AI summary by TwinMind
Stop Innovating (On The Wrong Things)
Y Combinator
11m 58s
Main Concept: Limited Innovation Energy
- Making startup work = miracle
- Limited "innovation juice" per company
- Focus miracle energy on single core miracle: making something people want
- Odds of performing 5 miracles simultaneously much lower
- Don't spread innovation across every problem encountered
Common Anti-Pattern #1: Corporate Structure Innovation
- Wyoming LLC incorporation as voluntary red flag on YC applications
- Smart people think they can improve on Delaware C Corp standard
- OpenAI nonprofit-owns-for-profit structure cited as cautionary example
- Sam Altman suggested this wasn't wisest choice
- Would not recommend this strategy
- Investment docs, vesting, share classes - stick to best practices
- Corporate governance not best place to innovate
Anti-Pattern #2: Contrarian Startup Advice
- "I'll only help customer if I can disprove all startup advice"
- Examples:
- Making small town center of startup ecosystem
- Hydrogen economy bet (founder believing hydrogen will replace gasoline/solar/wind)
- Completely unnecessary risk unrelated to core startup bet
- Putting personal preferences before customer needs
- "I'll only serve customer if I get to locate in New York"
- Business should be selfless, customer-first
Anti-Pattern #3: Technology Stack Innovation
- Asana example: wrote own programming language initially (later backed out)
- Choosing idiosyncratic programming languages because "it's fun"
- Reddit vs Digg comparison:
- Digg made high-risk technology choices
- Reddit focused on core: social news site that works
- Digg couldn't roll back when site didn't work after redesign
Anti-Pattern #4: Business Model/Pricing Innovation
- Not putting customer first in pricing strategy
- Cloud compute example: customers familiar with AWS pricing
- Should charge like AWS for customer comfort/recognition
- Instead: complex "floel credits" system customers can't understand
- Customers want to know cost simply
- Creating unnecessary friction in purchase process
Innovation Advice Source Mismatch
- Pop entrepreneur/Shark Tank advice comes from branding world
- Branding: being different/unique very important
- Technology world: successful companies do things 80% same, 20% different
- Fashion advice doesn't apply to software
- Usability example: innovating on cursor/button functionality unhelpful
Recommendations
- Put customer first, second, and third
- Save innovation experiments for second startup
- Get one successful startup first
- Focus on core: helping customers with important problems
- Use best practices for everything except core product innovation
