Ad Split Testing (3)
When testing Facebook multiple creative angles, are people still splitting up into 1 creative per ad set?
I'm having trouble with Facebook quickly throttling (within a day) half or more of my ads in the same ad set. I'm worried that this skews the data and gives me inaccurate results. I've had this issue ages ago, but was hoping that Facebook would fix this by now.
Originally Posted by flyingcake
The ad sets is the level where you define almost everything: audiences, budget, targeting, etc. So if your goal is to test different things like audiences, then it's better to split, otherwise it will give you inaccurate results. But if the idea is to test only different creatives (like different messages, images, and so on), then everything being in 1 ad set is fine
It all depends on your goals, and what exactly you would like to test.
Cheers!
In order for FB pixel to work and optimize you need to have at least 15-25 conversions before it starts to work as designed; ideally you want more (we aim for 100 conversions per week before FB pixel starts doing its magic)
Let's say you have 3 angles.
You should initially launch with 3 adsets, and one ad in each duped 2-3x. (The exact same ad duplicated within the adset)
You then run this for 48 hours - 72 hours, and then pick the winner angle/ad.
Now what you do is, you take this ad and you make copies where you change the image but leave the text (again in separate adsets, the ad duped 2-3x)
If you put too many ads into 1 adset fb will (as you've probably seen) give impressions to the first 3-4, and ignore the rest.
Home > Paid Traffic Sources > Facebook & Instagram