Anyone else feel like ad targeting has become a complete black box lately?
Feels like half the job now is just feeding Meta more budget and hoping the algo has a good week.
You upload conversions, broaden audiences, trust the machine… and when performance tanks nobody really knows why.
We’ve been testing a different approach recently using external data from Outra, and honestly it’s been refreshing just being able to understand who we’re targeting again.
Instead of purely relying on platform behaviour, we’re building audiences around real-world signals like:
- affluence
- life stage
- family setup
- moving intent
It’s made campaigns feel way less random and a lot easier to explain internally and allowing us to actually align creative which has been a lever that has worked with the right people.
Not saying platform algos don’t work, but I do think we’ve maybe over-corrected into “give the machine everything and ask no questions.”
Curious if anyone else is moving back toward more deterministic targeting/data layering vs pure black-box optimisation?