-
dkiesow They could also be changing the headlines based on editorial decision making instead of metrics, in which case I wouldn't call it a/b testing. I guess the question would be... how often does it change back to the original? That would be a useful indicator.