When it comes to starting a company, the best ideas don’t come from brainstorming sessions but from noticing real problems. This insight, shared by Y Combinator co-founder Paul Graham, highlights a powerful truth: great startups begin by solving problems that matter—often problems the founders themselves experience. Below, we explore Graham’s approach to identifying promising startup ideas, adding practical insights to help you apply this strategy.
Why Founders Should Solve Problems They Know
Great startup ideas tend to share three characteristics:
- They solve a problem the founders themselves face.
- The founders have the skills to build the solution.
- Few others recognize the idea’s potential.
Take Microsoft, Apple, or Facebook—each started as a small solution to a specific need, growing over time into transformative companies.
The Danger of “Made-Up” Ideas
Many failed startups fall into the trap of “made-up” ideas—concepts that sound plausible but solve no real problem. For example, a social network for pet owners might seem like a good idea: millions of pet owners, plenty of potential engagement. But when users don’t urgently need the solution, the idea flops.
The takeaway? Avoid ideas that sound good on paper but don’t solve a pressing problem.
Focus on “Wells,” Not Ponds
When evaluating ideas, Graham recommends looking for what he calls “well-shaped demand.” Instead of broad, shallow markets where many users have mild interest, target narrow, deep niches where a small group of users desperately need the product.
- Example: Facebook started by serving Harvard students—a small group with a strong need for campus-wide connection. Similarly, Microsoft began with Altair Basic, targeting a niche of hobbyists programming in machine code.
Key Question: Who needs this so badly they’ll use a scrappy, unfinished version from an unknown startup?
How to Develop a “Prepared Mind”
Graham argues that the best startup ideas are noticed, not invented. To notice opportunities, founders must live at the forefront of fast-changing fields.
- Immerse Yourself in the Future: Whether through programming, biotech, or another field, becoming deeply engaged helps you spot gaps others overlook.
- Pay Attention to Annoyances: Problems that frustrate you today might be solvable tomorrow.
Example: Drew Houston created Dropbox after repeatedly forgetting his USB drive. His personal frustration turned into a billion-dollar solution.
Practical Steps to Find Startup Ideas
- Solve Your Own Problems: Ask yourself, “What’s missing in my work or life that I’d pay to fix?”
- Explore Other Domains: Combining expertise from unrelated fields often leads to breakthroughs. For instance, computer scientists learning genetics might spot unique opportunities in biotech.
- Work on Interesting Projects: Side projects that seem like “toys” often turn into significant ideas. Early microcomputers, social networks, and search engines were all dismissed as trivial initially.
- Turn Off Filters: Ignore whether an idea seems unsexy or involves tedious tasks. Stripe succeeded because its founders tackled payment processing—an unglamorous but lucrative challenge.
How to Avoid Common Pitfalls
- Don’t Dismiss Small Markets: Niche products can grow into massive opportunities. Airbnb began as floor rentals during conventions but evolved into a global phenomenon.
- Ignore Competitors (Mostly): Focus on users, not existing players. A crowded market often means demand exists, and incumbents are leaving gaps.
- Beware the Schlep Filter: Many great ideas involve unpleasant work. Overcoming this aversion can lead to valuable opportunities.
Conclusion: Live in the Future, Build What’s Missing
The best startup ideas emerge organically from a “prepared mind” exposed to the right problems. Instead of forcing yourself to think of ideas, focus on becoming the kind of person who notices them. Work on interesting projects, dive into fast-changing fields, and pay attention to what frustrates you or others.
Ask yourself: What’s missing in my world, and how can I build it?
For a deeper dive, explore the original essay by Paul Graham here.