I think PG's larger idea here is simply: "build what you know." And that is excellent advice. Either build something that scratches your own itch, or if you're

I think PG's larger idea here is simply: "build what you know." And that is excellent advice. Either build something that scratches your own itch, or if you're trying to solve somebody else's problem, understand that problem like it's your own. My first funded start-up was an online English instruction website geared at students in East Asia. We had great software, a great idea, and the market was and still is huge, but I didn't learn English as a second language, and neither did my co-founder, and ultimately we just didn't understand the problem well enough from the user's point of view.