Build growth systems — activation funnels, referral programs, lifecycle automation, cohort analysis, and product-led growth patterns
npx claudepluginhub cure-consulting-group/productengineeringskillsThis skill uses the workspace's default tool permissions.
Before starting, gather project context silently:
Generates design tokens/docs from CSS/Tailwind/styled-components codebases, audits visual consistency across 10 dimensions, detects AI slop in UI.
Records polished WebM UI demo videos of web apps using Playwright with cursor overlay, natural pacing, and three-phase scripting. Activates for demo, walkthrough, screen recording, or tutorial requests.
Delivers idiomatic Kotlin patterns for null safety, immutability, sealed classes, coroutines, Flows, extensions, DSL builders, and Gradle DSL. Use when writing, reviewing, refactoring, or designing Kotlin code.
Before starting, gather project context silently:
PORTFOLIO.md if it exists in the project root or parent directories for product/team contextcat package.json 2>/dev/null || cat build.gradle.kts 2>/dev/null || cat Podfile 2>/dev/null to detect stackgit log --oneline -5 2>/dev/null for recent changesls src/ app/ lib/ functions/ 2>/dev/null to understand project structureBuild systematic, measurable growth into your product. Growth is not marketing — it is engineering loops that compound over time.
| Need | Output |
|---|---|
| Activation optimization | Time-to-value reduction plan + experiment backlog |
| Retention improvement | Engagement loop design + re-engagement automation |
| Referral system | Viral mechanics + invite flow + reward structure |
| Monetization experiment | Paywall strategy + conversion triggers + pricing tests |
| Lifecycle automation | Drip campaigns + behavioral triggers + segmentation rules |
| Cohort analysis | Retention curves + segmented metrics + experiment design |
Build your growth model top-down. Measure every stage:
Acquisition → How do users find you?
Metrics: CAC, signups/week, channel mix, organic vs paid ratio
Goal: lower CAC, diversify channels
Activation → Do they experience the core value?
Metrics: signup-to-activation rate, time-to-value, onboarding completion
Goal: >60% of signups reach activation within first session
Retention → Do they come back?
Metrics: D1, D7, D30 retention, DAU/MAU ratio, churn rate
Goal: D30 retention >20% (consumer), >40% (B2B SaaS)
Revenue → Do they pay?
Metrics: conversion rate, ARPU, LTV, LTV:CAC ratio
Goal: LTV:CAC > 3:1, payback period < 12 months
Referral → Do they invite others?
Metrics: viral coefficient (K), invite send rate, invite accept rate
Goal: K > 0.3 (assists growth), K > 1.0 (viral growth)
Model the funnel numerically:
1000 visitors → 200 signups (20% CVR) → 80 activated (40%) →
32 retained D30 (40%) → 10 paid (31%) → 3 referrals (30%)
Each referral brings 2 visitors → 6 new top-of-funnel
Track weekly. Build a spreadsheet. Every optimization compounds.
Activation is the most important growth lever. Fix activation before scaling acquisition.
Identify your aha moment:
1. List all user actions in first 7 days
2. Segment users: retained (active at D30) vs churned
3. Compare action frequencies between groups
4. The action with highest retained-vs-churned delta = aha moment
Time-to-value optimization:
- Measure time from signup to aha moment (median, p75, p90)
- Target: aha moment reachable in first session (<5 minutes)
- Remove every step between signup and core value
- Pre-fill data, use smart defaults, defer non-essential setup
Onboarding experiments (run sequentially, not in parallel):
1. Reduce signup fields (test social auth only vs email)
2. Add interactive walkthrough vs passive tooltips
3. Personalize first screen based on signup intent
4. Pre-populate with sample data vs empty state
5. Guided first action vs free exploration
Progressive disclosure rules:
- Day 0: core feature only, everything else hidden
- Day 1-3: introduce secondary features via contextual prompts
- Day 7+: full feature set available, advanced features highlighted
- Never show settings, preferences, or admin during onboarding
Empty state strategy:
- Every empty screen is a growth opportunity
- Show what the screen looks like when populated (preview/demo data)
- Single CTA: "Create your first [X]"
- Include social proof: "[N] users created their first [X] today"
Retention is a product problem, not a marketing problem. Build retention into the product.
Engagement loops (every product needs at least one):
Trigger → Action → Reward → Investment
Example (fitness app):
Trigger: push notification "Your workout streak is at 5 days"
Action: complete today's workout
Reward: streak badge + progress visualization
Investment: workout history + streak count (loss aversion)
Example (SaaS):
Trigger: email "3 teammates commented on your doc"
Action: open app, review comments
Reward: collaborative progress, social validation
Investment: content created, team context built
Re-engagement triggers:
At-risk detection (fire before user churns):
- No session in 2 days → push notification (value reminder)
- No session in 5 days → email (what they're missing)
- No session in 14 days → email (personal outreach from founder)
- No session in 30 days → win-back offer (discount, extended trial)
Implementation:
- Cloud Function / cron job checks last_active_at daily
- Segment users into: active, cooling, at-risk, dormant, churned
- Different messaging per segment — never generic blasts
Streak and habit mechanics:
- Show streak count prominently (don't hide it)
- Grace period: 1 missed day doesn't break streak (reduces anxiety)
- Weekly cadence is easier than daily for most products
- Milestone rewards at 7, 30, 100 (celebration + shareable badge)
Push notification strategy:
- Transactional: immediate (someone replied to you)
- Re-engagement: max 2/week, never between 9pm-9am local time
- Always include specific content ("Alex commented on your design")
- Never generic ("Check out what's new!")
- A/B test copy, timing, and deep link destination
Referral systems only work if the product is already good. Fix retention first.
