What should a brand safety report show, and how does it test the policy?
Part of Brand suitability
The short answer
A brand safety report shows four things beside the policy it was built on: incidents where ads ran against the policy, blocks the controls made, content nobody could classify, and the share of delivery the method could assess. Read together, they test whether the policy worked, overblocked or left gaps, which a single safe rate cannot show.
Key takeaways
- A suitability report restates the policy in force, then sorts delivery into incidents, blocks, unknowns and gaps against it.
- Post campaign reporting tests a policy in two directions: content it should have stopped and content it stopped without reason.
- Google says it cannot guarantee that all related content will be excluded, so a report states what was found in what was reviewed.
- Rates built on property level classification, unbilled impression counts or a hand drawn sample need their method beside the figure.
What should a brand safety report include?
A brand safety report is not a delivery report with a safety column added. It is a test of a written brand suitability policy against what happened, so it starts with the policy itself: the categories, the tier of tolerance for each, the controls used to enforce them and the dates they were in force. Without that page, a reader cannot tell an incident from a deliberate choice.
| Part | Question it answers | Evidence to attach |
|---|---|---|
| Policy in force | What did the brand say it would and would not fund? | Category and tier settings with effective dates and any approved exceptions |
| Incidents | Where did ads run against the policy? | Screen copies, URL strings and log rows for each placement |
| Blocks | What did the controls stop, and on whose rule? | Exclusion and block counts split by the rule that produced them |
| Unknowns | What could not be classified? | Delivery on unlabeled or unrateable content, kept out of safe totals |
| Coverage | How much of delivery could the method assess? | The share of the campaign the method reached, and how |
| Method | How were the judgments made? | Property or content level, census or sample, data age |
| Policy findings | What should change before the next flight? | Rules that missed, rules that overblocked, and the proposed edit |
The evidence column comes from the IAB and MRC Guidelines for the Conduct of Ad Verification (opens in a new tab) of February 2012, which list screen copies, tag and URL strings, ad log files, crawler logs and discussion records as appropriate evidence, and say that any altered material, such as an ad placed on an illustrative page, should be disclosed. Their project disclosures include "Evidence of Exceptions, Screen Shots, etc." and, for a census, the "Percent of total campaign able to be verified".
How can post campaign reporting test a suitability policy?
A policy can fail in two directions. It can let through content the brand meant to avoid, or it can stop content the brand would gladly have funded. Delivery reports show only the first kind, because blocked impressions leave no page to look at. TAG's Brand Safety Certified Guidelines (opens in a new tab), version 5.0 of July 2026, describe the same idea for certified companies: quarterly internal reviews should use "actual testing of data, both quantitatively and qualitatively" to confirm that controls are "designed correctly and operating effectively."
Four tests a post campaign report runs on the policy
- Check that the controls ran as configured. Compare the settings in force each week with the approved settings, and list each change with its date and author.
- Review the incidents. For each placement that broke the policy, name the rule that should have caught it and why it did not.
- Sample the blocks. Read a set of blocked pages and sort them into intended blocks and overblocks, since that is the only way blocked inventory becomes visible.
- State what could not be tested. Report unknown and unmeasured delivery separately, with its causes, so the other three tests are read at their true scope.
The block sample is where reports earn their keep. A rule that stops responsible reporting on health, crime or politics can show up as a pattern in blocked pages before anyone complains, and the fix is a policy edit, not a new exclusion.
Can any content suitability filter guarantee zero misplacements?
No platform documentation reviewed for this guide makes that promise. Google's content suitability help page (opens in a new tab) states "we can't guarantee that all related content will be excluded," and that content exclusions "may provide limited coverage of in-feed ads in YouTube" and of YouTube Shorts. The 2012 IAB and MRC guidelines say a verification service should try constantly to reduce the impact of "false positives and negatives or ambiguities from technological limitations", which assumes those errors exist.
The reasons are structural. Classification runs on samples and on data of a certain age, pages change after they are labeled, and some delivery arrives with masked URLs that a classifier cannot read. A report that says zero misplacements is claiming knowledge of pages nobody measured.
How should coverage gaps be reported?
A coverage gap is delivery the method could not judge, and it belongs in its own row, never folded into a safe total. The MRC's 2018 supplement on content level brand safety (opens in a new tab) says content not meeting baseline guardrails "must be classified as unknown or not yet classified". The MRC's October 18, 2025 policy (opens in a new tab) goes further for property level services, which must disclose that images, video and audio are excluded, when unclassified URLs are "rolled up" to domain level classifications, their time to live policies and crawl rates, and sampling and non sampling error.
| Gap | Documented cause | How to report it |
|---|---|---|
| Unlabeled content | Display & Video 360 names new sites, masked URLs and unsupported languages (digital content labels (opens in a new tab)) | Delivery on unlabeled content, with the cause where known |
| Rolled up URLs | Unclassified URLs or subdomains assume the domain's classification (MRC 2025) | The share of judgments made at domain level, not page level |
| Images, video and audio | Property level services exclude them from measurement (MRC 2025) | A line stating which content types the method read |
| Below reporting thresholds | Placements that miss minimum impressions or interactions are left out of the Where ads showed report | The Total: Other line, explained, beside listed placements |
| YouTube surfaces off the watch page | Ads not on the watch page, such as the Home feed, are reported as youtube.com in the Content suitability report | Named as unassessed at page level |
| Aging labels | Display & Video 360 updates its classification monthly | The data age behind each judgment |
Google's Where ads showed report (opens in a new tab) defines Total: Other as the difference between the placements it lists and the total for the account, campaign or ad group, caused by ads that did not run on YouTube or the Display Network, reporting filters, or placements below the minimum thresholds. A report that divides incidents by listed placements alone uses the wrong denominator.
