Programmatic client report: where ads ran, what changed and what is not known
Part of Programmatic media planning
The short answer
A programmatic client report should show where ads actually ran, which exclusions applied and why, the evidence behind each statement, what changed during the flight and which risks remain open. Each metric carries its platform definition and extraction date, and anything the data cannot show is labeled as unknown rather than reported as zero or as a result.
Key takeaways
- Lead the report with placement facts and changes, then delivery metrics, each with the definition the platform uses.
- Report exclusions by the level that set them and the reason, separating platform protections from the brand's own choices.
- Label each figure as recorded, measured or modeled, and keep causal claims out unless a test supports them.
- An unknown stated plainly is more useful to a client than a gap filled with a guess.
What should a programmatic client report show?
A client report answers the questions a brand manager will be asked by their own leadership: where did our ads run, what did we keep them away from, what changed and what are we still unsure about. Delivery and cost figures belong in the report, but they come after the placement facts and each one carries its definition, because the same word means different things in different platforms.
| Part | What it shows | Evidence |
|---|---|---|
| Where ads ran | The domains, apps and deals that carried delivery, matched to the approved list | Placement report by site or app, exchange and inventory source, with the extraction date |
| Exclusions and why | Each exclusion, the level that set it and the reason | Exported targeting lists, suitability settings, the exclusion log |
| Delivery and cost | Impressions, viewable impressions, video completions and cost, each defined | Platform report with a metric dictionary |
| Changes | What changed during the flight, when, by whom and why | Change history export with notes |
| Evidence | The files a reader can open to check any statement | An appendix list with file names and dates |
| Open risks and unknowns | What the data could not show and what is still being checked | Masked or unrateable rows, missing reach figures, pending reviews |
What evidence should accompany a client ready media plan?
A client report is only as checkable as the plan it refers back to. If the plan went out as a slide of logos, the report has nothing to compare delivery against. Send the plan with an evidence pack, and the report can then cite version numbers instead of describing the plan from memory. How to write the reason for each site is covered in the guide to explaining site selection; the document layout is in the media plan template.
Evidence pack sent with the plan
- The confirmed brief, with the date and name of the person who confirmed it.
- The approved list version number, its domain count and its approval date.
- One reason per site, and a separate list of sites left out with the reason for each.
- The suitability settings and avoid topics that will apply, with the level where each is set.
- Any forecast, labeled as a range with its sources and assumptions.
- The metrics the report will use and the platform definition of each.
How should exclusions be reported?
Exclusions come from several places, and a report that lumps them together misleads in both directions. In Display & Video 360, Google's brand suitability guide (opens in a new tab) separates brand safety, which automatically filters inappropriate or harmful content, from brand suitability, which lets an advertiser opt in or out of content some advertisers accept and others do not. A report that credits the brand's own settings for a platform level protection overstates what the brand chose.
| Source | Documented behavior | How to report it |
|---|---|---|
| Platform brand safety | Applied automatically; an exempt deal still keeps Google's automatic protections | One line stating it applied; not presented as a brand decision |
| Advertiser suitability settings | An exempt deal "only bypasses the brand suitability settings you control at the advertiser level" | Each setting and tolerance, and whether any deal was exempted |
| Inherited channels | Advertiser or partner channels can appear as blocked and not editable in a line item's Apps & URLs targeting (DV360 targeting guide (opens in a new tab)) | Named as set above the campaign, with the owner |
| Campaign exclusion list | Apps & URLs settings support inclusions and exclusions at the line item level | The list version, its size and the reason for each addition |
| Mid flight exclusions | Event history (opens in a new tab) records the time, date, description and user for a change | Date added, the incident that prompted it and the evidence file |
Which findings can be audited from platform reports after launch?
Some statements in a report can be checked by anyone with the same report; others depend on models or attribution rules. Mark the difference. The definitions below come from Google's metrics in reports (opens in a new tab) and dimensions in reports (opens in a new tab) pages for Display & Video 360.
| Field | Platform definition | What it supports |
|---|---|---|
| App/URL | "The name of a site, or mobile app" | Which properties carried delivery; not each page an ad appeared on |
| Impressions | The number of times a user's browser or device sent a ping to the ad server | Served volume by site, line item and date |
| Active View: Viewable Impressions | Impressions that were viewable "out of all measurable impressions" | Viewability for the measurable subset only; report measurable impressions beside it |
| Media Cost | "The raw cost for impressions purchased from an exchange" | Working media cost, before platform and partner fees |
| Revenue | The amount spent to purchase impressions and deliver ads through the platform | What the buyer was charged; not a sales figure |
| Post-View Conversions | Times a person landed on a success page after viewing an ad | A count under an attribution rule; not proof that the ad caused the visit |
Numbers move after the first pull. Google's report data freshness page (opens in a new tab) says impression and click data is updated every 3 to 4 hours and that Display & Video 360 can update data for up to 31 days while it confirms the validity of clicks, impressions and conversions. It keeps most reporting data for 24 months, but hourly data for up to 31 days and inventory availability for 30 days. Date each extract, and reissue a figure that changed with a note.
