Kills good design, anyway. I was reminded by the Bad Usability calendar (something you should check out, by the way) of a maxim I read from Jakob Nielsen that said "customization is not a silver bullet." He touched on a similar subject in an article on personalization vs. customization. I don't agree with everything he says, but he's spot on here, and this is why.
It's lazy
Customization is the lazy way out. Even here at ZURB we occasionally come to an impass on whether a feature should be implemented, or an interaction should work one way or another and we're tempted to just say 'well, we'll make it a preference.' Don't give in to the temptation to push your design problems onto the user – figure out what's best and go with that.
Users don't know
Toss in a Julian or Gregorian toggle and we're good to go!
Most users are not designers. There's a reason we get paid to make these kinds of decisions, and it's because they don't always have easy or obvious answers. The Bad Usability calendar uses a case where a user is asked to pick the date display format from a select list of...well they only show seven options but it's easy to imagine showing another 10, or 20, or who knows how many. One of those options is, in all likelihood, the most readable and intuitive for the largest possible group of people. Pick that. Figure out which one that is through user testing or research, and pick it for them. If one is a convention in America and another in Europe, design accordingly.
That's a fairly simple example, but the principle is sound. Customization is all well and good, and in fact in some cases it makes for a great experience (or at least a compelling one). But it's not a silver bullet, a catch-all for design problems – it's a tool we can use well or poorly.
Have any examples of customization (or personalization, to avoid a semantic dispute with Nielsen) we should see? Good or bad. Put them in the comments so we can check them out!