Nearly all the most interesting start ups I’ve seen around me have done this. They identified something that was frustrating and clunky, then picked away at it until there was a useful alternative and then kept iterating.