How to Start a Startup

Key ideas

Graham argues a startup needs three doable things: good people, a product customers actually want, and spending as little money as possible.

1
Execution and people beat the idea
Graham argues a startup doesn't need a brilliant idea, just a way to do something people already attempt without it being terrible. Ideas mostly matter as starting points for the people who had them.
2
Hire animals, and understand users
He recommends founding with two to four people and hiring those who are obsessive about their work. Founders should include technical people, and someone must be able to understand users.
3
Make what customers want, fast
In his view most failed startups made something nobody wanted; running out of money is just the immediate cause. The remedy is to put a prototype in front of users early and refine it from their reactions, ideally aiming at smaller customers and the low end.
4
Raise money, but treat it as a bet
Graham sees a startup as pass/fail, so trading stock for better odds is usually wise. Seed money is comparatively easy to get, valuations have no rational basis, and founders should check everyone's prior IP agreements.
5
Get big slow and stay cheap
He argues investor money isn't revenue, so founders should spend slowly, avoid unnecessary hires, and work from cheap spaces. Growing slowly gives time to learn from smart early users.

“Good people can fix bad ideas, but good ideas can't save bad people.”

“The only way to make something customers want is to get a prototype in front of them and refine it based on their reactions.”

“Financially, a startup is like a pass/fail course.”

Cover photo of Paul Graham, not from this work: Kevin Hale, CC BY-SA 2.0, via Wikimedia Commons.

Original source
Linked. Opens on the publisher’s site.
Player on this page
Not applicable. Written piece; nothing to play.
Full text or transcript
Not reviewed. No license to show a text version here.
Our summary
Draft. Written by us with AI from the full public text and checked against it; quotes are word for word. Not yet rights-reviewed.
Ask questions with AI
Not reviewed. Off until there is licensed text to work from.