[timjahn]

Don't build anything yet.

Go out and talk to your potential customers. If you don't know who those are, figure that out first. Learn about their problems, their pain points, how you could fix things for them.

Then start with the simplest of simple products. Could be a simple Google Form that submits info to a Google Doc spreadsheet. Or maybe it's a barebones HTML one-page site with a single feature of your idea.

Don't build anything complex until money is coming in the door.

We wasted almost a whole summer on matchist (http://matchist.com/talent) building features we thought our customers would want, only to talk to half of them after and realize we should have been talking to them the whole time. Needless to say, we'll never do that again. :)