Io be honest, i'm not much of an expert on this either. I've built several projects and have abandoned midway. I put down all of my ideas with a title and 2 li

Io be honest, i'm not much of an expert on this either. I've built several projects and have abandoned midway. I put down all of my ideas with a title and 2 lines of description on evernote whenever it strikes (mostly when seeing a problem, there could be other solutions that exists for this problem at this point). And whenever in an unrelated instance I encounter a variation of the problem again, which if I believe could've been solved by that idea that I had, I give it a serious thought into researching for solutions and if none exists, I get to build. I wouldn't call this an ideal process, it just works for me. The bad side-effect I have come to realise from this process is, I build products for which there doesn't exist a solution, which means either the market isn't there or isn't big enough. The previous project was the same situation (fileql.com, ended up selling it) and now RosterBird. They stay small for relatively long time as you also have to convince users that they have a problem in the first place.