I've been considering that myself quite a bit, and I find that its a combination of things: - Something where I can interact directly with the client/consumer.

I've been considering that myself quite a bit, and I find that its a combination of things: - Something where I can interact directly with the client/consumer. I dislike having to program by specs, when I know that they are written by someone who doesn't understand end users. Simple, effective touches are usually missing, and the feedback loop is so long that apps take years to evolve. - Something challenging. I want to stay up at night visualizing possible solutions. That kind of excitement makes the time spent worth every second. And with the opportunity to think and implement a solution that works really well for the problem. - Something that makes a difference. Doesn't have to be something ground-breaking, but it shouldn't be a CRUD front-end either. This is harder to quantify than I thought. I guess it depends on the novelty of the idea, the silken smooth end-user experience (think iPhone- I own a G1 myself, but man does the iPhone provide a great experience), and the quality/succinctness/simplicity of the code.