Working Media vs Non-Working Media: definitions, the ANA benchmark and the CTV cut

Table sources. ISBA cost categories and DSP fee: ISBA and PwC, May 2020. CMA scope: CMA, Appendix R, 1 July 2020. ANA waterfall lines and exclusions: ANA Benchmark, Q3 2025. Unknown delta in 2023: ANA, December 2023.

Read across the rows and two differences explain most of the gap between the studies. The ANA leaves agency fees out, which raises its number. It also deducts wasted impressions, which lowers it far more. ISBA does the opposite on both counts, so its 51% and the ANA's 45.1% measure different things even before the five years between them. The ad tech tax explained page walks through each hop of that chain.

The ANA ran its first full study in 2023. It covered 21 advertisers, $123 million of spend and 35.5 billion impressions, and defined a TrueImpression as one that is non-invalid, measurable and viewable under MRC standards (ANA, December 2023). In June 2024 the ANA, TAG TrustNet and Fiducia turned it into a quarterly benchmark built from advertisers' matched log-level data (TAG TrustNet, 17 June 2024). The quarterly releases also count spend on made-for-advertising (MFA) sites as waste (ANA, 27 May 2026).

The headline readings, with the CTV share of each sample, are below. The full quarterly table, with sample sizes, spend and seller revenue, lives in the working media benchmark tracker.

The 2023 study covered only web and mobile. When connected TV entered the data in full, the headline fell from 43.9% to 41.0%. The ANA put the drop down to CTV, with non-measurable spend up 3.8 points and non-viewable spend up 3.5 points (ANA Benchmark, Q1 2025). Transaction costs fell 3.7 points in the same report, which was not enough to offset the CTV effect.

CTV is not low-quality supply. It is supply that web verification tools struggle to measure. In Q3 2025 the median CTV advertiser kept 71.9% of spend as TrueAdSpend, close to 73.1% for web and mobile (ANA Benchmark, Q3 2025). Median CTV transaction costs were 21.0%, and median CTV seller revenue was 79.0%. Those medians are calculated differently from the headline waterfall, which uses sequential averages, so compare them only with each other.

Price is the other CTV factor. The overall CTV CPM was $16.91 in Q3 2025, against $5.43 for web and mobile, and nearly all CTV spend ran through private marketplaces (ANA Benchmark, Q3 2025). The ANA links CTV's higher non-viewable rates to server-side ad insertion (SSAI) and limited signals (Fiducia on ANA, 25 February 2026). If your plan is CTV-heavy, compare yourself with the all-environment series. If it is web and mobile only, use the web-and-mobile series. Vendors that measure CTV are compared in CTV measurement companies.

Suppose an advertiser spends $1 million through a DSP. Under ISBA's definition, about $510,000 reaches publishers, because 51% did in the 2020 study (ISBA and PwC, May 2020). Under the CMA's estimate, about $650,000 does, before trading desk, data and verification fees.

The ANA gives two answers from one waterfall. Seller revenue was 76.2% of spend in Q3 2025, so about $762,000 would reach sellers (ANA Benchmark, Q3 2025). TrueAdSpend was 45.1% in Q2 2026, so about $451,000 would buy impressions that pass every quality test. The remaining $549,000 is non-working by the ANA's definition: $272,000 in transaction costs at the Q2 2026 rate of 27.2%, and the rest in media quality losses (Fiducia on ANA, 6 August 2026).

The same million, read four ways, gives $451,000 to $762,000 of "working media". That spread is the reason to state the definition first. To model your own fee lines, use the ad tech fee calculator.

Four levers show up in the ANA data, and each sits in a different team. The first is supply-side fees. The SSP fee line fell from 14.7% to 12.8% of spend between Q2 and Q3 2025, which the ANA linked to supply path optimization (ANA Benchmark, Q3 2025). The second is viewability. Non-viewable spend dropped from 13.3% to 10.1% in Q2 2026 and drove the record reading (Fiducia on ANA, 6 August 2026).

