| Metric | Definition | Benchmark |
+|--------|-----------|-----------|
+| Tests/month | Experiments launched | >10 for mature teams |
+| Time to launch | Idea to live | <1 week ideal |
+| Win rate | % with positive result | 15-30% normal |
+| Coverage | % features/pages with active tests | >50% for leaders |
+
+---
+
+## 9. Retention & Hook Model
+
+### Retention as Foundation
+
+Without retention, acquisition is a leaky bucket.
+
+**Types of retention:**
+- **User retention** — User returns to the product
+- **Revenue retention** — Revenue is maintained/expanded
+- **Engagement retention** — Usage level is maintained
+
+### Hook Model (Nir Eyal, "Hooked" 2014)
+
+```
+Trigger --> Action --> Variable Reward --> Investment
+   ^                                          |
+   +------------------------------------------+
+```
+
+**1. Trigger:**
+- **External:** Push notification, email, ad, social mention
+- **Internal:** Boredom, loneliness, uncertainty, FOMO (the ultimate goal)
+
+**2. Action:**
+- Simplest behavior in anticipation of reward
+- Must be extremely easy (BJ Fogg: Behavior = Motivation x Ability x Trigger)
+- Examples: open app, scroll feed, type search
+
+**3. Variable Reward:**
+- Variability is crucial — predictable rewards lose effect
+- **Tribe:** Social recognition (likes, comments)
+- **Hunt:** Searching for resources/info (infinite scroll, search results)
+- **Self:** Personal mastery (gamification, progress)
+
+**4. Investment:**
+- User invests something (time, data, content, reputation)
+- Increases switching costs and return probability
+- Examples: complete profile, create content, add friends, customize
+
+### Lifecycle Marketing Stages
+
+| Stage | Objective | Channels | Examples |
+|-------|----------|----------|----------|
+| Onboarding (D0-D7) | Activation, first value | Email, in-app, push | Interactive tutorial, checklist |
+| Activation (D7-D30) | Habit, recurring use | Email, in-app | Feature tips, case studies |
+| Growth (D30-D90) | Expansion, upgrade | Email, in-app, sales | PQL triggers, premium trial |
+| Maturity (D90+) | Retention, advocacy | Email, community | NPS, referral, advanced content |
+| Decline (usage drop) | Re-engagement | Email, push, retargeting | "We miss you", new features |
+| Churn (inactive) | Resurrection | Email, ads | Win-back offer, feature announcement |
+
+### Churn Analysis
+
+**Types:** Voluntary (user decides), Involuntary (payment failure), Logo (headcount), Revenue (MRR lost)
+
+**Leading Indicators of Churn:**
+- Drop in login frequency
+- Reduced core feature usage
+- Unresolved support tickets
+- Non-adoption of new features
+- Reduced active users in account
+
+**Analysis Methods:**
+1. Survival analysis (Kaplan-Meier curves)
+2. Cohort analysis (retention by cohort)
+3. Behavioral segmentation (behaviors preceding churn)
+4. Exit surveys
+5. Predictive modeling (ML for at-risk users)
+
+### Resurrection Campaigns
+
+1. **What's new** — Communicate improvements since user left
+2. **Win-back offer** — Temporary discount to return
+3. **Personalized value** — Show data/content user created that they might lose
+4. **Social proof** — Show platform growth and peer adoption
+5. **New use case** — Present a different use case
+
+---
+
+## 10. Viral & Referral Mechanics
+
+### Virality Types
+
+| Type | Mechanism | Example |
+|------|-----------|---------|
+| **Inherent** | Product usage exposes non-users | Zoom, Calendly |
+| **Collaboration** | Collaboration requires inviting others | Google Docs, Figma, Slack |
+| **Word-of-mouth** | Product so good users talk about it | Tesla, Superhuman |
+| **Incentivized** | Rewards for referrals | Dropbox, Uber |
+| **Social** | Sharing generates exposure | Spotify Wrapped, Canva designs |
+
+### K-Factor (Viral Coefficient)
+
+```
+K = i x c
+
+i = average invites sent per user
+c = conversion rate of invites
+
+Viral cycle time = average time for a new user to invite others
+```
+
+| K Value | Interpretation |
+|---------|---------------|
+| K > 1 | Sustainable viral growth (rare and powerful) |
+| K = 0.5-0.9 | Virality significantly amplifies other channels |
+| K = 0.1-0.4 | Virality contributes but doesn't lead |
+| K < 0.1 | Insignificant virality |
+
+> **Cycle time matters as much as K-factor.** K=0.8 with 1-day cycle is more powerful than K=1.2 with 30-day cycle due to compounding.
