I am a big fan of Steve Blank's book. However, often in the past I would analyze, research, analyze and then ... nothing. I figured this time I would just get something up there - and then figure out what I'm building along the way.
I've found that customers are really crappy at telling you what they want, and product suggestions (and feature requests) should be treated with extreme skepticism.
However: they are absolutely brilliant at telling you what their problems are. Behind every feature request/product suggestion is a pain point of some kind; if you can find it, you usually can come up with a solution that is much better than what they would ever propose.
So: find a group of people who have a common problem. Once you do that, you're on your way.