The Restricted Category Playbook: Scaling Ads the Platforms Keep Rejecting

The first time an ad account of ours got restricted on a health brand, we lost nine days.

Not nine days of underperformance. Nine days of zero. The account went down on a Thursday, the appeal sat in a queue over a weekend, and by the time it came back the campaigns had lost every bit of learning phase progress they had accumulated.

The cause was one line of copy in one ad, written by someone who had written a thousand ads and had never worked in a restricted category before.

I run growth at a telehealth brand now. Weight management, hair loss, men's health. Categories where the platforms are strictest, the rejection rate is highest, and the downside of getting it wrong is not a paused ad set. It is the account.

Most operators treat compliance as a tax on creative. It is actually a structural advantage, and I will explain why.

A note before the substance: platform policies change frequently and vary by market. Treat everything below as a framework for how to think, and verify the specifics against current policy documentation and your own counsel before you launch.

Why the Normal Creative Workflow Breaks

In an unrestricted category, your workflow probably looks like this. Brief, produce, upload, launch, read the numbers.

In a restricted category, that workflow has a defect that only shows up at scale. Every asset that reaches the upload step is a compliance decision that nobody consciously made. You are running a review process, you just have not admitted it, and the reviewer is whoever happened to be uploading that day.

Three things make it worse than it sounds.

Rejections are not evenly distributed. They cluster. One angle that a reviewer flags will get flagged across every variant built on that angle, so a single bad concept can take out an entire production sprint.

Enforcement is cumulative. Individual rejections are cheap. What is expensive is the pattern. Enough of them in a short window and you are dealing with account level restriction, which is the outcome you actually have to design against.

Appeals are slow and unpredictable. You cannot plan a media calendar around a queue you do not control. Which means the only reliable strategy is not needing the appeal.

The fix is not to write more cautiously. Cautious copy in a restricted category is usually just bad copy that still gets rejected. The fix is to move compliance upstream, into the brief, where it costs almost nothing.

The Three Categories of Constraint

Every restricted category rule I have dealt with falls into one of three buckets, and they need different responses.

Constraint 1: What you may not imply about the person

The rule that catches the most operators off guard. On Meta specifically, ads must not assert or imply knowledge of a person's personal attributes, and health conditions sit squarely inside that. This is why "Struggling with hair loss?" is a problem and "Hair loss is more common than most men think" is not.

Same claim. Same audience. One addresses the viewer as a diagnosed subject, the other describes a category.

Once your writers internalize this single distinction, your rejection rate drops more than from any other intervention. It is one sentence of training and it is the highest ROI compliance work you will ever do.

Constraint 2: What you may not claim about the outcome

Prohibitions here typically cover before and after imagery, depictions of unexpected or exaggerated results, specific timelines, guaranteed outcomes, and anything implying a medical result the product cannot substantiate.

This is the constraint that feels most like it is taking away your best performing creative, because it usually is. Transformation content works. That is precisely why it is restricted.

The response is not to abandon proof. It is to move proof from the outcome to the process. Show the mechanism, the protocol, the clinical oversight, the ingredient, the practitioner. Show what someone experiences rather than what they end up looking like.

Constraint 3: Who is allowed to advertise at all

The gating layer, and it is not a creative problem. Certain categories require verification or certification before you can run at all in certain markets. Telehealth and online pharmacy advertising in the US, for example, typically requires third party certification before Meta or Google will let you run, plus written permission in some cases.

This one is boring, it takes weeks, and it is entirely front loadable. Do it before you build a single asset. I have watched brands produce a full creative library and then discover they were not eligible to run any of it.

Platform by Platform: What Actually Differs

The strategic error is treating your restricted category ads as one asset library deployed everywhere. The tolerance bands differ enough that this guarantees waste.

Meta. The personal attributes rule is the dominant constraint. Age gating applies to certain categories. Certification requirements gate entry for health services. Enforcement is largely automated at first pass with human review on appeal, which means literal pattern matching in copy matters more than intent.

