[seanos]

I agree that a PhD is not the only way to get domain knowledge. Working in industry and private research into a problem are other ways (amongst many). No matter how obtained, domain knowledge can be a great advantage for an entrepreneur.

Scott Adams attributes his success as a cartoonist to his rare combination of ability to both draw and write jokes well (http://dilbertblog.typepad.com/the_dilbert_blog/2007/07/care...). For a software entrepreneur, programming is analogous to drawing (as the base skill) and domain knowledge analogous to writing jokes (as the differentiator).

It’s also worth noting that technical PhD research is often based on a real problem; gives exposure to a broad range of hard problems and solutions; and teaches the process by which problems can be solved. Thus it can both help you find and solve problems (particularly those in the same field), whether those problems were the focus of your initial scholarly research or not.