I found that, FOR ME, the most commonly requested feature only affects my lowest common denominator of users.<p>It seems that the most requested feature on my primary app is a very "social" feature, meaning only the most "social" people were requesting it.<p>A competitor came out that STARTED with the social feature and it appears that very few people use it.<p>Not that the feature is off the table, but I didn't let users determine what goes to the top of the list. I've learned which features create more customers, and I'm sticking with that for now.
What I learned in building my app was how many different ways people used my app. I found that once I asked why they wanted certain features and learned a bit about their usage behavior, I could more clearly ascertain whether this was a smart idea or an aberration.
An interesting counter-example is Goodreads.com. The number of features on the site is amazing, and consistently growing.<p>At a first glance it seems to be a fairly basic site for book reviews, but the deeper you dig you find a full fledged used-book exchange, quizzes, lists, comparison shopping, profiles, trivia, quotes, videos, short stories, etc.<p>I'm not sure how they manage to keep it all under control or how they decide which features will be useful, but they do a great job with keeping users engaged in multiple activities on the site, no feature seems neglected or empty.