Viral coefficient calculation:
K = (invites sent per user) x (conversion rate of invites)
Example: each user sends 3 invites, 15% accept → K = 0.45
K > 1.0 = organic viral growth (rare, don't plan for it)
K > 0.3 = meaningful growth assist (realistic target)
Invite flow design:
1. Trigger: surface invite prompt after user experiences value
- After first success: "Share [X] with your team"
- After milestone: "You've completed 10 workouts — invite a friend"
- Never: during onboarding, during errors, randomly
2. Mechanism: pre-filled message with deep link
- SMS/iMessage (highest conversion on mobile)
- Email (B2B, professional context)
- Share sheet (let user pick channel)
- Copy link (fallback, lowest friction)
3. Landing: recipient sees personalized invite page
- "[Sender name] invited you to [Product]"
- Show what sender has achieved (social proof)
- One-tap signup (pre-fill referral code)
Reward structures:
Two-sided rewards outperform one-sided 2-3x:
"Give $10, get $10" > "Get $10 for each referral"
Reward types (in order of effectiveness):
1. Product value: extra storage, premium features, extended trial
2. Credit: account credit toward subscription
3. Cash: direct payment (expensive, attracts low-quality referrals)
Anti-fraud:
- Require referred user to activate (not just sign up)
- Cap rewards per user per month
- Detect duplicate devices/IPs
- Delay reward payout 7 days (reversible if fraudulent)
Platform-specific share integration:
iOS: UIActivityViewController, ShareLink (SwiftUI)
Android: Intent.ACTION_SEND with chooser, ShareSheet API
Web: Web Share API (navigator.share), fallback to copy-link
Cohort analysis framework:
Define cohorts by signup week (default) or by:
- Acquisition channel (organic vs paid vs referral)
- Platform (iOS vs Android vs Web)
- Plan tier (free vs paid)
- Geography (US vs international)
- Feature exposure (saw feature vs control)
Retention curve analysis:
Week 0 Week 1 Week 2 Week 4 Week 8 Week 12
100% 35% 28% 22% 18% 16%
Healthy curve: flattens after week 4 (found core users)
Unhealthy curve: never flattens (no product-market fit)
Compare curves across cohorts. If paid users retain 3x better
than free users, your activation problem is really a targeting problem.
A/B test design for growth:
1. Hypothesis: "Reducing signup to 1 field will increase completion by 20%"
2. Primary metric: signup completion rate
3. Guardrail metrics: activation rate, D7 retention (watch for quality drop)
4. Sample size: calculate before launch (minimum detectable effect, power)
5. Duration: minimum 2 weeks, capture weekly patterns
6. Analysis: check guardrails first, then primary metric
Statistical significance:
- p < 0.05 for primary metric (standard)
- Run until significance OR pre-determined end date
- Never peek and stop early on a positive result
- Use sequential testing if you must monitor continuously
Guardrail metrics (always monitor):
- Activation rate (are we getting worse users?)
- Revenue per user (are we cannibalizing?)
- Support ticket rate (are we creating confusion?)
- Error rate (is the variant broken?)
Experiment velocity:
Target: 2-3 growth experiments per week
Each experiment: hypothesis, design, ship, measure, document
Keep a log: experiment name, hypothesis, result, learning
Share learnings weekly — compound institutional knowledge
PLG means the product is the primary driver of acquisition, conversion, and expansion.
Freemium conversion:
Free tier must be genuinely useful (not a crippled demo):
- Enough value to create habit and dependency
- Limits hit naturally through usage growth
- Upgrade prompt at moment of need, not randomly
Effective limits:
Good: 3 projects (hit when engaged), 1GB storage (hit when invested)
Bad: 7-day trial (arbitrary), watermark on exports (punitive)
Usage-based upgrade triggers:
- Approaching limit: "You've used 8 of 10 projects" (banner)
- Hit limit: "Upgrade to create unlimited projects" (modal, not blocking)
- Team growth: "Invite your 4th teammate — upgrade to Team plan"
- Feature gate: "Export to PDF is a Pro feature" (show preview first)
Timing matters:
- Trigger at moment of highest intent (user trying to do the thing)
- Never interrupt a flow mid-action
- Offer monthly AND annual (anchor on annual savings)
Paywall optimization:
- Show value received so far: "You've created 12 designs with [Product]"
- Social proof: "Join 50,000 teams using Pro"
- Risk reversal: "14-day money-back guarantee"
- Price anchoring: show enterprise tier to make pro look reasonable
Trial design:
Reverse trial (recommended):
- Start with full features unlocked
- After 14 days, downgrade to free tier
- User has already experienced premium value → higher conversion
Traditional trial:
- 14 days of premium, then paywall
- Requires credit card upfront (higher intent) vs no card (higher volume)
- Cure default: no credit card required, reverse trial model
Expansion revenue:
- Seat-based: charge per user, team growth = revenue growth
- Usage-based: charge per API call, storage, compute
- Feature-based: unlock tiers as needs grow
- Best: hybrid seat + usage (predictable base + upside)
Generate growth infrastructure using Write:
analytics/cohort-retention.sql — BigQuery/PostgreSQL retention queryscripts/sample-size-calculator.ts — statistical significance calculatorsrc/referral/schema.ts — Firestore/PostgreSQL schema for referral trackingmonitoring/growth-dashboard.json — dashboard config with AARRR metrics/analytics-implementation — event taxonomy and tracking for growth metrics/feature-flags — experiment infrastructure for A/B tests/customer-onboarding — activation flow design (growth-engineering extends this)/notification-architect — push/email delivery for retention campaigns/saas-financial-model — unit economics that growth must satisfy/stripe-integration — payment and subscription mechanics for monetization