[mindcrime]

Yeah, I think you have a point. I'm not sure existing data is enough to generate ideas about all potential new products though, since behavior is shaped by what things/ideas/knowledge we have access to today. That is, what data is going to tell you that somebody wants a product that they can't even imagine yet?

So I'd think consumer behavior data / demographic data / etc. is a useful guide, but I'm not sure it's sufficient in and of itself.

All of that said, Alan Kay said something interesting in a set of lectures[1][2] he did a year or so ago. He brought up the idea of basing new product ideas on "Human Univerals", or universal aspects of human nature that really don't vary. A writer named Donald Brown wrote a book literally titled Human Univerals[3] which catalogs a number of these (a lot of them are pretty obvious though: food, shelter, sex, etc.). Kay suggested reading this book, pick one of the human univerals, and figure out a way to use technology to address that.

[1]: https://www.youtube.com/watch?v=id1WShzzMCQ

[2]: https://www.youtube.com/watch?v=1e8VZlPBx_0

[3]: https://www.amazon.com/Human-Universals-Donald-Brown/dp/0070...