How do you report what is not known?
Unknowns are part of the result. A blank cell read as zero becomes a claim that nothing happened. Use fixed labels, define them once at the top of the report, and keep them out of totals.
| Label | Meaning | Documented example |
|---|---|---|
| Not available | The platform did not return a value for this cut | A dash in the DV360 Unique Reach report means "the model couldn't make an accurate calculation" (reach reports (opens in a new tab)) |
| Lower confidence | A modeled figure the platform marks as less certain | Instant reports can include estimated numbers with lower confidence, filterable by Unique Reach Sample Size |
| Not classifiable | The content could not be rated | Unrateable impressions in DV360 brand suitability |
| Pending | The data window is still open | Reach metrics can take up to 3 days to appear; data can update for up to 31 days |
| Measured zero | The data is complete and the count is zero | Use only when the source confirms coverage for the full period |
The inferred tier matters beyond the meeting room. When a brand repeats a campaign claim in its own advertising, the FTC's policy statement on advertising substantiation (opens in a new tab) of November 23, 1984 says firms must have a reasonable basis for objective claims before they are disseminated. A report that labels attribution counts as counts, not as proof of effect, keeps that distinction visible.
How does PlacementGraph supply the planning evidence a report refers back to?
What does a client report look like for one flight?
Fictional example
A Bay Area ballet company's season subscription report
A fictional ballet company in San Francisco runs a six week display flight for its season subscriptions across the Bay Area and Sacramento, on arts, culture, dining and local events sites. The plan went out with an evidence pack: the confirmed brief, list version 2 with its domain count, a reason per site and the advertiser level suitability settings.
The report opens with where ads ran, matched to version 2, and notes two domains added in week three as version 3 with the reason. The exclusions section separates the platform's automatic protections from the company's own settings, names a partner level channel inherited from the agency, and states that no deal was exempted from suitability settings.
Delivery figures follow with their definitions and extraction date. Viewable impressions sit beside measurable impressions. Unique reach for one line item is labeled not available, and post view conversions are labeled as attribution counts. The open risks section lists one exchange whose URLs were masked and a pending review of a culture blog's new comment section.
Frequently asked questions
How often should a client receive a programmatic report?
Match the cadence to the decisions the client can make. A short weekly note on placements and changes suits a flight where lists can still be edited; a full report at the end of the flight, after the data window closes, carries the final figures. Reissue any figure that changed between the two.
Should the report include raw placement data?
Attach it as an appendix file rather than pasting it into the body. The body summarizes where ads ran and what stood out; the file lets the client check any line. Name the file, its date range and its extraction date so a reader can match it to the summary.
Can a client report compare this campaign with the last one?
Only where the definitions, platforms and settings match. If viewability was measured on a different subset, or the list changed size, say so beside the comparison. A comparison across different measurement rules shows the change in rules as much as any change in the campaign.
Who should sign off a client report before it is sent?
Someone who did not build it and can check the figures against the source reports, such as the account lead or a senior analyst. The sign off confirms that each figure matches its extract and that unknowns are labeled, not that the campaign succeeded.
Sources
- Metrics in reports (opens in a new tab)Display & Video 360 Help
- Dimensions in reports (opens in a new tab)Display & Video 360 Help
- Report data freshness and availability (opens in a new tab)Display & Video 360 Help
- Brand suitability (opens in a new tab)Display & Video 360 Help
- App, channel, and website targeting (opens in a new tab)Display & Video 360 Help
- Reach reports (opens in a new tab)Display & Video 360 Help
- Event history (opens in a new tab)Display & Video 360 Help
- FTC Policy Statement Regarding Advertising Substantiation (opens in a new tab)US Federal Trade Commission, November 23, 1984
Researched and written by PlacementGraph Research. Read how we research, source and update these pages. Corrections: diana@placementgraph.com.