· 7 min read · Jason Youngjin Koh
Why Meta rejects clinic ads
A rejection is not a mystery and it is rarely bad luck. It is a named policy applied to a specific sentence in your ad. Once you can see which sentence, the fix takes an hour instead of a fortnight.
Most clinic ads that get rejected on Meta are rejected for the grammar of one sentence, not for the treatment being advertised. The offer is fine. The audience is fine. The sentence asserts something about the person reading it, and that is the thing the policy is written to stop.
This is our side of the table rather than yours, so it is worth setting out plainly. Everything below was checked against Meta's published Advertising Standards on transparency.meta.com in September 2026. Meta changes wording and enforcement without notice, so treat this as the shape of the system rather than a permanent map.
The policy that catches clinics first
It is called Privacy Violations and Personal Attributes, and it states that ads must not contain content that asserts or implies personal attributes. The list of attributes it protects includes physical or mental health, and medical conditions sit inside that.
Meta publishes its own examples, and they are more useful than any explanation. Not allowed: Do you have diabetes? Not allowed: Depression getting you down? Get help now. Allowed: Depression counselling. The subject matter is identical in all three. What changes is whether the ad tells the reader something about themselves.
Apply that mechanism to clinic creative and the pattern falls out immediately. A headline that opens with Do you hate your smile? asserts a condition of the person seeing it. A headline that reads Smile makeovers in Parramatta describes a service that exists. Same practice, same audience, same landing page, and one of them will be reviewed as an assertion about a stranger's health.
The instinct that produces the first version is a good marketing instinct in every other category. Naming the reader's problem back to them is how direct response has worked for a century. In health advertising on Meta it is the specific thing the classifier is built to catch, and the more precisely you name the problem the more reliably you get caught.
| The line a clinic wants to run | Why it fails | What we run instead |
|---|---|---|
| Do you hate your smile? | Asserts a personal attribute about the reader | Smile makeovers, Parramatta. Consultations this month. |
| Missing teeth holding you back? | Asserts a health condition and implies a consequence | Implant consultations with a named practitioner in Chatswood |
| Struggling with adult acne? | Second person assertion about a medical condition | Skin consultations at our Bondi clinic |
| Finally feel confident again | Implies an existing negative state in the reader | Book a consultation to talk through your options |
| Look ten years younger by Christmas | A promise of a specific outcome within a set timeframe | Consultations available before the end of December |
The right hand column is not softer, it is more specific. It names the suburb, the practitioner and the next step. That tends to be the version that performs better anyway, which is the part clinics find hardest to believe until they see two weeks of data.
Imagery, and the trap of Meta permitting something
The relevant policy here is Health and Wellness. It permits before and after transformation imagery for general cosmetic products, procedures and surgeries where the ad is targeted to adults. It prohibits close ups on a specific body area, giving the example of pinching fat in weight loss advertising. It prohibits statements of inferiority about physical appearance, which it defines as terms, descriptions or questions that are negative and attack an individual's appearance, specific body parts or hygiene. And it restricts sensational language with exaggerated or extreme claims, or promises of specific outcomes within a set timeframe without disclaimers.
In practice this means the imagery argument usually resolves before it starts. The creative that satisfies AHPRA's matched lighting, angle, framing and warning requirements is almost never the creative that would have won on Meta anyway, so we spend the budget on clinic footage, practitioner pieces and plain treatment explanation instead.
Why an account gets restricted, not just an ad
This is the part that costs real money, and it is the part clinics almost never see coming. A rejection is an event. A pattern of rejections is a record.
Meta describes its enforcement as proportional to the severity of the violation, the history of violations on the account, and the risk or harm involved. It states that for most violations, if you continue to post content that goes against its standards after repeated warnings and restrictions, it will disable the account. Restrictions escalate rather than reset, and the ability to create ads is one of the things that gets removed.
So the cost of a rejected ad is not the rejected ad. It is the third rejected ad, made by someone who kept resubmitting slight variations of a line that was never going to pass, on an account that also holds your page, your pixel, your custom audiences and your entire learning history. Rebuilding that from a fresh asset is weeks, and during those weeks you are not advertising at all.
24 hours
Meta states that ad review is typically completed within 24 hours, although it may take longer in some cases. Every resubmitted variation of a non compliant line spends that cycle again, and lands on the same account record. Source: Meta Advertising Standards, transparency.meta.com, accessed September 2026.
What actually gets reviewed
More than the ad. Meta states that its review system examines images, video, text and targeting information, as well as an ad's associated landing page or other destinations. It also states that ads remain subject to review after they are approved and may be rejected later.
Two things follow. First, a clean ad pointing at a page with a testimonial on it, or a price list for a prescription only treatment, can be rejected for the page. Clinics rewrite the headline four times without touching the page, and cannot work out why nothing changes. Second, an ad that ran fine for six weeks can stop, which is not a bug and is not personal.
The appeal path is a request for review in Account Quality. An edited ad is treated as a new ad and goes back through review from the start.
Why we rebuild rather than argue
An appeal is a queue you do not control, with no date you can plan around and no person you can talk to. A rebuild is an hour of work with a known result. When the rejection is a genuine policy hit, and for clinic creative it usually is, the appeal is asking a reviewer to agree that the policy does not mean what it says.
There is a narrow case for appealing. If the ad is plainly inside the policy and the rejection looks like a classifier error, and especially if the same creative has run before, request the review. Do it once, on one ad, and keep advertising with a rebuilt version in the meantime so the account is not sitting idle waiting for an answer.
The question is never whether Meta is right. It is whether you would rather be correct in four days or live in three hours.
What we do before anything goes live
- Strip every second person assertion out of the copy. If a line tells the reader something about their own body or health, it gets rewritten as a description of a service.
- Check the destination as carefully as the ad, because Meta reviews it too. Testimonials about clinical care, prices for prescription only treatments and self select treatment menus all come off first.
- Build outcome imagery to AHPRA's requirements rather than Meta's, and assume the Australian rules are the binding ones.
- Keep a small set of pre cleared, deliberately plain creative in the account so there is always something running while a new concept is in review.
- Log every rejection with the policy named, so the pattern is visible after three of them instead of after thirty.
- Never resubmit a rejected line with the adjectives changed. If it failed on the mechanism, it will fail again, and the account carries the record.
None of this is a workaround. It is the opposite. The policies are stable, published and specific enough to design against, which makes them a constraint rather than a risk. The clinics that struggle on Meta are almost always the ones treating each rejection as an unlucky one off instead of as the same rule, stated again.
Meta's policies are only one of the two boxes. For the other: the Australian rules that sit underneath all of it