Duplicate Until You’re Happy
January 31, 2018
January 31, 2018
Duplicate at Ad Set Level or Ad Level depending on budget
We’ve talked about duplicating your ad sets before, probably a few times, but we’ve been seeing some great results with it recently. It’s almost like a refresh button that you can click until you’re happy with the CPA.
Duplicating at Ad Set Level
If you have ads that are running and working but your CPA is a bit too high, you ideally want to duplicate without turning the original ad set off. If it’s working, leave it on until you’ve got a better ad set. If your budget doesn’t allow you to duplicate and leave the original ad set on, then you need to make a decision. The duplicate could bring lower CPA’s, but it could also completely fail and bring no results…
We’ve found that duplicating an ad set that’s bringing results but costing a bit to much 3 times, so there’s 4 ad sets running in total, is the best way to bring costs down. You run all 4 ad sets for at least 3 days (ideally 7) and then look for the ad set with the lowest CPA. Turn the other 3 off, then duplicate the ad set 3 times. Keep repeating this until your CPA won’t get any lower. If you simply don’t have the budget to be running 4 ad sets at once while you wait for them to decide which one is going to perform the best, you can duplicate at ad level…
Duplicating at Ad Level
Duplicating at Ad Level hasn’t been as effective as at Ad Set Level for us, but it can still bring costs down. One thing to note is that where possible, it’s important to make sure the duplicates are using the same post ID so that you keep the same engagement across all ads. You basically want to do the same as above:
Duplicate your ad 3 times, so there’s 4 ads running in total within your ad set. You run all 4 ads for at least 3 days (ideally 7) and then look for the ad with the lowest CPA. Turn the other 3 off, then duplicate the ad set 3 times. Just keep repeating this every week for a while until you’re happy. You’ll either have 4 ads running within your ad set that are running equally well, or 1 ad set that is out performing the others that you can leave on and switch the others off.