Google. Search intent means the user has already declared the condition themselves, which changes the copy problem substantially. The constraint shifts toward certification, landing page compliance, and what the destination page claims. Your landing page is part of the ad here in a way that operators from a paid social background consistently underestimate.

TikTok. The most restrictive of the three for health. Entire product categories are prohibited outright in many markets rather than merely restricted, and the rules vary meaningfully by country. Check market by market before you build, not after.

Reddit and the rest. Generally more permissive on category, and the community context does its own enforcement. An ad that survives review and gets destroyed in the comments is not a win.

The practical implication is that you build a core creative concept and then produce platform specific executions of it. Not one master file with different aspect ratios.

The Compliance Layer Inside the Brief

This is the actual system, and it is four additions to the creative brief you already use.

Addition 1: The claim ladder

Before anyone writes a word, define three tiers for the concept.

Tier 1, always safe. Category level education, mechanism explanation, general statements about the condition space. No addressing, no promising.

Tier 2, conditionally safe. Requires specific substantiation, a disclaimer, or a particular framing to run. Usable, but only with the supporting element attached.

Tier 3, do not build. Outcome claims, before and after, direct personal address, anything implying diagnosis.

Writers work from the ladder, not from memory or vibes. New writers are productive in days instead of months, which matters enormously if you are running a lean creative operation.

Addition 2: The phrasing swap table

A living document, maintained per platform. Left column is a construction that gets rejected. Right column is the compliant version that preserves the persuasive intent.

This is the single most useful artifact in the whole system, because it converts hard won rejection experience into something a new hire can use on day one. Every rejection you take should add a row. If it does not, you paid for the lesson and threw away the receipt.

Addition 3: Pre flight review, not post rejection review

One named person reviews every asset before upload against the ladder and the swap table. Fifteen minutes per batch.

The objection I hear is that this slows down velocity. It does the opposite. Rejections cost you an upload cycle, a resubmission, a possible appeal, and a hit to the account pattern. A pre flight check costs fifteen minutes and catches most of it.

Addition 4: Documented substantiation

For every Tier 2 claim, a file that says what backs it up. Study, internal data, practitioner sign off, whatever it is. Stored where the writer, the reviewer, and the person handling the appeal can all reach it.

When an appeal comes, you have thirty seconds of attention and one shot to be convincing. Having the substantiation ready is the difference between a same day resolution and a week.

Account Architecture as Insurance

The creative system reduces frequency. The architecture reduces blast radius. You need both.

Separate ad accounts by risk profile. Highest risk category in its own account. If it goes down, the rest of the business keeps spending.

Business Manager hygiene. Verified business, clean admin structure, no shared logins, no assets floating between entities. Most account level enforcement I have seen escalate faster than it should escalated because the account structure looked suspicious independent of the ads.

Domain separation where it is legitimate. Different verified domains for genuinely different product lines. Not to obscure anything, which does not work and makes things far worse when discovered. To limit the correlation between one product line's issue and another's delivery.

A cold standby. A second verified, warmed account with the pixel, catalog, and audiences already in place, running a small always on budget. It exists so that the answer to an outage is hours instead of the nine days I opened this piece with.

Landing page parity. Your destination page must claim exactly what your ad claims, no more. Compliant ad, aggressive landing page is one of the fastest ways to lose an account, and it is a mistake made almost entirely by teams where the media buyer and the landing page owner never talk.

A Worked Example

Anonymized, from a health services brand. Two consecutive quarters, same team, same spend level. The system described above was implemented at the start of Q2.

Nothing about the team, the budget, or the production capacity changed. The only change was moving the compliance decision from after upload to before the brief.

Effective creative velocity went up 63% without producing a single additional asset. In a category where creative volume is the primary constraint on scale, that is the whole game.

Why the Constraint Is Actually an Advantage

Here is the part most operators miss.

The reason restricted categories stay profitable is that the compliance overhead is a barrier to entry. Every competitor who has not built this system is running at a 35% rejection rate, burning production budget on assets that never see delivery, and periodically going dark for a week at a time.

