You need to work on your it sucks radar. In my experience all software sucks (even what I've written). The user is typically looking for a piece of software whe

You need to work on your it sucks radar. In my experience all software sucks (even what I've written). The user is typically looking for a piece of software where something works, anything. So everything that you see can be improved. What you want to find our improvements which change the market? Sometimes you can see this before you start working on the problem. I can only see the possibilities when I consider the problem deeply. So my suggestion is to find a problem that you have, and figure out what is the least that you have to do to solve it. (You really want this to be 1 weeks work). Then develop the solution and figure out how it sucks, and it does, all software sucks. Get feedback from your friends and make improvements.