Even the most graduate-school-inspired businesses are about 5% pre-existing research and 95% solving a problem people will pay for. Academic research is, almost

Even the most graduate-school-inspired businesses are about 5% pre-existing research and 95% solving a problem people will pay for. Academic research is, almost by definition, solving problems that people will not yet pay for (I say this having been a graduate student and professor, and now in the startup game). I'd encourage you to think about what marketable problems you are well-equipped to solve, rather than thinking about what you can do to sell your pre-existing research. The former is the essence of a startup and the latter is, on average, a recipe for disappointment. On the legal matters: the University will almost certainly license the idea to you. Universities are not in the business of business, so just holding the patent is not particularly useful to them. They want someone to go out and start a business around the idea. You would be a good candidate for that. There is essentially no chance that they won't license it to you if you are credibly starting a business around it (though they'll probably want some kind of stake in the company). This is exactly the situation Google has: Stanford owns the PageRank patent and licenses it to Google in return for (I believe) a portion of Google's early stock.