A few weeks ago, we tried a little experiment. We tested two different headlines for one of our blog posts. The first used a headline that teased the post’s content: The second used a similar headline, but we made it clear that the post was backed by data: The results were more than a bit surprising: Now, I admit, we expected the variant that mentioned research to win. But we didn’t expect the delta to be more than 40%! For us, that small test shaped a new way of thinking about what makes