BackStartups

Book · How companies get built

The Cold Start Problem

Andrew Chen · 2021

How network effects actually get started, sustained and defended — the most rigorous available treatment of the term everybody uses loosely.

Why it is here

Network effects are claimed in a very large proportion of pitches and understood in a much smaller one. This book takes the concept apart: how a network gets its first users, why it stalls, what makes it defensible, and why most claimed network effects are local rather than global.

That last point is the one an angel most needs. A marketplace with a network effect in each city is a collection of small businesses, not one large one, and it is attackable city by city. Our marketplace diligence guide asks the question; this book explains why it is the right one.

The author spent years working on growth at a company where these dynamics were the whole business, so the material is operational rather than theoretical.

Read it if

You look at marketplaces, social products or anything where value depends on other users being present.

Skip it if

You invest exclusively in business software sold seat by seat.

Pair it withZero to OnePeter Thiel with Blake MastersA short, argumentative book about monopoly, competition and why the interesting companies are the ones doing something nobody else is.

The Cold Start Problem: common questions

What is the most common misuse of "network effect"?
Claiming a national one for a business that really has many local ones. If each new user only makes the product more valuable to users in the same city or category, the company has a set of small, separately attackable networks rather than one defensible large one.
Is this useful outside marketplaces?
Yes, for any product where value depends on other users — social products, collaboration tools, developer platforms. It is least useful for software sold seat by seat to unrelated customers.