+
+### Referral Program Design
+
+**Essential Elements:**
+1. Clear incentive
+2. Double-sided rewards (referrer AND referred)
+3. Minimal friction (1-2 clicks to share)
+4. Right timing (after "aha moment", not during onboarding)
+5. Transparent tracking (user sees progress and rewards)
+6. Social proof (show how many friends already use it)
+
+### Classic Referral Cases
+
+| Company | Mechanism | Result |
+|---------|-----------|--------|
+| Dropbox | 500MB free per referral (both sides) | 3900% growth in 15 months |
+| PayPal | $10 referrer + $10 referred | 7-10% daily growth initially |
+| Uber | Ride credit for both | Primary initial expansion engine |
+| Airbnb | $25 travel credit | 25% of new users via referral in mature markets |
+| Revolut | Free card + premium features | 55% of new customers via referral |
+
+### Behavioral Economics in Incentive Design
+
+- **Loss aversion** — "You'll lose X" > "You can gain X"
+- **Social proof** — "32 of your contacts already use it"
+- **Reciprocity** — "Your friend gave you a gift"
+- **Scarcity** — "Limited invite supply"
+
+---
+
+## 11. CRO -- Conversion Rate Optimization
+
+### Foundation
+
+```
+Conversion Rate = Conversions / Visitors x 100
+```
+
+Doubling conversion rate = doubling revenue with the same traffic. CRO ROI is typically higher than incremental acquisition ROI.
+
+### Landing Page Elements
+
+1. **Headline** — Clear value proposition in <10 words
+2. **Sub-headline** — How you deliver the promised value
+3. **Hero image/video** — Visual demonstrating the product
+4. **Social proof** — Logos, testimonials, numbers
+5. **Benefits (not features)** — What the user gains
+6. **Single clear CTA** — One button, one action
+7. **Anxiety reduction** — Guarantee, security, no commitment
+8. **Congruence** — Ad message = landing page message
+
+### Form Optimization
+
+- Minimum fields (name + email to start)
+- Progressive profiling (collect more data over time)
+- Smart defaults (pre-fill with geolocation, etc.)
+- Inline validation (immediate error feedback)
+- Multi-step forms (sunk cost effect)
+- Social login (maximum friction reduction)
+
+### Pricing Page Best Practices
+
+1. **3 tiers** — Anchor, main option (highlighted), premium
+2. **Highlight recommended plan** — Visually differentiated
+3. **Annual vs Monthly toggle** — Show savings ("Save 20%")
+4. **Feature comparison table** — Transparency
+5. **FAQ** — Address common objections
+6. **Social proof** — Logos, customer count, testimonials
+7. **Free trial/freemium CTA** — Low-commitment option
+
+### Pricing Psychology
+
+| Technique | Mechanism |
+|-----------|-----------|
+| Anchoring | Show expensive plan first to make middle seem reasonable |
+| Decoy effect | Plan that exists only to make another look better |
+| Charm pricing | R$97 vs R$100 |
+| Value-based framing | "R$3/day" vs "R$90/month" |
+
+### Checkout Optimization
+
+Average cart abandonment: **~70.2%** (Baymard Institute, 50 studies, 2025-2026)
+
+| Cause | % Abandonment | Solution |
+|-------|---------------|---------|
+| Unexpected extra costs | 48% | Full cost transparency upfront |
+| Account creation required | 26% | Guest checkout |
+| Too complex process | 22% | Reduce steps, progress indicator |
+| Doesn't trust site | 18% | Security badges, SSL, reviews |
+| Slow delivery | 16% | Clear shipping options |
+| Errors/crashes | 13% | Performance, error handling |
+
+### Social Proof Types
+
+| Type | Example | Effectiveness |
+|------|---------|---------------|
+| Expert | "Recommended by [authority]" | High for YMYL |
+| Celebrity | Public figure endorsement | High for B2C |
+| User | Testimonials, reviews, ratings | High universal |
+| Wisdom of crowds | "50,000+ companies trust us" | High for B2B |
+| Wisdom of friends | "3 of your friends use it" | Highest (social) |
+| Certification | Badges, awards | Medium-high |
