How to evaluate a startup idea: problem, solution, and unfair advantage

Executive overview

Most founders either lack conviction in their idea or can't articulate why it will grow quickly. Investors don't look for what's wrong — they look for how a company could possibly win.

A startup idea is a hypothesis about why a company can grow quickly. It has three parts: the problem (initial conditions), the solution (the experiment), and the insight (the unfair advantage that explains why you'll win).

Without a clear unfair advantage, an investor has no reason to choose you over anyone else.

What makes a problem worth solving

• Aim for at least one of: popular, growing, urgent, expensive, mandatory, or frequent
• Frequency matters most — it creates repeated opportunities to convert users (motivation + ability + trigger must align simultaneously)
• Ideal signals: millions of people affected, market growing 20%+ per year, people need to solve it right now, costs billions in aggregate, driven by regulatory change, used multiple times daily

Solution: start with the problem, not the technology

• SISP (solution in search of a problem) is a common trap — building around a technology first makes growth much harder
• Ask: are you doing this because you care about the technology, or because you're committed to solving a real user problem?
• Use whatever is necessary to solve the problem — don't shoehorn a problem into a preferred solution

The five types of unfair advantage

1. Founder — are you one in ten (or one in a million) people who can solve this? A PhD with a relevant patent qualifies; being an engineer at a big tech company does not
2. Market — growing 20%+ per year means you grow by default, but this alone is the weakest advantage
3. Product — is your product 10x better than the competition? 2–3x is not enough; it must be an obvious order-of-magnitude improvement
4. Acquisition — paid acquisition doesn't count; the best companies grow by word of mouth at zero cost
5. Monopoly — does your company get harder to beat as it grows? Network effects and winner-takes-all dynamics qualify

Threshold beliefs vs miracle beliefs

• A threshold belief is table stakes: can they even build it? For engineering-heavy startups, this is assumed
• A miracle belief is what determines success: for a B2B startup, can the founders sell? Evidence of sales ability matters more than product completeness

YC and Wufoo as worked examples

• YC's unfair advantages: exceptional founders (PG's audience + prior SaaS exit), falling cost of starting tech companies, free content-driven acquisition, and an alumni network that compounds in value
• Wufoo's advantages: 10x faster drag-and-drop form builder, 100x cheaper than hiring a developer, freemium drove viral spread, 100k developer audience pre-launch, embedded forms spread the product organically
• Wufoo raised only $118k total and returned over 30,000% — capital efficiency as proof of the model