What is working media, and how should a ratio be reported?
Part of Supply path transparency
The short answer
Working media is the part of an advertising budget that pays for ad exposure itself, as opposed to fees, data, tools and services. A working media ratio divides that amount by a stated base, such as total programmatic investment or money entering a DSP. The ratio means little until the report names its numerator, its denominator and its scope.
Key takeaways
- Working media has no single public standard definition; ISBA used publisher revenue, while the ANA measures spend that delivers qualified impressions.
- Platform media cost is not the same as publisher revenue: in DV360 it is the raw cost of impressions bought from an exchange, before the exchange's own share.
- The denominator decides the answer, so the same campaign can show very different ratios against advertiser spend and against money entering a DSP.
- The ANA's TrueAdSpend counts spend on impressions that are fraud free, measurable, viewable and MFA free, among benchmark participants only.
- A usable report states the numerator, the denominator, the channels, the period, the currency and the data source.
What is working media?
Definition
- Working media
- Spend that buys the ad exposure itself. Non working media is everything else charged against the same budget: agency and platform fees, data, ad serving, verification and any amount no party can account for.
In programmatic buying the money passes through several companies before a publisher is paid, which makes the line hard to draw. Public sources use it in at least two ways. The ISBA programmatic supply chain transparency study (opens in a new tab) of May 2020 described "the proportion of advertiser spend reaching publishers as 'working media'." The ANA instead tracks how much spend produces impressions that pass quality tests, a measure it calls TrueAdSpend.
| Definition | Numerator | Strength | Weakness |
|---|---|---|---|
| Publisher revenue (ISBA 2020 usage) | Money the publisher received for the impressions | Shows exactly how much the intermediaries kept | Needs seller side data, and a paid impression can still be fraud or unviewable |
| Qualified impressions (ANA TrueAdSpend) | Spend on impressions that are fraud free, measurable, viewable and MFA free | Counts waste as well as fees | Depends on the quality criteria and the vendors measuring them |
| Platform media cost | The media cost line in the buying platform | Available in standard reports, no outside data needed | Includes the exchange's share and says nothing about quality |
Which costs count as non working media?
Google's Display & Video 360 metrics reference (opens in a new tab) gives a concrete set of cost lines to sort. It defines media cost as "the raw cost for impressions purchased from an exchange." Its Total Media Cost adds to that media cost the fees for third party audience data and partner costs, which it describes as "any partner costs, such as the fees for using Display & Video 360," a third party ad server or a verification service. Despite its name, Total Media Cost therefore mixes working and non working lines.
| Cost line | Publisher revenue reading | Qualified impressions reading |
|---|---|---|
| Publisher's share of the clearing price | Working | Working only for impressions that pass the quality tests |
| Exchange or SSP share inside media cost | Non working | Counted inside spend; not separated |
| DSP platform fee | Non working | Non working |
| Third party data fees | Non working | Non working |
| Ad serving and verification fees | Non working | Non working |
| Agency fee or markup | Non working | Non working |
| Media cost on fraud, unmeasurable, unviewable or MFA impressions | Working, if the publisher was paid | Non working |
The last row is where the definitions part ways. An impression on a made for advertising site can pay its publisher in full and still fail the quality test. A team that reports only fee based working media can look efficient while the budget lands on pages no reader chose. The made for advertising sites hub covers how those sites are found.
Agency fees need one more step. ISBA found agencies paid through a fixed fee for all services, a fixed fee for programmatic, a commission, or a fee bundled with ad tech, and converted each into an estimated percentage to place it in the waterfall. It noted that these models "prevent direct cost comparisons of agencies." A working media report for a client on a fixed fee should therefore show how the fee was turned into a share of spend, and over which period.
What is the denominator of a working media efficiency claim?
A ratio has a base, and an efficiency claim without one cannot be checked. Three published bases show how much it matters. ISBA divides by advertiser spend, the gross amount before the agency's fee. The ANA's December 2024 benchmark used money at the DSP: its 2024 release (opens in a new tab) said "For every $1,000 entering a DSP, 43.9 percent now reaches consumers," a historical figure for that year's participants. TrueAdSpend divides by "total programmatic investment," per the ANA's Q4 2025 release (opens in a new tab) of February 25, 2026.
| Denominator | Who uses it | What sits outside it |
|---|---|---|
| Advertiser spend | ISBA and PwC, May 2020 | Nothing on the buy side; the agency fee is inside |
| Money entering a DSP | ANA 2024 benchmark | Agency fees and any costs paid before the DSP |
| Total programmatic investment | ANA TrueAdSpend, 2025 and 2026 releases | Whatever the participant does not route as programmatic spend |
| Platform Total Media Cost | DV360 reporting | Lines added at invoicing, such as surcharges and taxes |
What does the ANA TrueAdSpend index measure?