Which metrics mislead in brand safety reports?
Each figure below is accurate on its own terms and misleading when read as a verdict on brand safety. The fix in each case is to print the method beside the number.
| Figure | Why it misleads | What to print beside it |
|---|---|---|
| A brand safe rate from domain level classification | The MRC's 2025 policy bars accredited property level services based on text from positioning their results as brand safety | Property level or content level, and which content types were read |
| A high block rate shown as success | Blocks include overblocks; Google notes the Limited inventory type "may limit your reach and performance" | A sample of blocked pages sorted into intended blocks and overblocks |
| Impressions from the Content suitability report | Google says counts "don't reflect the impressions that you're billed" and Shorts impressions may not be deduplicated | The billing source used for totals |
| Incident rate on report totals | The suitability and Where ads showed reports are "not regularly updated and may include invalid traffic" | The denominator, its source and its date |
| Zero incidents from a hand review | A judgmental sample cannot be generalized across a campaign, per the 2012 guidelines | Selection method, sample size and the period sampled |
| Invalid traffic filtered pre bid | A fraud measure in Display & Video 360 brand controls (opens in a new tab), not a content judgment | Reported in its own section, apart from suitability |
Google's Content suitability report page (opens in a new tab) is direct about its scope: "The Performance Max placement report and the Content suitability report's purpose is strictly for brand suitability." Use it to find where ads ran, and take impression and cost totals from billing.
How does PlacementGraph state the tolerance a report tests?
What does a suitability report look like for one flight?
Fictional example
A cereal maker reviews its back to school flight in Maryland and Delaware
An unnamed breakfast cereal maker runs an eight week back to school campaign in Baltimore, Annapolis and Wilmington, on family food, parenting and local news sites and on YouTube. Its policy avoids tragedy and conflict at medium tier and above, and allows news about schools and nutrition.
The report opens with that policy and its dates. Two incidents are listed, each with a screen copy and URL, both on one recipe site page later edited to cover a fatal kitchen fire. The cause is a label older than the edit, so the finding is about data age, not the rule.
The block sample tells a different story. A set of local news pages about school lunch nutrition was blocked by a content keyword the agency added for a different client's policy. The report flags it as an overblock and proposes removing the keyword. A coverage row states that YouTube delivery off the watch page could not be assessed at page level, and the report says so instead of counting it as safe.
Frequently asked questions
Should a brand safety report name the publishers involved in incidents?
Inside the client report, yes, since the evidence needs a URL. The 2012 IAB and MRC guidelines also encourage sharing known exceptions with sellers to enable process correction, so the publisher can fix a page or a feed. Whether the brand shares names beyond that is its own choice.
Can a client check the incidents in a report for itself?
It should be able to. The 2012 guidelines say evidence supporting significant findings should be retained so all parties can examine it, and that the customer above all should be able to review it. Attach the screen copy, URL and time for each incident, and disclose any image that was recreated.
Should incidents the client found itself be in the report?
Yes, with their source marked. An incident spotted by a brand manager is as real as one found in a sample, and leaving it out makes the report look stronger than the monitoring was. Recording who found each incident also shows which checks are catching problems.
Is the Content suitability report a billing record?
No. Google says its impression counts do not reflect the impressions billed in Google Ads and points to billing invoices for those figures. Use the report for where ads ran by source and channel, and take totals and spend from billing.
Sources
- Guidelines for the Conduct of Ad Verification (opens in a new tab)IAB and Media Rating Council, February 14, 2012
- MRC Supplement to IAB Guidelines for the Conduct of Ad Verification: Enhanced Content Level Context and Brand Safety (opens in a new tab)Media Rating Council, September 2018
- Policy for Property-Level Ad Verification Representations (opens in a new tab)Media Rating Council, October 18, 2025
- Brand Safety Certified Guidelines, version 5.0 (opens in a new tab)TAG (Trustworthy Accountability Group), July 2026
- About content suitability (opens in a new tab)Google Ads Help
- About the Where ads showed report (opens in a new tab)Google Ads Help
- About the Content suitability report (opens in a new tab)Google Ads Help
- Digital content labels in Display & Video 360 (opens in a new tab)Google Display & Video 360 Help
- View and edit brand controls for your advertiser (opens in a new tab)Google Display & Video 360 Help
Researched and written by PlacementGraph Research. Read how we research, source and update these pages. Corrections: diana@placementgraph.com.