+1 for Matt's comment-- for a lot of businesses, the fundamental problem (or at least the first problem that you need to solve) isn't a technical one. Famous example: Digg. Famous example #2: TechCrunch (or any content play, really).

I know of a guy who is minting money doing long-tail adwords for car buying phrases, collecting contact information of the people who click on the ads, and selling that contact info to car dealerships for $20 per. It's not evil-- the pitch is: "enter your contact info any we'll get you three quotes from local dealers".

Bang, no impressive technology required (to prove the initial theory).