+
+---
+
+## 12. Brazilian Growth Context
+
+### Market Overview (2025-2026)
+
+| Metric | Value | Source |
+|--------|-------|--------|
+| Online population | ~183 million | DataReportal Jan 2025 |
+| Internet penetration | ~86.2% | DataReportal Jan 2025 |
+| Smartphones | ~170 million | GSMA |
+| Average time online/day | ~9h30 | DataReportal |
+| E-commerce GMV | ~R$200 billion/year | ABComm |
+| Social media users | ~144 million | DataReportal Jan 2025 |
+
+**Unique Characteristics:**
+- **Mobile-first** — 60%+ of traffic is mobile
+- **WhatsApp dominant** — 99% penetration, #1 communication channel
+- **Social media heavy** — Brazil is top 3 global in Instagram, TikTok, YouTube usage
+- **Google dominance** — 97%+ market share (vs 88% global)
+
+### PIX Impact on Conversion
+
+PIX (launched November 2020 by Banco Central) transformed Brazilian e-commerce:
+
+| Impact | Value |
+|--------|-------|
+| Checkout abandonment reduction | 30-40% |
+| Conversion vs boleto | +15-25% |
+| Merchant cost | ~0% (vs 2-5% credit card) |
+| Settlement | Instantaneous |
+| Financial inclusion | Users without credit cards can now pay online |
+
+> **For growth in Brazil, offering PIX is not optional — it is mandatory.**
+
+### LGPD Compliance
+
+Lei Geral de Protecao de Dados (Law 13.709/2018) impacts growth directly:
+
+| Area | LGPD Impact | Adaptation |
+|------|-------------|-----------|
+| Email marketing | Requires explicit consent | Double opt-in, preference center |
+| Analytics | Cookie consent required | Consent banner, cookieless analytics |
+| Retargeting | Limits data use for ads | First-party data strategy |
+| Personalization | Requires legal basis | Consent or legitimate interest |
+| Lead generation | Data use transparency | Privacy policy, clear purpose |
+
+**User Rights (Art. 18):** Access, correction, deletion, portability, consent revocation.
+
+### SEO in Brazilian Portuguese
+
+1. **Search volume** — Lower than English but less competitive
+2. **Accents** — Google treats "acai" and "acai" as equivalent (generally), optimize for both
+3. **Regionalisms** — "biscoito" vs "bolacha" impact keyword research
+4. **Competition** — Less quality PT-BR content = opportunity
+5. **Tools** — Semrush and Ahrefs have good Brazilian coverage
+
+### Brazilian Platforms
+
+| Platform | Category | Growth Relevance |
+|----------|----------|-----------------|
+| **RD Station** | Marketing automation | Inbound leader in Brazil (50K+ clients, acquired by TOTVS) |
+| **Hotmart** | Digital products | Largest infoproduct platform in LatAm ($10B+ cumulative GMV, 188 countries) |
+| **Eduzz** | Digital products | Hotmart alternative, affiliate focus |
+| **Monetizze** | Digital products | Third major infoproduct platform |
+| **VTEX** | E-commerce | Brazilian enterprise e-commerce |
+| **Nuvemshop** | E-commerce SMB | Brazilian Shopify for SMBs |
+
+### Infoproduct Growth Model (Brazil-specific)
+
+```
+Producer creates course → Affiliates promote → 30-70% commission
+  → Affiliates reinvest in ads → More sales → More affiliates → ...
+```
+
+**Growth Mechanisms Used:**
+- **Launches** — Jeff Walker's Product Launch Formula adapted to Brazil
+- **Webinars/Lives** — Free content, pitch, sale
+- **Affiliate marketing** — Armies of affiliates with commission
+- **WhatsApp groups** — Community for engagement and social proof
+- **Video testimonials** — Heavy social proof (Brazilian culture values this)
+
+---
+
+## 13. Prioritization Frameworks
+
+### ICE Scoring (Sean Ellis)
+
+| Criterion | Description | Scale |
+|-----------|-------------|-------|
+| **Impact** | Expected impact on target metric | 1-10 |
+| **Confidence** | Confidence in estimate | 1-10 |
+| **Ease** | Implementation ease | 1-10 |
+
+**Score = (I + C + E) / 3.** Simple, fast, ideal for small teams. Downside: subjectivity.
