The "Interest Targeting Doesn't Work" Scam

Here is the claim you have heard a hundred times. Interest targeting is broken, Meta's algorithm is smart enough now, so turn everything off and let the machine find your buyer. It gets repeated because it is partly true, and that is exactly what makes it dangerous. The people selling it rarely tell you the half where it falls apart.

Where the "it doesn't work" crowd is actually right

They are describing a real thing. On accounts with meaningful daily spend and a well fed pixel, layering ten interests on top of Meta's own modelling usually does nothing except shrink your pool and raise your CPMs. Broad genuinely outperforms in that situation. That part is honest.

Broad targeting tends to win when:

• Your pixel has strong purchase data, usually 50 or more conversions per week
• Your creative is high quality and clearly signals who the product is for
• Your budget is large enough to exit the learning phase quickly
• Your product has wide appeal rather than a very narrow niche

The scam is the part they leave out

The trick is the leap from "broad beats interests on big, mature accounts" to "interest targeting does not work, full stop." That second sentence is the one that gets sold, and it is the one that torches new accounts, tiny niches, and anyone whose pixel has seen 40 purchases instead of 4,000.

Why blanket "go broad" advice backfires

Meta's algorithm optimises toward the signal it is given. When it has thousands of conversions to learn from, broad is just handing a very good student a very big library. When it has almost none, broad is handing that same student an empty room and asking them to guess.

Early on, a sharp interest is not a cage. It is the first real hint the system gets about who is actually worth chasing. Broad is not a strategy you start with. It is a reward you unlock once you have given the pixel something to be smart about.

Warm audiences are not proof that broad works

The comment section on the original reel nailed the other quiet truth. Customer lists and retargeting audiences convert not because the targeting is clever, but because those people are already yours. Attributing that win to broad prospecting is how a warm audience result gets laundered into a cold traffic promise. Judge prospecting on prospecting numbers only.

How to actually decide: interests or broad

Skip the ideology. The question was never "interests or broad." It is "which one matches the data I have right now." Run it like this:

1. Check your conversion volume first. Under roughly 50 pixel conversions a week, you do not have enough signal for broad to be smart. Give it a strong interest to anchor on.
2. Match the audience to the budget. A tiny daily budget spread across a broad open audience learns slowly and expensively. Concentrate it while you are still small.
3. Use interests as a testing scalpel. Even on broad friendly accounts, a sharp interest is the fastest way to validate a new angle, hook, or product before you scale it wide.
4. Graduate to broad on evidence, not vibes. Once the pixel is fed and a concept is proven, open it up and let broad do what it is genuinely good at, which is finding more of a buyer you have already defined.
5. Never confuse warm with cold. Customer lists and retargeting will always look like heroes. Keep them in their own bucket when you read results.

If you want the deeper breakdown, we wrote a full comparison of broad vs. interest targeting on Meta ads, and a practical guide on how to lower your Meta ads customer acquisition cost once your targeting is sorted.

The takeaway

None of this is exciting enough to go viral, which is exactly why the "targeting is dead" version keeps winning the algorithm. Nuance does not sell courses. Results do. Pick the approach that fits the data you have today, and change it when the data changes, not when a guru tells you to.