In short: Ad platforms sort products into prohibited, restricted and open groups, using a process built to refuse a legitimate business rather than approve a harmful one. A restricted classification is a map of where to build distribution, not a review to argue with or evade.
Key points
- Restricted categories, such as financial services, crypto, gambling and health, run only with conditions like approval or licensing.
- Reviewers see the ad, landing page and account, never the product in use.
- A wrong approval costs the platform far more than a wrong refusal, so policy favours refusal.
- Plan in three bands: build in open channels, file restricted paperwork with nothing forecast, mark prohibited as closed.
- Disguised accounts, cloaked landing pages and vendors promising approvals lead to permanent bans.
Every major advertising platform maintains a policy that sorts products into three groups: those it will run without question, those it will run with conditions, and those it will not run at all. The middle group is what the industry calls restricted categories, and it is larger than most founders expect until they are in it.
The three groups
Prohibited. Products the platform will not advertise under any conditions. The list varies but typically includes weapons, illegal products, counterfeit goods, certain adult content, and, on some platforms and in some geographies, entire financial or health sub-categories. No documentation changes the answer.
Restricted. Products the platform will advertise only with conditions: prior written approval, proof of licensing or registration, certification, geographic limits, age restrictions, specific disclaimers, or limits on claims. Financial services, cryptocurrency, gambling, alcohol, pharmaceuticals, health and supplements, dating, political content, and several others sit here on most platforms, with the exact rules differing by platform and country.
Unrestricted. Everything else, subject to the general policies on claims, creative, and landing pages that apply to all advertisers.
How the decision is made
The platform's policy is written by its legal, trust-and-safety, and policy teams, and it is applied by a combination of automated classifiers and human reviewers who see the ad, the landing page, the account history, and sometimes the business documents, but never the product in use.
That last point explains most of the frustration. A reviewer is asked to decide in seconds whether an account looks like a category that has caused the platform legal or reputational problems before. A legitimate consumer fintech app looks, to a classifier, much like a predatory one. An AI product that mentions medical or financial use cases can be sorted into those categories by a keyword. A Web3 wallet is a crypto product regardless of what it holds.
Appeals exist, and sometimes succeed. They are slow, opaque, and decided by the same process.
Why legitimate companies get caught
The platform's cost of a wrong approval is high: regulatory action, press coverage, advertiser boycotts. The platform's cost of a wrong refusal is one lost small advertiser. The policy is written accordingly. Companies in reviewed categories are not being judged; they are being sorted, by a system designed to prefer refusing a legitimate business over approving a harmful one.
Recognizing that is useful. It means the rejection is not personal, is not evidence the product is wrong, and is not likely to be overturned by better copy.
What the classification means for a distribution plan
A company that learns it is in a restricted category has learned something valuable early: which channels it cannot plan around.
The practical response is a map. For each platform and geography that matters, record whether the category is prohibited, restricted, or open, what documentation a restricted approval requires, and whether the company can supply it. Then plan in three bands: build first in the channels that carry the category without approval, affiliate and partner programs, creator distribution with disclosure, clipping, owned email and content, and much of programmatic; file the restricted-band paperwork with nothing forecast against it; and write the prohibited channels down as closed, so nobody proposes them in the next planning cycle.
What it does not mean
It does not mean finding a way around the review. Accounts created to disguise a category, landing pages that show reviewers one thing and users another, and vendors who promise approvals are all routes to permanent bans, lost account history, and, in regulated categories, worse. A restricted classification is a fact about where to look for distribution. It is not a puzzle to be solved.
lowob takeaway: Platforms sort products into prohibited, restricted, and open, using a process built to refuse a legitimate business rather than approve a harmful one. Treat the classification as a map of where to build, not as a review to be argued with.
Common questions
What is a restricted category in advertising?
Products a platform will advertise only with conditions, such as prior approval, licensing proof, geographic or age limits, disclaimers or claim limits. Financial services, crypto, gambling, alcohol and health sit here on most platforms.
Why was my legitimate fintech or crypto ad account rejected?
Platforms sort accounts quickly by how they look, and a wrong approval costs them far more than a wrong refusal. The rejection is not personal and is unlikely to be overturned by better copy.
How should a restricted business plan its marketing channels?
Map each platform and geography as prohibited, restricted or open. Build first in open channels like affiliate, creator, clipping, owned email and content, and much of programmatic.
Should I try to get around ad platform category restrictions?
No. Disguised accounts, landing pages that show reviewers something different, and vendors who promise approvals lead to permanent bans and, in regulated categories, worse.