You are not competing on media buying skill in these categories. You are competing on how much usable creative you can get into the auction consistently over time. Which means the operator with the boring documented review process wins against the more talented operator without one.

That is a durable advantage. Creative talent gets poached. A twelve month old swap table built from your own rejection history does not.

What I Look At Monthly

1. Rejection rate by platform. Trend, not absolute. A rising rate is an early warning of a policy shift, and you want to catch it before enforcement does.
2. Rejection rate by writer. Not to punish. To find who needs an hour of training.
3. Rejection reason clustering. If one reason dominates, the ladder or the swap table has a gap.
4. Time from rejection to resolution. Measures whether your appeal documentation is working.
5. Swap table growth. If it stopped growing, someone stopped feeding it.
6. Standby account health. Verified, warmed, spending a small amount. Check it quarterly, because a cold standby that has gone stale is not a standby.
7. Landing page audit. Every live destination against every live ad claim. Monthly. This is where drift happens silently.

FAQ

What is the single highest impact change I can make this week?

Train every writer on the difference between addressing a person's condition and describing a category. That one distinction accounts for a large share of health category rejections on Meta.

Do I need legal review on every ad?

No, and it would be unworkable. You need legal review on the claim ladder and the substantiation file. Once those exist, individual asset review is an operations task, not a legal one.

How do I handle a rejection I believe is wrong?

Appeal once with your substantiation attached, and while it is pending, ship a Tier 1 version of the same concept so the concept keeps spending. Never let a single asset's appeal block a whole angle.

Is testimonial content usable in restricted categories?

Sometimes, with care. The risk is that a testimonial makes an outcome claim on your behalf that you could not make directly. Brief creators explicitly on the ladder, and review their content against the same standard as your own.

What do I do if my account is already restricted?

Stop submitting. Additional submissions during an active review can compound the problem. Audit everything live, appeal once with documentation, and route spend to the standby account. If you do not have a standby, build it while you wait, because this will happen again.

Does this apply outside health?

Yes, with different specifics. Financial services, supplements, alcohol, gambling, dating, and anything touching social issues run on the same structure: an entry gate, claim restrictions, and an audience restriction. The system transfers even when the rules do not.

How much does the review layer actually slow down launch?

In our case, roughly fifteen minutes per batch, and it made net launch velocity faster because the resubmission cycle disappeared. If your reviewer is a bottleneck, the swap table is too thin, not the process wrong.

Should the same person buy media and review compliance?

No. The buyer wants the asset live. Give the review to someone whose incentive is the account staying up.

Closing

The instinct in a restricted category is to feel constrained, and to blame the platform for the ads you cannot run.

I have watched teams spend more energy litigating why a rule is unfair than it would have taken to build the system that operates cleanly inside it. The rule is not going to change because you are annoyed by it, and the competitor who accepted it six months ago is currently taking your impressions.

The other thing worth saying plainly: most of these rules exist for reasons. Health advertising that addresses people as diagnosed subjects and promises outcomes it cannot substantiate causes real harm to real people. Building a system that produces honest, mechanism led, substantiated creative is not just risk management. It happens to also be how you build a brand people trust enough to buy something medical from.

Compliance moved upstream is not a constraint on creative. It is the thing that lets creative run.

For related reading on how the creative system underneath this works, see the creative testing system that produces real winners and how to brief a UGC creator for performance.

Keep reading

Pieces I've written on related topics that pair well with this one:

• The Creative Brief Template I Use for Every Ad Campaign, the brief framework this compliance layer sits inside.
• Creative Velocity: How Many New Ads Should You Launch Monthly, the spend tier benchmarks for creative volume.
• How to Build a High-Output Creative Team Without 15 People, the lean production model.
• How to Design a Landing Page That Converts Cold Traffic, the destination side of the claim parity problem.
• The Creative Testing System That Produces Real Winners, the infrastructure that turns tests into compounding learning.