Facebook's ad system concentrates budget on just 1-4 creatives out of 10-15. To test efficiently, launch many ads per set, then duplicate the set and remove underperformers before relaunching with fresh creatives.
that Facebook will grab, as usually Facebook does, will grab 2 to 3, sometimes 4, sometimes 1 of those ads and just run with them. And out of 10 or 15 ads that I'm launching on each ad set, maybe again, 2, 3, 4 of those will get the bulk of the budget. But here's the trick is that now you have an extra 10 ad sets, 12 ad sets, 15 ad sets that you can just duplicate and relaunch in case they didn't get any budget. Okay, so that's kind of an easy way of launching new creatives. Sometimes you can do that. So if you launch an ad set with 20 ads in it and only 5 got any sort of meaningful spend and they still didn't perform, don't be— no big deal. You just stop that ad set, duplicate the ad set, get rid of those 5 ads that were kind of proven or tested already, and you can repeat the process
