> I would want to see evidence to show blind is more than just a collection of anecdotes before calling it "data".

Didn't Bezos say something like if anecdotes and data disagrees, its the anecdotes that are usually right?