+
+### RICE Scoring (Intercom)
+
+| Criterion | Description | Calculation |
+|-----------|-------------|-------------|
+| **Reach** | Users impacted per period | Absolute number |
+| **Impact** | Impact per user | 0.25, 0.5, 1, 2, 3 |
+| **Confidence** | Estimate confidence | % (100%, 80%, 50%) |
+| **Effort** | Effort in person-months | Absolute number |
+
+**Score = (R x I x C) / E.** More rigorous than ICE. Includes Reach and Effort in absolute terms.
+
+### Retention-First Prioritization (Elena Verna)
+
+1. Map the retention curve (retention by cohort over time)
+2. If curve doesn't stabilize → Focus on product-market fit, NOT growth
+3. If curve stabilizes → Focus on optimizing each loop stage
+
+### Jobs-to-be-Done Applied to Growth
+
+- **Functional Job** — What the user wants to do
+- **Emotional Job** — How they want to feel
+- **Social Job** — How they want to be seen
+
+For growth, JTBD helps:
+1. Identify the correct "aha moment"
+2. Segment users by job, not demographics
+3. Position product vs alternatives (including "do nothing")
+4. Create messaging that resonates with real motivation
+
+---
+
+## 14. Data-Driven Decision Making
+
+### OKRs for Growth
+
+```
+Objective: Become the preferred platform for designers in Brazil
+
+Key Results:
+  KR1: Increase MAU from 5K to 15K
+  KR2: Improve NRR from 95% to 110%
+  KR3: Reduce time-to-first-value from 15min to 5min
+  KR4: Achieve K-factor of 0.5 in referral program
+```
+
+**Principles:** Objectives are qualitative and inspirational. KRs are quantitative. 3-5 KRs per Objective. 60-70% achievement is "healthy" (stretch goals). Quarterly cadence with weekly check-ins.
+
+### KPI Tree Example (SaaS)
+
+```
+MRR
++-- New MRR
+|   +-- Leads
+|   |   +-- Organic traffic x Conversion rate
+|   |   +-- Paid traffic x Conversion rate
+|   |   +-- Referrals x Conversion rate
+|   +-- Trial-to-Paid rate
+|   +-- Average deal size
++-- Expansion MRR
+|   +-- Upgrade rate
+|   +-- Cross-sell rate
++-- Churned MRR (negative)
+    +-- Logo churn rate
+    +-- Downgrade rate
+```
+
+### Vanity Metrics vs Actionable Metrics
+
+| Vanity Metric | Actionable Metric |
+|---------------|-------------------|
+| Total registered users | MAU (monthly active users) |
+| Total page views | Engagement rate, time on site |
+| App downloads | DAU/MAU ratio |
+| Social media followers | Engagement rate, click-through |
+| Total revenue | MRR growth rate, NRR |
+| "Impressions" | CTR, conversions |
+
+**Test:** If the metric went up, do you know what to do differently? If not, it's probably vanity.
+
+---
+
+## 15. Email & Lifecycle Marketing
+
+### Email as Growth Channel
+
+ROI: **$36-38 for every $1 invested** (DMA 2026)
+
+**Advantages:** Owned channel (no algorithm dependency), high ROI, personalizable, measurable, accompanies full lifecycle.