The ANA's Q4 2025 release defines TrueAdSpend as "the share of total programmatic investment that delivers impressions meeting preset quality criteria," listed as "fraud-free, measurable, viewable, and MFA-free." The latest public edition, the Q1 2026 release (opens in a new tab) of May 27, 2026, reports a market level TrueAdSpend Index; that figure, with its participant count, is set out on the programmatic supply chain page. The benchmark is built on log level data matched through TAG TrustNet and describes its participants, not the whole market.
The Q1 2026 release also splits participants into cohorts. Higher performing advertisers "converted 54.0 percent of programmatic spend into qualified impressions," while the lower performing cohort "converted just 32.1 percent." The higher group lost 19.0 percent of spend to media quality issues and the lower group 38.4 percent. These are cohort results within the benchmark, not targets for a single campaign. The previous edition, for Q4 2025, gave 56.7 percent for advertisers with disciplined quality governance and 37.5 percent for the lower performing group, though the two releases describe their cohorts in different words.
How should working media be reported?
Six parts of a working media line in a report
- Name the numerator. Publisher revenue, qualified impression spend or platform media cost, in those words.
- Name the denominator. Advertiser spend, money entering the DSP, or total programmatic investment, with the agency fee shown as inside or outside.
- State the scope. Channels, environments such as web, app and CTV, and whether deals and open auction are combined.
- Fix the period and currency. DV360 reports can show advertiser currency, partner currency or USD; a Canadian client's report should say which.
- Give the data source. Platform reports, matched log level data, or a benchmark; say whether seller side data was included.
- List what was excluded. Taxes, refunds, creative production or any line not in the base.
Before accepting a working media figure
- The report says which definition of working media it uses.
- The denominator is named, and agency fees are placed inside or outside it.
- Quality losses and fees are shown as separate lines.
- Benchmark figures are labeled as participant results, with their quarter.
- The currency and period match the invoices they are checked against.
Fees are one input to this report and are mapped layer by layer on the ad tech tax page. Quality losses are the other input, and they need verification data rather than contracts.
What does a working media report look like on a real plan?
Fictional example
A children's music school group in New Brunswick
An unnamed group of children's music schools in Moncton, Fredericton and Saint John ran English and French display in CAD on parenting, arts and local news sites ahead of fall term enrollment. Its agency's first report showed one working media percentage with no base.
The client asked for the six part version. The revised report named platform media cost as the numerator and money entering the DSP as the denominator, in advertiser currency, and showed the agency fee on a separate line. A second line used qualified impression spend from the verification data, so fees and quality losses sat side by side. No performance figures are implied.
How does PlacementGraph screen sites for made for ads risk before the buy?
Frequently asked questions
Is a higher working media ratio always better?
Not on its own. A ratio can rise because fees fell or because the base changed. Under a fee only definition it can also rise while more spend lands on unviewable or made for advertising pages. Read it with the quality line and the stated denominator.
Is creative production working or non working media?
The programmatic measures on this page, from ISBA and the ANA, are built from media spend data, so production sits outside them. If a client's budget does include it, the report should list it as an exclusion or as a separate non working line.
Can a buyer compare its ratio with the ANA benchmark?
Only with care. TrueAdSpend uses the ANA's quality criteria and its participants' data, matched at log level. A buyer's own ratio built from platform reports uses a different base and different measurement, so the two numbers describe different things even when both are percentages.
Does a curated deal raise working media?
It can change both parts of the ratio. A deal limited to chosen sites can reduce spend on low quality pages, while a curation fee adds a non working line. Report the curation fee separately so its effect on the ratio is visible.
Sources
- ANA Q4 2025 Programmatic Transparency Benchmark Shows Media Quality Is Now Driving Measurable Performance Outcomes (Feb. 25, 2026) (opens in a new tab)ANA
- ANA Programmatic Transparency Benchmark Reveals Widening Gap Between Top and Bottom Advertisers (May 27, 2026) (opens in a new tab)ANA
- ANA Releases 2024 Programmatic Benchmark Study (December 12, 2024) (opens in a new tab)ANA
- ISBA programmatic supply chain transparency study, executive summary (May 2020) (opens in a new tab)ISBA, in association with the AOP, carried out by PwC
- Metrics in reports (opens in a new tab)Google, Display & Video 360 Help
- Fees and billable cost (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.