Featuritis vs. the Happy User

With the primary sphere of software innovation now being online offerings, it seems like we have an awesome opportunity to do what leaders of focus groups have dreamed about since the term focus group was invented -- observe usage during natural activities.

Would it be an invasion of my users' privacy to record the frequency of use of various features, if that info was not linked to the user in any way? If that was acceptable to users, we'd have the opportunity to expand on the areas that were getting the most use, while (all the while battling the marketing dept -- "It's already there, what's the harm???") dropping the features that were rarely used, or (even more telling!) got initiated but not completed.

That could lead to some radical simplifying of applications, all to the advantage of the user. But I wouldn't want to be compared to some Orwell character. Whadda people think -- reasonable and beneficial research, or nightmarish Big Brother tactic?

Admittedly I'm guilty of this...adding too many features to any app be it web or desktop. Just goes to show you that the KISS principle is still alive and kicking.