Part 1: Product-market fit, MVPs and A/B testing

Module 8 · Sat 5 Sep

Product-market fit (PMF) = a small but meaningful group deeply values your product because it solves a real, painful problem well, at the right time. Signals: retention, recommendations, organic spread, willingness to pay. Revenue alone is not PMF. PMF comes before scaling.

• Deep problems are painful and persistent - people act on them. Shallow problems sound good but behaviour doesn't change. Niche is fine if the problem is deep.
• Class example - Superhuman asked "How would you feel if you could no longer use this?" and aimed for ~40% "very disappointed". ChatGPT's organic spread with tiny marketing spend was another PMF signal.
• An MVP is the smallest version that teaches you the most - learning, not completeness. Class examples: Airbnb (one listing), Dropbox (demo video), Zappos (manual photos and shipping), WhatsApp-based ordering.

Start from a hypothesis: population + treatment + outcome metric. Keep the treatment atomic (one change). Pick the cheapest test that answers the question.

A/B testing proves cause, not just correlation: control vs treatment, random assignment, sample size based on effect size and confidence. Watch for bias and spillover, track guardrail metrics, and run sanity checks before trusting results.