growth-experimenter

Growth Experimenter

Run systematic experiments to grow faster through data-driven optimization.

Core Philosophy

Growth = Experimentation Velocity × Win Rate × Impact per Win

• Run more experiments
• Increase your hit rate through better hypotheses
• Focus on high-impact areas

Growth Model (AARRR / Pirate Metrics)

Acquisition → Activation → Retention → Revenue → Referral
    ↓             ↓            ↓          ↓         ↓
  Traffic      Sign Up      Day 30    Upgrade   Invites
  100%          40%          50%        20%       10%

Example: 10,000 visitors/month
→ 4,000 signups (40%)
→ 2,000 active at D30 (50%)
→ 400 paying (20%)
→ 40 referrals (10%)

Improve ANY metric by 10% = 10% more customers

Where to focus first: The leakiest bucket

• If 40% sign up but only 10% are active at D30 → Fix retention
• If 80% are active but only 5% pay → Fix monetization
• If 2% visitors sign up but 60% convert to paid → Get more traffic

Experiment Framework

1. Identify the Problem

Good problem statements:

• "Only 2% of homepage visitors sign up" (specific metric)
• "50% of trials don't complete onboarding" (clear drop-off)
• "Users who invite teammates have 3x retention, but only 10% invite" (known behavior)

Bad problem statements:

• "We need more growth" (too vague)
• "Conversion is bad" (no baseline)
• "Users don't understand the product" (not measurable)

2. Form a Hypothesis

Hypothesis template:

We believe that [change]
will result in [outcome]
because [reason/evidence]

Examples:

✅ Good:
We believe that adding social proof (testimonials) to the pricing page
will increase trial signups by 10%
because visitors currently have low trust and need validation.

✅ Good:
We believe that sending a Slack notification when user completes setup
will increase D7 activation by 20%
because users forget to come back after initial signup.

❌ Bad:
We believe that changing the button color will improve conversions
(no reason why)

❌ Bad:
We believe that improving the product will increase retention
(too vague, not testable)