I acknowledge that a lot of bad decisions based on asking the people are a result of simply taking the answers literally, ie. not asking the five whys. Taking an example from Tommi’s post on S60 Address Book, someone asked to be able to undo edits in the address book. Well, a nice idea! But the only reason was because the user was deleting contacts accidentally — so the better approach would be to fix the problem of accidental deletion rather than do what is being asked, and add yet another feature, the undo. I guess this is the kind of examining you are thinking of?

As said, I do believe (without any grounds ;) that by watching the users, even more radical improvements to the user experience are possible. Improvements on stuff that isn’t perceived by the users. Stuff that has become status quo, things that “are the way they are because they’ve always been that way”. Improvements that take features to the next level rather than tinker with the old way of doing things.

It sounds like I’m talking about radical R&D and experimental prototypes (which I do love to talk about in a different context ;) but really I’m not. I believe in small changes which — if found — can radically change the user experience. And I don’t believe those can be found by asking.

I could be wrong. Would love to be wrong, actually. :)