The third is supply concentration. Higher performers in Q2 2026 bought across about 19,000 fewer domains and apps than lower performers (Fiducia on ANA, 6 August 2026). Kimberly-Clark, named in the same release, cut its cost per quality impression by 31.4% through impression-level optimisation (company case, Fiducia on ANA, 6 August 2026).

The fourth lever is new waste. A Q1 2026 analysis found AI-generated "slop" sites took 1.3% to 2.4% of open-web programmatic spend, about the same as MFA (Fiducia on ANA, 10 August 2026). Those impressions often pass viewability and fraud checks, so they can count as working media while adding nothing.

1. Name the definition. Say whether you mean money reaching publishers or money buying quality impressions.
2. List what is excluded. Agency, ad serving and managed service fees sit outside the ANA waterfall; say whether yours includes them.
3. Split fees from waste. Transaction costs and media quality losses need different fixes: contracts for the first, supply and measurement for the second.
4. Show CTV separately. Report all-environment and web-and-mobile figures side by side.
5. Compare like with like. Use the same quarter of the ANA series and the same cohort logic. An audit of agency buys follows the same steps; see how to audit your agency's programmatic buys.

At budget level the stakes look smaller but are not. Ad Age, citing the ANA's 2024 work, estimated that open-web programmatic is about 10% of industry media spend, with DSPs and SSPs taking roughly 30% of it (Fiducia on Ad Age, 22 September 2025). That is about 3% of all media spend going to programmatic intermediaries. Fee benchmarks by platform are in DSP pricing explained, and the economics hub is in-house programmatic.

1. ISBA and PwC, "Programmatic Supply Chain Transparency Study, executive summary" (May 2020). https://www.isba.org.uk/system/files/media/documents/2020-12/executive-summary-programmatic-supply-chain-transparency-study.pdf
2. UK Competition and Markets Authority, "Appendix R: fees in the adtech stack", Online platforms and digital advertising market study (1 July 2020). https://assets.publishing.service.gov.uk/media/5fe49625e90e071207e10eff/Appendix_R_-_fees_in_the_adtech_stack_WEB.pdf
3. ANA, "Programmatic Media Supply Chain Transparency Study: Complete Report" (December 2023). https://www.adslot.com/wp-content/uploads/2023/12/ana-programmatic-media-supply-chain-transparency-study.pdf
4. ANA and TAG TrustNet, "Programmatic Transparency Benchmark: Q1 2025 Benchmark Findings" (2025). https://s3.amazonaws.com/media.mediapost.com/uploads/ANAQ12025Programmatic.pdf
5. ANA and TAG TrustNet, "Programmatic Transparency Benchmark: Q3 2025 Benchmark Findings" (November 2025). https://s3.amazonaws.com/media.mediapost.com/uploads/Q32025ProgrammaticTransparencyBenchmark.pdf
6. ANA, "ANA Introduces Brand Safety & Suitability Metrics, Achieves 99.1% Low-Risk Programmatic Spend" (5 November 2025). https://www.ana.net/content/show/id/pr-2025-11-transparency
7. Fiducia, "ANA Q4 2025 Programmatic Transparency Benchmark Shows Media Quality Is Now Driving Measurable Performance Outcomes" (25 February 2026). https://www.fiducia.eco/post/ana-q4-2025-programmatic-transparency-benchmark-shows-media-quality-is-now-driving-measurable-perfor
8. ANA, "ANA Programmatic Transparency Benchmark Reveals Widening Gap Between Top and Bottom Advertisers" (27 May 2026). https://www.ana.net/content/show/id/pr-2026-05-programmatic
9. Fiducia, "ANA Programmatic Transparency Benchmark: Q2 2026 Working-Media Efficiency Reaches a Record 45%" (6 August 2026). https://www.fiducia.eco/post/ana-programmatic-transparency-benchmark-q2-2026-working-media-efficiency-reaches-a-record-45
10. Fiducia, "Procter & Gamble, Bayer scrutinize ad tech as brands demand accountability and lower costs" (22 September 2025, excerpt of Ad Age). https://www.fiducia.eco/post/procter-gamble-bayer-scrutinize-ad-tech-as-brands-demand-accountability-and-lower-costs