Do Things That Don't Scale

Paul Graham challenges a core assumption many founders make: that successful startups must be built on scalable processes from day one. The essay argues that the most common Y Combinator advice is to do things that don't scale—manually recruiting users, obsessively delighting early customers, and performing labor-intensive work that seems inefficient. These unscalable activities aren't just acceptable; they're essential to building momentum and discovering what actually works. Successful companies like Stripe, Facebook, Pebble, and Wufoo all began with founders doing things that couldn't possibly scale.

3 sections · 10 lessons

Course outline

Acquiring and Serving Early Users

  1. Manual Recruitment
  2. Extreme Attention
  3. Narrow Focus
  4. Launch Fallacy
  5. Acquiring and Serving Early Users

Building and Operating Early Products

  1. Startup Fragility
  2. Imperfect Excellence
  3. Manual Operations
  4. Consulting Discovery
  5. Building and Operating Early Products

Strategic Frameworks and Methods

  1. Hardware Bootstrap
  2. Vector Framework
  3. Strategic Frameworks and Methods

Start learning with Wondering