Ad policy and privacy

Policy is not a legal footnote — it is an operational constraint. A rejected ad costs a day; a restricted account costs the quarter. Both are usually avoidable by knowing what the review models look for.

Why ads get rejected

Most rejections come from a short list, and almost all of them are about the claim rather than the picture:

Reason	Typical form
Unsupported claims	Guaranteed results, specific income or weight figures
Personal attributes	Copy that implies knowledge about the reader ("your diabetes")
Before / after imagery	Body, skin, or condition transformations
Misleading presentation	Fake interface elements, fake play buttons, fake system alerts
Landing page mismatch	The page sells something other than the ad, or is broken
Restricted content	Alcohol, gambling, supplements, financial products, health

The second most common cause is not the ad at all — it is the destination. Review crawls the landing page, so a page with an unsupported claim can take down an ad whose copy was clean.

Restricted categories

Some categories are allowed but constrained: extra verification, limited targeting, mandatory disclaimers, or minimum audience ages. Health, finance, housing, employment, and political or social issue advertising are the usual ones, and the exact rules differ by platform and by country. Check the requirement before you produce the creative, because the constraint often changes the concept.

Appeals and account health

Handle a rejection as a diagnosis, not a dispute:

1. Read which policy was cited before editing anything
2. Fix the specific element, including the landing page
3. Appeal once, with the change described plainly
4. Do not re-upload the same ad repeatedly — repeated violations weigh on the account

Repeated rejections accumulate against the account, not just the ad. That is the real cost: an account with a history gets slower reviews and less benefit of the doubt.

Consent and measurement

Consent rules and platform tracking limits changed what you are able to see. The practical effects:

• A share of conversions is modelled rather than observed
• Click-based attribution windows are shorter than they were
• Consent banners reduce the events that reach the platform at all
• First-party data — your own customer list, server-side events — carries more weight than it used to

None of this makes measurement impossible; it makes single-source reporting unreliable. Read the consequences in measurement and attribution.