Do Things That Don't Scale
Manually recruit users and create extraordinary experiences that seem inefficient but drive early growth.
Explanation
Paul Graham's counterintuitive advice: successful startups require founders to personally engage in activities that seem inefficient or unscalable in the short term. These manual efforts help you understand users deeply, build the right product, and create a foundation for future automated growth.
Real-World Example
Stripe founders would physically install their payment system on users' laptops during meetings rather than waiting for them to try it later ('Collison installation'). Pebble hand-assembled their first watches before scaling to millions on Kickstarter. Reddit founders created fake user accounts to populate content and make the site feel alive.
How to Apply
Recruit first users personally—don't wait for organic growth. Create an impossibly good experience for early customers, even if manual. Focus on narrow markets initially until you dominate them. Avoid big launches; they're often lazy shortcuts. Do manual operations to learn what to automate later. Remember: 100 users growing 10% weekly becomes 14,000 in a year.