Building companies
Zero to One
Notes on startups: real progress comes from creating something new (0 to 1), not copying (1 to n).
Key concepts
- What important truth do very few people agree with you on?
- Competition is for losers; aim for a monopoly in a small market, then expand
- Durable advantage: proprietary tech, network effects, economies of scale, brand
- Power law: a few bets drive most returns
- Founder and team matter more than the plan
How I use it
What stuck with me isn’t a framework, it’s the split itself, and I’ve extended it into how I think about jobs. Zero to one is when there’s nothing. Whatever you build is a huge improvement, but it asks you to live with ambiguity and mess, be scrappy, and get the outcome right before worrying about scale or governance. What I call one to two is when something already works and has to go to the next level. That’s often harder, because you’re fighting settled ways of working and breaking the status quo.
When I read it, I realised almost every role in my career had been one to two. Building my own businesses is the first time I’m doing zero to one, and I’m still learning it. One to two has feedback loops, because there’s a baseline. Zero to one often has none, and some days it’s hard to see whether it will work at all.
I use the split when I look at teams too: is this a zero-to-one person or a one-to-two person? They need different ways of thinking, and a different appetite for risk.
Connected books
The Hard Thing About Hard Things
Elon Musk