How Google Ads Policies Affect Campaigns & How to Overcome Issues

A policy flag can stop a profitable campaign overnight, and it rarely happens for the reason the team assumes. Most disapprovals are structural — something about the destination, the claim or the account setup — rather than an attempt to advertise something obviously prohibited.

Understanding how the review system actually works turns a recurring emergency into a process. The teams that suffer least are not the ones that never get flagged; they are the ones that expect it and have a response ready.

The three levels of enforcement

Not every flag is the same size, and treating them all as equally urgent wastes the attention needed for the one that matters.

• Disapproved — a specific ad or keyword cannot run. Contained, usually fixable, and the most common outcome by far.
• Limited — the ad runs but delivery is restricted, often by geography or audience. Easy to miss, because nothing appears broken.
• Account-level action — suspension of the account itself. Rare, severe, and almost always the result of repeated or systemic issues rather than one bad advert.

The middle case causes the most silent damage. A limited ad keeps spending and keeps reporting, while reaching a fraction of the audience it should. Nobody investigates, because the campaign is technically live.

Where disapprovals actually come from

A handful of causes account for most of what teams encounter, and very few of them involve deliberately breaking a rule.

Destination problems are the largest single category. The landing page has to work, load properly on mobile, and deliver what the ad promised. A page that redirects unexpectedly, gates content behind a form the ad did not mention, or fails to load for the reviewer will be flagged regardless of how compliant the ad copy is.

Misrepresentation covers claims that cannot be substantiated, unclear pricing, or a business identity that is hard to verify. In regulated categories this is where good-faith marketing language most often crosses a line — an outcome stated as a certainty rather than a possibility.

• Destination mismatch — the page does not deliver what the ad offered.
• Unsubstantiated claims — guarantees, outcomes or comparisons with nothing behind them.
• Restricted categories — health, finance, insurance and similar verticals with certification or eligibility requirements.
• Trademark use — competitor names in copy, which is judged differently from bidding on them.
• Missing disclosure — terms, conditions, licensing or identity information the category requires.

Regulated categories need work before the campaign

Health, financial services, insurance and similar verticals carry requirements that exist before a single ad is written: certifications, verified identity, sometimes restrictions on which geographies can be targeted at all.

These are not obstacles to argue with after a disapproval. They are prerequisites, and campaigns launched without them will fail review repeatedly no matter how carefully the copy is written. Confirm eligibility during planning, not during a launch week.

How to appeal without making it worse

Appeals work, and they work better when the submission is specific. A flag names a policy; the appeal should address that policy directly rather than asserting in general terms that the ad is fine.

Fix the underlying issue first, then appeal. Appealing an unchanged ad in the hope of a different reviewer wastes the attempt and builds a pattern the account does not benefit from. If the destination was the problem, the page needs to change before the appeal is worth filing.

• Read which policy was cited, rather than guessing from the ad text.
• Change the thing that was flagged, and note what changed.
• Appeal once, clearly, with the specific policy addressed.
• If it fails twice for the same reason, the interpretation is wrong — rebuild rather than resubmit.

An appeal is a correction, not an argument. Change something first or there is nothing to review.

Reducing exposure before it happens

Most of the cost of policy enforcement is unplanned downtime, so the useful work is structural and done in advance.

Keep the destination stable during review periods. Avoid pushing landing page changes at the same time as a large creative refresh, because a flag then has two possible causes and diagnosis takes twice as long. Stage new claims on a small set of ads before rolling them across the account.

Structural separation helps as well. Keeping regulated and unregulated activity apart limits how far a problem can spread, and means an issue in one area does not put unrelated spend at risk.

How review actually works

Almost everything is reviewed automatically on submission. Systems compare the ad, the destination and the account against policy at scale, which is why a disapproval often arrives within hours and why the stated reason can feel oddly literal.

Human review generally enters on appeal or where automated classification is uncertain. This is the practical reason a well-argued appeal can succeed on an unchanged policy: the first decision was made by a system reading patterns, and the second by a person reading context.

It is also why edits trigger re-review. Changing an approved ad resubmits it, which means a routine copy tweak during a sensitive period can take a working advert offline. Batch edits deliberately rather than adjusting live ads through the week.

The copy patterns that get flagged

In regulated verticals, most flags are caused by ordinary marketing language rather than anything anyone would recognise as a violation. The recurring problem is certainty — stating an outcome as guaranteed when it depends on circumstances the advertiser does not control.

• Outcomes promised rather than described — approval, savings, results, eligibility.
• Urgency implying consequence, particularly around deadlines, benefits or coverage.
• Implied endorsement or official status where none exists.
• Superlatives and comparisons with nothing published to support them.
• Pricing that omits the conditions attached to it.

The rewrite is usually small. Describing what the service does, who it suits and what happens next will clear review where the same claim framed as a guarantee will not — and it tends to convert better, because it sets an expectation the landing page can meet.

The destination is reviewed too

Teams optimise ad copy endlessly and forget the page is part of the submission. A reviewer, automated or otherwise, will load it — from an unexpected location, on a device you did not plan for, without the session context a normal visitor arrives with.

• The page loads reliably, including from outside your usual geography.
• Content matches the promise in the ad, not a broader version of it.
• Required disclosures, licensing and contact details are present and findable.
• No unexpected redirect, interstitial or gate that the ad did not disclose.
• A privacy policy and terms that actually correspond to what the page collects.

The first hour of a suspension

Account-level action is the scenario worth rehearsing, because it is the one where instinct is most often wrong. The instinct is to open a new account and keep spending. That is the single action most likely to turn a recoverable suspension into a permanent one.

Work in order instead. Read the stated reason carefully and take it literally. Identify what changed recently across ads, pages and billing. Correct the cause before submitting anything. Then appeal once, describing precisely what was wrong and what was changed — an appeal that only asserts good intentions gives the reviewer nothing to act on.

Build the recovery path before you need it

The difference between an hour of disruption and a week of it is usually preparation. An account that can identify what was flagged, when it changed and who owns the fix will resolve most issues quickly. An account where nobody is sure which version of the landing page was live will not.

• A record of what changed and when, across both ads and destination pages.
• A named owner for policy issues, so nothing waits for someone to notice.
• A monitoring habit that catches limited status, not only outright disapproval.
• A fallback set of conservative, pre-approved creatives to keep delivery alive during a fix.

That last point is the one most teams skip and the one that saves the most money. Enforcement will happen eventually in any serious account. Having something compliant ready to run means a flag costs a few hours of performance rather than a week of silence.