Our in-house A/B testing tool allows us to optimise components on the site and validate new features. In collaboration with the newsroom, we took this one step further and began testing the actual content of the site, starting with headlines on FT.com. After some initial curiosity tinged with a hint of apprehension from the editors, a few key results and some interesting surprises have led to the tool being embraced in the newsroom and used multiple times a day.
Accompanied by lots of examples and some (voluntary!) audience participation, I’ll discuss the differences between testing content versus features, and the technical changes we made in order to facilitate this new type of testing.
Required audience experience
Objective of the talk
This session will provide an understanding of: the type of A/B testing we do at the FT and how we do it; the technical challenges of real-time headline A/B testing; the full workflow from headline writing to implementation of the winning variant; and the future direction of the headline project.