Policies
Content policy
How content is rated, which flags exist, and how it is decided whether a specific ad may appear beside a specific page. It is the mechanism behind everything this site claims about brand safety.
The scale
Five values. It is the same number that is stored in the database, not a translation for the website.
| Value | Name | Meaning |
|---|---|---|
| 0 | general | Safe for anyone. It is the default everywhere. |
| 1 | teen | Mild themes. It is the ceiling on what machine classification can assign on its own. |
| 2 | mature | Adult themes, artistic nudity, strong language. |
| 3 | adult | Explicit content. Requires an attested age gate on the site. |
| 4 | extreme never served | Illegal content, non-consensual material, or anything involving minors. It is a refuse-and-report state, never monetizable. |
Four dimensions, not one
Confusing "this page is strong" with "this ad is strong" is the mistake nearly every brand-safety failure comes from. Here they are four different things.
- What the publisher's page is: its rating and its flags, which can be narrowed per placement.
- What the ad depicts: the creative's rating and flags.
- What the advertiser tolerates: the strongest page it accepts appearing beside.
- What the publisher accepts: the strongest creative it admits in each placement.
The matching rule
- The creative's content rating does not exceed the maximum the placement accepts.
- The creative carries no content flag the publisher has blocked.
- The page's content rating does not exceed the maximum the campaign tolerates.
- The page carries no content flag the advertiser has blocked.
The site/creative asymmetry
The sentence most worth understanding on this page.
The platform serves ads on adult content but never serves adult ads. A placement may be rated up to and including explicit, provided the site declares an age gate and the country permits it. A creative, by contrast, can never reach the approved state above mature, and a schema constraint is what stops it.
It exists so that communities whose mature work no mainstream network will accept can monetise their inventory, without that meaning their pages fill up with adult ads. The cap on creatives is what keeps the second half of that sentence true.
The flag registry
Two groups with different purposes. Taste flags describe what content depicts; regulated flags carry legal obligations independent of taste.
Taste flags
| Bit | Key | Label | Human review |
|---|---|---|---|
| 0 | nudity_artistic | Artistic nudity | Yes |
| 1 | nudity_explicit | Explicit nudity | Yes |
| 2 | sexual_content | Sexual content | Yes |
| 3 | violence | Violence | Yes |
| 4 | gore | Gore | Yes |
| 5 | profanity | Profanity | No |
| 7 | fetish | Fetish | Yes |
| 8 | tragedy_news | Tragedy and hard news | No |
| 9 | religion | Religion | No |
| 10 | dating_mainstream | Mainstream dating | Yes |
| 11 | dating_adult | Adult dating | Yes |
Regulated flags
| Bit | Key | Label | Human review |
|---|---|---|---|
| 16 | gambling | Gambling | Yes |
| 17 | alcohol | Alcohol | Yes |
| 18 | tobacco | Tobacco and vaping | Yes |
| 19 | cannabis | Cannabis | Yes |
| 20 | drugs | Drugs | Yes |
| 21 | weapons | Weapons | Yes |
| 22 | pharma | Pharmaceuticals | Yes |
| 23 | financial_services | Financial services | Yes |
| 24 | crypto | Crypto assets | Yes |
| 25 | political | Political | Yes |
The defaults
All of them point the same way: what has not been reviewed is not permitted.
- The default rating is general, everywhere.
- No flag is set by default and no tolerance starts above general.
- Country policy permits nothing until it is reviewed. An unreviewed country enables nothing.
- Machine classification can propose, but it cannot carry anything above teen on its own. Anything above that level requires a named reviewer.
Reporting an ad
Anyone can, with no account and no session.
Every served ad carries a report link tied to that specific impression. The available reasons are misleading, offensive, adult content, malware or scam, broken, and irrelevant. Reports of adult content and of malware or scam cross a far lower threshold before escalating, precisely because those are the two categories where waiting is expensive.