24/09/2021
I'm testing a new concept called "audience splitting" and I'm liking the results so far. It is a simple concept that I actually learned from my systematic stock trading mentor. Here is the idea. Once you have one audience that works you then come up with an idea for refinement of that control audience to get improved conversions. Here is how it works. You take your base control audience and narrow it down by a new interest AND you also create a second audience that excludes that interest. Here is my first time trying this, but if this works it should tell you if that interest is significant or not. I'm running this test using a Traffic campaign objective as opposed to using the Facebook AI using conversions. This test just started, but I'll share updated results in about a week once we have more traffic.