Broad vs interest targeting in 2026

go broad and let the creative do the targeting. on 2026 ad platforms, broad audiences with strong creative consistently beat interest stacking, because the delivery systems now read the ad itself to decide who should see it. interest layers mostly restrict the algorithm’s search space and raise your costs. narrow targeting still earns its keep in a few real cases, tiny niches and hard geo constraints, and almost nowhere else.

What changed, and why interest stacking stopped working

the platforms got better at targeting than your interest list ever was. interest stacking made sense in an era when the advertiser knew more about the audience than the algorithm did, so you told Meta who to show the ad to and it obeyed. that era ended quietly over the 2023 to 2025 platform cycles. the delivery systems now predict conversion from thousands of signals you cannot see, and the audience signal in your interest stack is a blunt subset of what they already know. stacking “AI tools” on “productivity” on “early adopters” does not find you a smarter audience. it fences the algorithm into a smaller pen, makes the auction inside that pen more expensive, and excludes the converters who never told the platform they liked those things.

The targeting is now inside the creative

the ad itself decides who sees it, which moves the craft from the audience tab to the creative brief. when delivery systems read the hook, the visuals, the spoken words and the early engagement pattern to work out who responds, the person on screen and the problem named in the first two seconds do more targeting than any interest layer. an ad that opens on a student panicking about an exam finds students. an ad that opens on a founder staring at a dashboard finds founders. this is why creative volume is the real targeting lever: every distinct angle you test is effectively a new audience definition, tested at auction prices, and it is why the testing cadence matters more than anything in the audience settings.

format is part of that signal too. Business of Apps data has 9:16 video converting about 12% better than static, which fits the same logic: full-screen native video carries far more for the delivery system to read, and more for a human to respond to, than a static frame. it is also exactly the format UGC-style creative naturally produces, which is convenient, because that is the cheapest creative to make in volume.

What broad requires from you

broad targeting only works if the creative is strong enough to do the job you just handed it. that is the honest trade. weak generic creative on a broad audience gives the algorithm nothing to read, and delivery drifts toward whoever is cheapest to reach rather than whoever converts. if your ads cannot articulate who they are for in the first two seconds, broad will feel like throwing money into fog, and the fix is not adding interest layers back. it is making ads with a sharper point of view. broad does not lower the bar on creative. it moves the entire bar there.

Where narrow still makes sense

narrow targeting survives where the platform genuinely cannot infer your constraint. a b2b-ish product inside a consumer platform, where the buyer is one job title the creative cannot reliably select for. a product that only works in three cities, where geo targeting is not strategy but fact. a language or regulatory constraint that hard-limits who can use the product at all. and retargeting, which was never interest targeting to begin with, just a list of people who already know you. the test for any restriction is simple: is this a real-world constraint on who can become a user, or is it my guess about who will convert. constraints stay in the settings. guesses belong in the creative, where the auction will grade them honestly.

run broad, put every guess you have into the ads themselves, and read the results weekly. the rest of the system this fits into is the paid ads playbook. if you want your account structure sanity-checked before you flip to broad, the breakdown call is free. Let’s talk Growth.