David Karp on how his team keeps Tumblr lean:

After we launch a new feature, I keep a close eye on how many people are using it. If it’s unpopular, we’ll discontinue it and try something else. Every feature has some maintenance cost, and having fewer features lets us focus on the ones we care about and make sure they work very well. For every new feature we add, we take an old one out. A lot of big sites don’t do that, and it’s a problem. Twitter started as a beautifully simple product, but it’s now going the same route as Facebook. The drive to innovate can overencumber and destroy a product. My goal is to keep Tumblr very focused.