Interest Targeting Isn't Dead — It's Just Not Doing the Heavy Lifting Anymore
Meta has quietly removed hundreds of thousands of granular targeting options over the past couple of years and pushed hard toward broad, AI-driven audiences. The natural conclusion a lot of advertisers draw is that interest targeting is dead. It isn't — it's just no longer doing what people think it's doing.
The old model: interests as a precision instrument
The old playbook treated interest targeting like a scalpel: stack five or six specific interests and you'd get a laser-focused audience of exactly the right people. That approach fights the current delivery system more than it helps it — Meta's algorithm now draws on far more behavioral and conversion signal than any stack of stated interests can represent, and narrowing the audience artificially just restricts how much room it has to work.
What interest targeting actually is now
A starting signal. A way to nudge the algorithm toward a general direction before it takes over using real engagement and conversion data. It's not the thing doing the precision work anymore — it's the on-ramp.
What this means in practice:
Stop testing ten interest-stack combinations — test one or two broad starting signals and let the algorithm's own data do the narrowing.
Give the system creative diversity instead of audience granularity — more genuinely different concepts move performance more than tighter interest stacks do.
Judge interest-based ad sets by how well they feed conversion data back to the system, not by how narrow or clever the interest combination looks.
Save true precision targeting for retargeting and lookalike audiences built from real customer data — that's still a legitimate use of narrow targeting.
Interest targeting still has a job. It's just a smaller, earlier one than most advertisers are still budgeting for.
Comments