I had a great — great! I tell you – 30 second conversation with Ken Varnum (of RSS4Lib fame) that went something like this (much paraphrasing, obviously):

B: You're gonna have to fix that interface. The standard header won't work.
K: Well, no, we're going leave it as it is.
B: It's not gonna work.
K: We've decided to make it all consistent.
B: OK, you can keep saying that, but I'm really, really smart and I say users are going to be confused.
K: We've done user testing. They weren't confused. And here's our plan to see if they are confused once we go live.

And then I finally shut the hell up. While I’m never crazy about being just plain wrong, it was so so SO refreshing to have someone say, “Well, actually, we’re making this decision based on data and not just pulling answers out of our pants like so many flying monkeys.”

Where, oh where in the library is the dedication to making actual data-based decisions? Besides Ken’s office, I mean?