This strikes me as a market research tool - trying to find out what people want, the needs, the problem to be solved, the pain to address. It's tricky for users

This strikes me as a market research tool - trying to find out what people want, the needs, the problem to be solved, the pain to address. It's tricky for users to make the connection between problem and solution (hell, it's tricky for anyone - so why not users?) Here's my suggestion: target developing knock-offs (copies of existing products). There's a long and glorious tradition of this in open source (linux, gimp, open office, samba, mySQL, etc). It works because users already know they need it, and how to use it, it's already integrated into their behaviour and their other processes and other software - and programmers have a clear specification to work from. This narrowing the kind of project you seek to fund (i.e. a niche) will help you in many ways, as any business book will recommend (even though narrowing your options seems like a limiting thing...) But please don't use this to undermine good-guy programmers, who took the risk of working to understand a problem, design a solution and get it accepted - use your bounty powers for good, not evil. That is, use it to break monopolies that are hurting customers in small industries. If they have any sense, the monopolist will back down, and some good will be done (whether the open source product is actually used or not). Sounds like good, heroic fun!