Skip to content
Deepak Gupta markBooks

Books

The Cold Start Problem

Editorial pick

How to Start and Scale Network Effects

By Andrew Chen · Harper Business · 2021

The first serious book on network-effect product strategy that isn't recycled blog content.

Framework 350+ pages(368p) Intermediate Published 2021
Amazon Publisher Goodreads Author site

Editorial take

Chen — a16z partner, formerly Uber's growth lead — wrote the canonical book on how network-effect businesses actually start, scale, and decay. The framework (cold start → tipping point → escape velocity → ceiling → moat) gives you precise vocabulary for problems that are usually waved at: why marketplaces die, why social products plateau, why some networks compound and others don't. The Uber war stories are the strongest case study material. Read it before you propose a network-effect strategy at your company; it will save you from a lot of magical thinking.

Last hand-checked 2026-05-18.

Read if you …

  • are building a marketplace, social, or collaboration product with network dynamics
  • lead growth at a product where 'engagement' is more important than 'acquisition'
  • evaluate marketplace investments and want a structured diagnostic lens

Skip if you …

  • you build pure SaaS with no network effects — most chapters don't apply
  • you wanted hands-on growth tactics — this is strategy-first, tactics-light

If you only read one chapter

The Atomic Network

The cleanest articulation in print of how networks must start as tiny dense atoms, not broad shallow launches.

Key ideas

  • Start with the atomic network — the smallest unit where utility exists.
  • Hard side first: every network has a hard side (drivers, hosts, creators) and an easy side; the hard side rate-limits everything.
  • Tipping point is reached one micro-network at a time, not in one giant push.
  • All networks face anti-network effects at scale — saturation, congestion, low quality. Plan for them.

About the book

Andrew Chen's synthesis of a decade running and observing network-effect products at Uber and a16z. The book introduces a five-stage framework — Cold Start, Tipping Point, Escape Velocity, Ceiling, Moat — and works through each with extended case studies of Uber, Slack, Tinder, Dropbox, Zoom, and others.

The writing is essayistic; the book reads more like a structured book of essays than a tight monograph. The frame stays useful long after specific examples date.

Pairs with

If The Cold Start Problem works for you, these likely will too.