[dshah]

Thanks for the compliment. I've struggled with articulating my position on this issue, and still haven't nailed it yet. Here's another pass:

I completely get the fact that projects are not a bad thing. Most of the best businesses actually started out as projects.

It makes total sense for me that startups should pursue building something people want (and make that the focus). However, a subset of these are projects that seek to build something people want AND that they will pay for. I'd further argue that within this particular subset of ideas, the optimal choice might be to build something people will pay for from the get go. In these cases, it would be sub-optimal to simply strive to build something people want.

I guess what I'm trying to say is that in some situations, trying to find the problem that people want solved AND will pay for ends up putting the startup on the right idea wheresa if they simply tried to build something people want, they may wind up in the wrong place.

Not all software is best served by creating something that will be subsidized by another activity (like advertising). I think it's jumping too quickly to that assumption that I think has the chance of being misguided. Sometimes the right answer is building something that people will part with money for and you have to start with that end in mind to really get there.