Nobody knows how to evaluate startup ideas better than Y Combinator. YC partner Kevin Hale gives an excellent framework for assessing if a startup idea has potential to grow quickly.
Summary
Your startup idea is a hypothesis about why you will grow quickly.
This hypothesis is made of up three things:
Problem: What is the setting for this company that allows it to grow quickly?
Solution: What is the experiment that you're running within those conditions for it to grow quickly?
Insight: What's your explanation why your experiment is going to be successful?
Always start with the problem. Don't be a solution in search of a problem.
You need an unfair advantage. You want at least one of these:
Founders: Are you 1 in 10 people in the world that can solve this problem?
Market: Is your market growing 20% a year? This is one of the weakest advantages to have because it's not unique to you.
Product: Is your product 10x better than the competition?
Acquisition: Paid acquisition is bad. Word of mouth is best.
Monopoly: Do you get stronger as your company grows? Marketplaces and network effects.
Hale goes on to apply this framework to two startups he knows well: Y Combinator and Wufoo.
Example 1: Y Combinator
Problem: Hard for founders to raise money without knowing someone in venture capital.
Solution: Y Combinator invests in companies through an open application.
Insight / Unfair advantage:
Founders: YC founders had outstanding accomplishments and knowledge of the space. First worm. LISP textbook. First SaaS company.
Market: They saw that future billion-dollar companies would be technology / software startups. Software startups need less money than traditional companies.
Product: Covering 3 months of living expenses is enough to attract hackers. They want to do this. No coworking space is needed. Other angels and VCs want to invest in these startups.
Acquisition: Paul Graham wrote books and online essays with a large audience of target users: hackers. Building an audience is a massive advantage.
Monopoly: They didn't realize this at the time, but the YC alumni network became more valuable than partners. Easier to recruit. Scales advice. Better beta.
Example 2: Woofoo
Problem: Websites need to collect custom information, but you need to know how to code or hire a programmer.
Solution: Wufoo is a WYSIWYG online form builder that lets non-technical people create forms and surveys.
Insight / Unfair advantage:
Market: Almost every website needs to collect data.
Product: WYSIWYG + AJAX is 10x faster than direct competition. 100x cheaper/faster than hiring a developer.
Acquisition: Wufoo started with an audience of 100K developers. They built this audience by blogging for a year. Wufoo forms can be embedded on a website and Wufoo brands the confirmation page, meaning users refer Wufoo to their users automatically.
You can watch the talk in full here: