How to Evaluate Startup Ideas

Hale says a startup is designed to grow quickly, so an idea should explain how the company could achieve that growth. YC considers companies at the idea stage, including Zenefits, Reddit, and Wufoo. He frames an idea as a hypothesis that an investor can assess.

Hale breaks an idea into a problem, a solution, and an insight about why the solution can succeed. Strong problems are popular, growing, urgent, expensive, mandatory, or frequent; an idea need not have every trait. He also emphasizes triggers that bring people back to a product, drawing on B. J. Fogg’s model of motivation, ability, and triggers.

Hale warns against starting with a technology and searching for a problem to fit it, which he calls a solution in search of a problem. He describes five possible advantages: founder expertise, a growing market, a product that is 10 times better, low-cost acquisition such as word of mouth, and a competitive position strengthened by network effects. A startup should identify at least one advantage connected to its ability to grow.

Hale distinguishes threshold beliefs, such as whether a team can build the product, from the more important “miracle” belief about what could make the company take off. He illustrates the framework with YC, whose advantages included its founders’ expertise, an open application, Paul Graham’s audience, and an alumni network that grew in value. Wufoo combined a broad need for web forms with a visual editor, a developer audience, and forms that users could embed on their sites.

Hale asks founders to map their problem, solution, and advantage, then identify gaps in the story. The next step is to decide what evidence to gather and what assumptions to test by talking to users.