
The Cold Start Problem summary
How to Start and Scale Network Effects
The book in one sentence
Products that get more valuable as more people use them are brutally hard to launch and hard to kill, and the way to win is to spark one small self-sustaining network, then grow and defend it through a set of predictable stages.
Overview
Some products are only good once other people use them too. A dating app with no matches is just a lonely screen, and a chat tool is dead without your coworkers. Andrew Chen, a partner at Andreessen Horowitz, explains why these networked products are so hard to start and how the winners finally broke through.
Key idea 1: Why Networked Products Feel Empty at First
Some products are only good when other people already use them. A chat app is dead without your coworkers on it. A rideshare app is useless with no drivers. A dating app with no matches is just a lonely screen. This is the chicken-and-egg trap at the heart of networked products: nobody joins because nobody's there, and nobody's there because nobody joins. Chen goes further and names a hidden force that makes it worse. He calls it the anti-network effect. A nearly empty network doesn't just sit quietly waiting to fill up. It actively pushes people away. An empty chat room feels abandoned. A rideshare app that shows no cars for half an hour makes you delete it. A newcomer shows up to nothing and leaves, and the network gets a little weaker each time. That is the opposite of the pull that makes huge networks feel unstoppable. The whole book is a map for getting from zero, where the network works against you, to the point where its own size does the work for you. Chen splits a network's life into five stages: the cold start, the tipping point, escape velocity, the ceiling, and the moat. He spends the most care on the first one because it is where nearly every networked product dies. Picture Uber in a city on day one. A handful of drivers means long waits, so riders give up, so the few drivers sit idle and stop logging on. Both sides drain at once. And you can't buy your way out with ads, because ads only bring people to a product that still has nothing to offer them. The fix is a small cluster of users dense enough to make the product feel alive to each other. Broad marketing to a broad audience won't do it. That first climb out of zero is the hardest thing a networked product ever does, and it sets up everything that follows.
7 more key ideas in the app
- 2. Start With the Smallest Network That Works
- 3. Build for the Hard Side of the Network
- 4. Come for the Tool, Stay for the Network
- 5. Tricks for Filling an Empty Network
- 6. Growth Runs on Three Engines at Once
- 7. Why Growth Stalls, and What Is Really Happening
- 8. Networks as a Moat, and How Rivals Attack Them
Did you know?
- Chen led rider growth teams at Uber before becoming a partner at the venture firm Andreessen Horowitz, and many of the book's examples come straight from launching Uber city by city.
- The book grew out of years of his widely read essays on startups. One of them, the Law of Shitty Clickthroughs, notes that the web's first banner ad in 1994 drew a 44 percent clickthrough rate, while today's barely register.
The E-Myth Revisited
The Everything Store
The First 90 Days
The Four
The Hard Thing About Hard Things