How to Get and Evaluate Startup Ideas | Startup School

Résumé vidéo

Overview

This talk provides a systematic framework for evaluating and generating startup ideas, drawing on Y Combinator's extensive experience. It begins by outlining common mistakes founders make, such as building solutions without real problems or getting stuck on superficially plausible "tar pit" ideas. The core of the talk presents a 10-question framework to assess any startup idea, with a strong emphasis on founder-market fit. Finally, it offers practical "recipes" for generating new ideas, prioritizing methods that leverage a team's unique expertise or personal experiences over abstract brainstorming.

Timeline Summary

🎯 Introduction and Common Mistakes

• The speaker aims to provide conceptual tools for thinking about startup ideas in a sophisticated way, similar to how YC evaluates them.
• The advice is based on analyzing top YC companies, Paul Graham's essays, and patterns from thousands of rejected applications.
• The most common mistake is building something that doesn't solve a real, acute problem for users, known as a "solution in search of a problem."
• Another critical error is pursuing "tar pit" ideas—common, seemingly easy problems with hidden structural reasons why they are very hard to solve.

❓ The 10-Question Evaluation Framework

• The first and most important question is about founder-market fit: whether the team is uniquely suited to execute on the idea, as exemplified by PlanGrid.
• A good market can be either large now or small but rapidly growing, like Bitcoin was for Coinbase in its early days.
• A strong problem is one where the alternative to your solution is "literally nothing," as was the case for Brex when providing corporate credit cards to startups.
• The presence of existing competitors is often a positive sign, indicating real demand, as seen with Dropbox entering a crowded market.

💡 Counterintuitive Good Ideas and Organic Generation

• Ideas that seem "bad" because they are hard to start, boring, or have competitors are often great opportunities, as they scare away less determined founders.
• Stripe is a prime example of an idea that required overcoming significant initial "schlep" (hard, tedious work) that others avoided.
• The best startup ideas are usually noticed organically through work or life experience, not by sitting down to brainstorm.
• To set yourself up for future organic ideas, become an expert in a valuable field or work at a startup to see problems firsthand.

🧪 Practical Recipes for Generating Ideas

• The best recipe is to start with what your team is uniquely good at, ensuring automatic founder-market fit, as demonstrated by Rezi.
• Look for problems you've personally encountered, especially from an unusual vantage point, like the Vetcove founders who saw their veterinarian father's inefficient supply ordering.
• Systematically research a fertile "idea space," like the AtoB founders did with the trucking industry, by talking to potential customers and existing founders.
• If uncertain after evaluation, the best way to know if an idea is good is to launch it and find out through real-world feedback.

Key Points

• 🔍 Fall in Love with a Problem, Not a Solution:The most common mistake is starting with a cool technology (like AI) and looking for a problem to apply it to, which often leads to superficial ideas. Instead, begin with a high-quality, specific problem that people genuinely care about.
• 🧱 Beware of "Tar Pit" Ideas:These are common, seemingly easy startup ideas (like a social planning app) that have trapped founders for years. They are dangerous because a hidden structural reason makes them much harder to execute than they appear.
• 🎯 Founder-Market Fit is Paramount:The single most important criterion is whether you and your team are the right people to build this company. Your search should be for a good ideafor your specific team, not just a good idea in the abstract.
• 📈 Evaluate Markets Dynamically:A good market isn't just one that's big now; it can also be a small but rapidly growing market, as was the case for cryptocurrency when Coinbase started.
• 🚀 "Bad" Signs Can Be Good:Ideas that seem hard to start, exist in a "boring" space (like payroll software), or have existing competitors are often excellent opportunities because they deter less serious founders, leaving the field open.
• 🌱 The Best Ideas Are Often Organic:Most successful YC companies got their idea organically through work or life experience, not from forced brainstorming sessions. Deliberate idea generation is possible but more prone to producing common, low-quality ideas.
• 🗺️ Hunt in Fertile "Idea Spaces":Instead of fixating on one specific idea, position your team in a broader category of related opportunities (e.g., "software for trucking"). This increases your chances of finding a good problem even if your first concept needs to pivot.

Frequently Asked Questions (FAQs)

1. What is the most common mistake founders make with startup ideas?
   Building a "solution in search of a problem"—developing a product for a need users don't genuinely care about.
2. Is having competitors a bad sign for a startup idea?
   No, counter-intuitively, it's often a good sign. It validates that a real market exists. The opportunity lies in having a new insight or better execution than the incumbents.
3. How can I tell if I have "founder-market fit"?
   Ask if you and your co-founders are the obvious right team to work on this idea, typically due to unique expertise, experience, or passion related to the problem.
4. What is a "tar pit" idea?
   A common, superficially plausible startup idea that seems easy to solve but has hidden structural complexities, causing founders to get stuck for long periods without progress.
5. What's the best way to come up with a startup idea?
   Start by leveraging your team's unique expertise or by identifying problems you've personally encountered from an unusual position that others might miss.
6. What should I do if I'm unsure whether my idea is any good?
   The most reliable way to know is to launch a minimal version and get real user feedback, rather than theorizing indefinitely.

Conclusion

This framework shifts the focus from chasing abstract, "cool" ideas to a more disciplined, evidence-based approach. Success hinges on deeply understanding a real problem, ensuring the founding team is uniquely equipped to solve it, and having the courage to pursue opportunities others overlook because they seem too hard or boring. The process is iterative; even a good starting idea will likely morph based on market feedback.Action Suggestion: Systematically audit your past experiences and expertise to list problems you are uniquely positioned to see, then evaluate the most promising one using the 10-question framework.