+
+### Segmentation Types
+
+| Type | Criterion | Example |
+|------|----------|---------|
+| Demographic | Role, company, sector | "CTOs at startups" |
+| Behavioral | Product actions | "Used feature X but not Y" |
+| Lifecycle stage | User phase | "Trial expires in 3 days" |
+| Engagement | Email interaction | "Opened last 5 emails" |
+| RFM | Recency, Frequency, Monetary | "High value, low frequency" |
+| Predictive | Action probability | "High churn probability" |
+
+### Drip Campaign Framework
+
+```
+Day 0:  Welcome + Quick Win (immediate value)
+Day 2:  Feature highlight #1 (core value)
+Day 5:  Case study (social proof)
+Day 8:  Feature highlight #2 (secondary value)
+Day 12: Educational content (thought leadership)
+Day 15: Upgrade offer / main CTA
+Day 20: Final follow-up (urgency)
+```
+
+### Deliverability Factors
+
+| Factor | Action |
+|--------|--------|
+| SPF, DKIM, DMARC | Configure all three |
+| Sender reputation | Monitor with Google Postmaster |
+| Bounce rate | Clean list regularly |
+| Spam complaints | Keep < 0.1% (Gmail threshold) |
+| Engagement | Segment by engagement level |
+| List hygiene | Remove inactives after 6 months |
+
+### ESP Comparison
+
+| ESP | Focus | Best For |
+|-----|-------|----------|
+| **Resend** | Developer-first, modern API | Transactional + marketing for devs |
+| **Customer.io** | Behavioral automation | SaaS with complex lifecycle |
+| **Brevo** | All-in-one, competitive pricing | SMBs, European/BR market |
+| **ActiveCampaign** | Automation + CRM | SMBs with active sales |
+| **HubSpot** | Full marketing suite | Enterprise, integrated CRM |
+| **Klaviyo** | E-commerce | Online stores (Shopify) |
+| **RD Station** | Brazil-focused | Brazilian companies |
+
+---
+
+## 16. AI & Growth
+
+### AI-Powered Personalization
+
+| Application | Description | Example |
+|-------------|-------------|---------|
+| Product recommendations | ML suggests items by behavior | Netflix, Spotify Discover Weekly |
+| Dynamic pricing | Prices adjusted by demand/profile | Airlines, Uber surge |
+| Content personalization | Homepage/emails customized per user | Amazon, Netflix UI |
+| Onboarding personalization | Flow adapted to user profile | Welcome survey, personalized path |
+
+### Predictive ML Models
+
+| Model | Application | Impact |
+|-------|-------------|--------|
+| Churn prediction | Identify at-risk users | Proactive intervention |
+| LTV prediction | Estimate future user value | Optimize CAC by segment |
+| Propensity scoring | Conversion/upgrade probability | Prioritize outreach |
+| Next-best-action | Which action maximizes engagement | Personalized triggers |
+| Anomaly detection | Identify anomalous changes | Automatic alerts |
+
+### GenAI for Content — Recommended Approach
+
+**Use AI as co-pilot, not autopilot.** Accelerate drafts and variations, but always with human editing, original data, and real expertise.
+
+**Risks:** Average quality, E-E-A-T penalties, commoditized content, brand voice inconsistency, hallucinations.
+
+### LLM-Era SEO Strategy
+
+- Create content that LLMs cite (original data, authority)
+- Optimize for Perplexity and ChatGPT (structured data, authoritative sources)
+- Focus on experience and expertise AI can't replicate
+- Diversify traffic sources (don't depend solely on Google)
+
+---
+
+## 17. Key People & Bibliography
+
+### Key People
+
+| Person | Contribution | Company |
+|--------|-------------|---------|
+| Sean Ellis | Coined "growth hacking" | Dropbox, LogMeIn, Eventbrite |
+| Andrew Chen | Growth essays, The Cold Start Problem | a16z (ex-Uber) |
+| Brian Balfour | Reforge, Growth Loops, RARRA | Reforge (ex-HubSpot) |
+| Casey Winters | Growth advising, retention | Grubhub, Pinterest |
+| Lenny Rachitsky | #1 growth/product newsletter | Lenny's Newsletter (ex-Airbnb) |
+| Nir Eyal | Hook Model | Author of Hooked |
+| Peep Laja | CRO, experimentation | CXL Institute |
+| Elena Verna | PLG, growth advising | Miro, Amplitude, Dropbox |
+| Wes Bush | Product-Led Growth | ProductLed |
+| Rand Fishkin | SEO, founder transparency | SparkToro, Moz |
+| Eric Ries | Lean Startup | Author |
+| Dave McClure | AARRR/Pirate Metrics | 500 Startups |
+
+### Essential Books
+
+| Book | Author | Core Contribution |
+|------|--------|-------------------|
