Do Things That Don’t Scale
Source: https://paulgraham.com/ds.html Author: Paul Graham Date: 2023-11-01 (saved)
Summary
Paul Graham’s classic essay on why early-stage startups should manually do things that don’t scale — recruit users one at a time, provide exceptional white-glove service, build features for specific users. The lesson: growth comes from doing the un-scalable first.
Key Claims
- Startups fail by trying to scale before finding what works at small scale.
- Manual recruitment: Airbnb founders visited hosts personally, photographed apartments, built relationships. This doesn’t scale but creates the quality foundation.
- The turn: you go from unscalable to scalable after you’ve learned what users actually need.
- Exceptional service is a feature: early users remember you treated them personally. This builds loyalty that scales through word of mouth.
- YC pattern: most successful YC companies did profoundly unscalable things in their early days.
Concepts
- This is a product/startup essay — tangentially relevant to AI product development