The Cold Start Problem
Author: Andrew Chen (GP at a16z, former head of Uber rider growth) Published: 2021 URL: https://www.coldstart.com/
Summary
The definitive book on how networked products solve chicken-and-egg problems. Chen breaks network growth into 5 stages (Cold Start → Tipping Point → Escape Velocity → Ceiling → Moat) and introduces critical concepts: the atomic network (smallest self-sustaining unit), the hard side vs easy side (sellers are harder to get than buyers), and anti-network effects (the destructive force that drives new networks to zero). Draws from interviews with founders at Slack, Clubhouse, Zoom, Twitch, Tinder, Reddit, Uber, Airbnb, PayPal.
The 5 Stages of Network Growth
- Cold Start — Chicken-and-egg problem. Anti-network effects destroy the network unless you intervene.
- Tipping Point — Network has enough density to self-sustain in one atomic network.
- Escape Velocity — Network grows exponentially as cross-side effects compound.
- Hitting the Ceiling — Growth slows due to saturation or degradation. Requires active management.
- The Moat — Network becomes defensible. Competitors can’t replicate without your scale.
Key Claims
- The atomic network is the smallest stable, engaged network that can self-sustain
- Anti-network effects are real — new networks want to go to zero, not grow
- The “hard side” (sellers on marketplace, creators on platform) must be solved first
- Facebook (colleges), Uber (cities), Slack (teams) all used atomic network strategy
- Network effects asymptote — plateau management is its own discipline
- The moat isn’t the network itself — it’s the network’s engaged quality