+| Hacking Growth (2017) | Sean Ellis | Growth hacking process |
+| The Lean Startup (2011) | Eric Ries | Build-Measure-Learn, MVP |
+| The Cold Start Problem (2021) | Andrew Chen | Network effects |
+| Hooked (2014) | Nir Eyal | Habit-forming products |
+| Product-Led Growth (2019) | Wes Bush | Complete PLG framework |
+| Influence (1984/2021) | Robert Cialdini | 7 persuasion principles (CRO base) |
+| Traction (2015) | Weinberg & Mares | 19 traction channels |
+| Lean Analytics (2013) | Croll & Yoskovitz | Metrics by business type and stage |
+| Crossing the Chasm (1991) | Geoffrey Moore | Technology adoption lifecycle |
+
+### Learning Resources
+
+| Resource | Type | Focus |
+|----------|------|-------|
+| Reforge | Course/Community | Growth frameworks (premium) |
+| CXL Institute | Course/Certification | CRO, analytics, growth marketing |
+| Lenny's Newsletter | Newsletter | Growth, product management |
+| GrowthHackers | Community | Growth hacking discussions |
+| NFX | VC/Blog | Network effects |
+| OpenView Partners | VC/Blog | PLG research |
+
+---
+
+## 18. Actionable Checklists
+
+### Pre-Launch Growth Checklist
+
+- [ ] North Star Metric defined with guardrails
+- [ ] At least one compound growth loop identified
+- [ ] Onboarding flow designed for <5min to first value
+- [ ] Analytics stack configured (GA4 + product analytics)
+- [ ] Event taxonomy documented
+- [ ] Cohort analysis dashboards set up
+- [ ] Referral/sharing mechanism designed into product
+- [ ] Content strategy with pillar-cluster architecture
+- [ ] Email welcome series (5+ emails) created
+- [ ] PIX payment enabled (Brazil)
+- [ ] LGPD consent mechanism implemented (Brazil)
+- [ ] Mobile-first design validated
+
+### Weekly Growth Review Checklist
+
+- [ ] NSM trend reviewed (vs previous week and target)
+- [ ] Retention curve by latest cohorts analyzed
+- [ ] Active experiments status checked
+- [ ] Experiment results documented and decided (ship/iterate/kill)
+- [ ] New experiment hypotheses prioritized (ICE/RICE)
+- [ ] Channel performance reviewed (organic, paid, referral)
+- [ ] Funnel conversion rates checked for anomalies
+- [ ] Churn leading indicators monitored
+
+### SEO Audit Checklist
+
+- [ ] Core Web Vitals passing (LCP <2.5s, INP <200ms, CLS <0.1)
+- [ ] Mobile-first indexing verified
+- [ ] XML sitemap submitted and up to date
+- [ ] Robots.txt correctly configured
+- [ ] Canonical tags on all pages
+- [ ] Structured data (Schema.org) implemented
+- [ ] Internal linking structure reviewed
+- [ ] HTTPS enforced
+- [ ] Page titles and meta descriptions optimized
+- [ ] Image alt text present
+- [ ] 404 errors monitored and redirected
+- [ ] Content freshness audit (update evergreen content)
+
+### Experimentation Launch Checklist
+
+- [ ] Hypothesis documented (If/Then/Because)
+- [ ] Primary metric and guardrails defined
+- [ ] Sample size calculated (alpha=0.05, power=0.80)
+- [ ] Expected duration estimated
+- [ ] Variants implemented and QA'd
+- [ ] Tracking verified (events firing correctly)
+- [ ] Experiment documented in tracking system
+- [ ] Results review date scheduled
+
+### CRO Quick Wins Checklist
+
+- [ ] Single CTA per landing page
+- [ ] Social proof visible above the fold
+- [ ] Form fields reduced to minimum
+- [ ] Page load time <3 seconds
+- [ ] Mobile experience tested
+- [ ] Value proposition in headline (not company name)
+- [ ] Guest checkout available
+- [ ] Trust badges/security indicators present
+- [ ] PIX as first payment option (Brazil)
+- [ ] Exit-intent mechanism configured
+
+### PLG Health Checklist
+
+- [ ] Self-serve signup works without friction
+- [ ] Time-to-first-value measured and optimized
+- [ ] Activation rate tracked by cohort
+- [ ] PQL scoring model defined and calibrated
+- [ ] Freemium limits set to encourage upgrade (not frustrate)
+- [ ] In-product upgrade prompts at right moments
+- [ ] Usage-based triggers for sales outreach configured
+- [ ] Viral loops embedded in core product flows
+
+---
+
+> **Data sources:** 42+ verified sources including DataReportal, BrightEdge, SparkToro/Datos, Baymard Institute, Seer Interactive, DMA, Reforge, CXL Institute. Full source list in MS-004 research document.
+>
+> **Last verified:** 2026-04-07 by @research-orqx (Prism)