Sensitive interest targeting is gone on Meta

On Wednesday, January 19, 2022, the detailed targeting box in Ads Manager got smaller, and if you advertise a practice, a clinic or a treatment, it got a lot smaller.

Meta announced the change on November 9, 2021, ten weeks before it took effect. Detailed targeting options that touch sensitive subjects are gone: health causes and conditions, sexual orientation, religious practice, political affiliation, race and ethnicity. The stated reason is that these options let advertisers reach people based on things those people never volunteered.

If you sell implants, clear aligners, weight management, dermatology, physical therapy or anything adjacent, the audiences you have been running for years are either gone already or will stop when Meta’s grace period for running ad sets ends.

What actually went away

Worth being precise, because the panic in the forums is wider than the change.

• Gone. Interests, behaviors and demographics that imply a health condition, a treatment someone might need, a cause they support, their religion, their politics, their sexual orientation or their ethnicity.
• Still there. Location and radius, age, gender, language, and the broad interests that do not imply anything sensitive.
• Still there. Custom audiences from your customer list, website audiences from the pixel, engagement audiences, and lookalikes built from any of those.
• Still there. Detailed targeting expansion, which lets delivery reach past whatever you picked.

Ad sets already running on removed options get a short grace window before they have to be edited. Do not let a stalled ad set on a Friday afternoon be how you find out. In the accounts we run we are rebuilding audiences in the week after the change rather than waiting for the grace window to close.

Saved audiences are the quiet trap. If a saved audience contains a removed option it does not keep working because nobody opened it, and duplicating last year’s ad set carries the problem forward into the new one.

The honest version for a local advertiser

Here is the part that gets lost in the outrage: for a single-location business, interest targeting was never doing most of the work.

A practice serving the drive time around one office has a finite pool of adults to reach. Layering “interested in cosmetic dentistry” on top of that does two things. It shrinks an already small pool until frequency climbs and costs follow, and it tells you a story about precision that the delivery system stopped believing years ago.

The accounts that lose the least from this change are the ones that were already broad. Tight geography, wide audience, creative doing the qualifying. The accounts feeling it hardest are the ones stacking four interests inside a small radius and wondering why the same people keep seeing the ad.

What we are running instead

In order of how much they matter.

1. Geography and age, then nothing. A radius that matches where people actually drive from, an age range wide enough to learn from, detailed targeting left empty. Then read cost per booked consult rather than cost per click.
2. Your own data. A customer list upload is now the most valuable audience you own and the most reliable seed for a lookalike. Match rates are lower than they were before the tracking prompt, so upload every field you are allowed to: name, email, phone, city, ZIP code.
3. Website and engagement audiences. People who reached the booking page and did not book. People who watched most of a video. People who opened a lead form and abandoned it. With the browser pixel seeing less than it used to, the Conversions API is worth setting up properly rather than as an afterthought.
4. Creative as the targeting layer. If the first three seconds say “invisible braces” or “knee pain”, the wrong people scroll past at no cost and the right people stop. This is the lever most local advertisers have never really pulled, and it is the one that replaces interests.
5. The follow-up. A lead from a broad audience is colder on average than a lead from an interest stack. That is fine if somebody calls within minutes and a text goes out when the call is not answered. It is not fine if the form lands in an inbox nobody opens until Tuesday.

The policy rules that did not relax

Removing targeting options changes nothing about the ad rules that already applied to health and aesthetics. The ones we see cause rejections are the same as ever.

• No implying you know something personal about the reader. “Struggling with missing teeth?” reads as a personal attribute to a reviewer. “Implants placed in one visit” does not.
• No before and after images, and no tight crops that do the same job. This is still the most common rejection in aesthetics accounts.
• The landing page counts, not just the ad. Claims you would not put in the headline should not be sitting under it either.
• Keep the offer literal. A free consult is a consult, and saying so converts better than pretending otherwise.

If your account carries a rejection history, the targeting change is a good excuse to rewrite the creative rather than appeal it again.

What to do this month

• Open Ads Manager, filter for ad sets flagged over removed targeting, and rebuild them in the next week or two rather than waiting for the March cutoff.
• Delete or rebuild every saved audience that contains a sensitive option, so nobody duplicates one later.
• Upload a current customer list and build one lookalike from it. Put a monthly refresh on the calendar.
• Check that the Conversions API is sending events and not just the browser pixel.
• Write two new creative concepts that name the problem in the first line, and let a broad audience find them.
• Fill in your own lead form on a Tuesday afternoon and time how long the call back takes.

None of this is a downgrade if the offer and the follow-up are strong. The businesses that struggle with broad targeting are usually the ones where the ad was carrying work the rest of the funnel should have been doing. That is where most of our Meta ads work goes, and it is why med spa accounts and wellness clinics tend to see the difference in booked appointments before they see it in cost per lead.