Chapter 6: Let The Data Do The Talking
The Ads Manager Breakdown Menu is one of the best tools for gaining a clearer picture of what your core demographics look like, and where you can start to optimize your targeting. We call this process “trimming the fat”, or the process of manually excluding or refining your targeting options to maximize conversion potential based on demographic data. In the breakdown menu, you can filter your top performing campaigns by conversion based on age, gender, device used, placement, location, and more, all of which are directly targetable options when creating new ads.
Let’s say for example, you spend $100 and receive 100 conversions. When looking at your conversion breakdown by age, you notice 80/100 conversions came from men age 18-35, and only $50 of ad spend went to that demographic. Conversely, 20/100 conversions were from women age 35+, and also cost $50 of ad spend. While both demographics received conversions, the first demographic (Men, 18-35) had a much lower cost per conversion ($0.62) than the second demo (Women, 35+, $2.50/conversion). This data would be an indication that on our new ad sets, we should set the targeting suggestions to exclusively men, aged 18-35.
Intuitively it might not make sense to cut the female demographic out entirely, since they are indeed converting on the ads. However, if we ran the same ad targeting only men 18-35, with the average cost per conversion of $0.62, the same $100 spend would yield 160 conversions instead of the original 100. This is a simplistic view of the concept of “trimming the fat”.
The Meta ads algorithm is great in a lot of ways, but it is constantly trying to expand your audience and gain more data. This means that it intrinsically must lose money on some demographics in order to learn what works and what doesn’t. If you have existing data that supports throttling in the algorithm by making targeting restrictions, like the above example, then you can minimize the amount of ad spend that automatically goes into targeting less profitable customers.
This concept can be applied to targeting options like age, gender, location, device used, placements, and individual interest groups, but only when enough data has been previously obtained to justify such a restriction. When we refer to “Ad Optimization”, this is typically the process we are referring to.