Do Things That Don't Scale
by Paul Graham · read the original
Startups don't take off by themselves; founders make them take off. PG's canonical essay on manual recruitment, delighting early users, and why the unscalable work you're avoiding is the actual job.
- 1
Recruit Users Manually
Manual, one-at-a-time user recruitment feels beneath you and pointless. Compound growth says otherwise: 100 users growing 10% weekly is 2 million in two years.
- 2
Fragile Is Not the Same as Doomed
Don't dismiss your own startup because it's small. Judge it by what it could become with great execution, then go execute.
- 3
Delight: Be Insanely Attentive
Over-serving early users is never the mistake. Small size lets you deliver a level of care incumbents can't, and it hardens into culture.
- 4
Perfect the Experience, Not Just the Product
An incomplete product delivered with extreme attentiveness beats a polished product delivered with indifference, because attentiveness is also how you learn.
- 5
Start a Contained Fire
Dominate a deliberately small pond first. Critical mass in a narrow market beats invisibility in a big one.
- 6
Pulling a Meraki, and the Consulting Move
Build the first units and serve the first customer by hand. The learning is the point, and it de-risks scaling. Just don't start billing hourly.
- 7
Do It by Hand Before You Automate
Be the Wizard of Oz. Manual operations launch faster and give you muscle memory for what the software should actually do.
- 8
Big Launches Don't Matter; Ideas Are Vectors
Launches and partnerships are not growth strategies. Your idea = what you build + the unscalable effort to ignite it. Plan both.