A common media buying thought process in which I disagree with is to not optimize your campaigns for:
-Placement
-Age
-Time of day

While FB's machine learning mechanisms may work over time, with this strategy you can maximize ROAS AND spend more money simultaneously.
An example which illustrates this is as follows:

We ( @snowagencyco) recently onboarded a client and noticed that after the first 7 days:
-95% of the budget was going to FB with a 1.56 ROAS
-5% of the budget was going to IG with a 2.2 ROAS
In order to test if the IG placement would have substantially different results (with the same content) we:
-duplicated these ad sets to serve ONLY to IG.

The result: 4-8x ROAS on all ad sets duplicated ($50-200 budget per adset)
On the same ad account we also noticed an interesting thing:

The time of day between 1am-9am was under a .5x ROAS

In order to test this and give more ad spend to the rest of the times of day we duplicated the top performing audiences to exclude these times.

RESULT: 6-8x ROAS
Now I already know what some of you are about to say:

"BUT WHAT ABOUT THE CPM INCREASES?!?"

-Something SUPER interesting happened to these segmented ad sets:

THE CPMS WERE 40-50% CHEAPER THAN THE CONTROLS!!!! 🤯🤯🤯🤯🤯🤯🤯🤯🤯
You can follow @iamdansnow.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: