A client at work wanted a specific feature. It would be cool, but nobody knew really how useful it would be because nobody had asked, “Well, alright, but who would use it? When would they use it, and how?” So we poked around, called some people, and at this point it seems like a full-fledged feature wouldn’t be the best solution, because use cases seem to be restricted to mainly a few specific types.
This was a serendipitous example, because Prof. Jacob (and my mentor) had talked about this very problem: non-designers tend to think in terms of things they had seen before, like scroll bars and widgets. Scroll bars and widgets are great, but they’re not always the best solution for a given interaction; the best way to start is to set aside the interface and think about what the damned thing is supposed to do. And this is a lesson that applies to any other field of design, I should think.