OpenX can put half of CTV deal spend in supply-side targeting while most bid requests still lack show-level content.

Two ledgers, one impression

Deal spend shares and bid-request observability are not the same instrument. OpenX's report draws on what it calls the complete deal-level transaction record of its marketplace, classified under a taxonomy OpenX wrote. PPC Land's November 1 article quotes the rule that matters for mixed packages: when a deal layers site lists, third-party segments, and content metadata, OpenX assigns the entire deal to whichever signal it ranks as most advanced, because there is no practical way to split credit by contribution.

That is rational for an SSP selling supply-side intelligence. It is awkward for a buyer engineer trying to explain why a DSP filter on content.title or content.genre still returns empty on production CTV traffic while vendor slides claim half of CTV dollars moved to supply-side targeting. The report itself places show, genre, and rating selection under quality within supply-side targeting, not under curation-only inventory bundling. PPC Land notes that if those content-object deals had been counted as curation, the CTV supply-side targeting figure, and much of the growth since 2024, would look materially different, and OpenX does not publish the same charts under an alternative classification.

OpenX's October 15 press release emphasized four figures: supply-side targeting share rising from 11% of spend in the fourth quarter of 2023 to 37% two years later, 17% of 2025 deals including outcome measurement, 500% growth in CTV spend using supply-side targeting from the first half of 2022 to the first half of 2025, and roughly 11% average quarterly growth in the U.S. and EMEA since early 2025. PPC Land's follow-on work adds the CTV reversal detail the press release skipped: supply-side targeting's share of CTV deal spend peaked near 56% in the third quarter of 2025 and fell to about 47% in the fourth quarter, while curation's CTV share rose in the same window. Non-CTV deal types kept moving the other direction. None of that quarter-to-quarter motion appears in a single annual share printed on a landing page.

What the 51% actually measures

The 51% figure is a share of deal spend on OpenX's U.S. marketplace, not of open exchange impressions across every SSP, and the report still omits dollar volume. PPC Land walks through the internal inconsistency: executive summary language near 32%, trend chapter at 37% for the fourth quarter of 2025, closing copy near 40%, and a CTV paragraph that once prints 53% where elsewhere 51% appears three times. Readers reconciling vendor charts to finance models need to know which cut they are looking at before they argue with their own fill logs.

Inside supply-side targeting, the mix shifted as fast as the top-line share. Third-party audience data accounted for about 78% of supply-side targeting spend in 2022 and roughly 40% in 2025, while quality rose to about 35% of U.S. supply-side targeting spend, driven overwhelmingly by connected TV, according to PPC Land's summary of the report. A falling share is not falling spend: PPC Land compounds OpenX's published growth rates for supply-side targeting deal spend and estimates third-party audience dollars on OpenX still grew on the order of fifteenfold between 2022 and 2025 even as quality's share climbed. Headlines about third-party data dying on the supply path can coexist with more absolute segment money on the same platform.

Qualitative claims rest on eight one-on-one interviews with senior buy-side executives, seven from agencies and one from a Fortune 50 brand, all described as OpenX clients, with research carried out by W Media Research in partnership with OpenX. That context does not invalidate the transaction record, but it does bound how far to treat the narrative as market-wide truth versus a well-documented story about one independent SSP's deal book. Jounce Media's September 2026 supply coverage ranking, cited in the report, is part of OpenX's scale argument, not a third-party audit of the taxonomy.

Content objects on the wire

OpenX's explanation for connected TV acceleration is structural: streaming lacks browser cookies, device identifiers fragment, and two impressions from the same app can differ sharply by show and rating. Content objects, metadata describing title, genre, episode, and rating, are the lever the report credits for quality supply-side targeting growth. PPC Land ties that story to industry plumbing: IAB Tech Lab's OpenRTB Content Object and Content Taxonomy 3.1, announced in December 2024 with a CTV genre subset, and OpenX's claim that it normalizes genre variants into nine categories and 132 subcategories on its side of the auction.

The wider market did not wait for one SSP deck. PPC Land cites Peer39 finding in March 2026 that only 40% of CTV bid requests carry usable program-level signals, and notes September 2026 moves from Wurl writing content signals into requests and Pixalate shipping a pre-bid API that adds show titles, alongside buy-side show-level data partnerships. Those are patches on a sparse field, not proof that half of industry CTV spend already travels with rich content blocks your bidder can require without throttling yourself out of the auction.

When a deal is classified as supply-side targeting because genre and rating filters sit in the SSP product, the OpenRTB request your exchange forwards may still carry an empty content object or a bundle-only identity that downstream EPG splicing tries to repair. We wrote separately about bundle-only show ranking and about pre-bid APIs that inject titles after the fact. OpenX's 51% share describes how buyers bought through OpenX deal products, not how often content.title arrived populated on the path you log in Splunk.

What to check before you re-traffic filters

Supply-side targeting growth is a real product story: buyers consolidating paths and applying filters closer to publishers is compatible with privacy pressure and with connected TV's share of deal spend on OpenX rising past half the marketplace in 2025, per PPC Land's reading of the report charts. The mistake is importing the SSP's deal taxonomy into OpenRTB validation rules as if the two were already aligned.

• Separate deal reporting from request shape: track what percentage of your cleared CTV impressions include non-empty content.title, content.genre, or taxonomy fields you actually bid on, independent of vendor supply-side targeting spend charts.
• Read mixed deals skeptically: a package that combines inventory lists with a segment may be counted entirely as supply-side targeting upstream while your bidder still sees only the bundle and app identifiers on the wire.
• Watch quarterly CTV reversals: if supply-side targeting share fell from roughly 56% to 47% on connected TV in the fourth quarter of 2025 on OpenX while non-CTV supply-side targeting kept climbing, your own seasonality may not match a single annual headline.
• Compare EMEA lag honestly: the report puts EMEA supply-side targeting near 23% against 37% in the U.S. at year-end 2025, roughly two to three quarters behind in narrative terms, with choppier quarterly charts than the summary implies.
• Validate before hard-blocking: paste production connected TV requests into the bid request tester after any policy change driven by supply-side intelligence marketing, so blocks target missing or inconsistent fields, not vendor classification alone.

Vocabulary fights are bidstream fights

OpenX asks the industry to reserve curation for site or inventory bundling and treat audience, quality, measurement, and optimization signals as supply-side targeting. Germany's BVDW and the IAB Tech Lab Curation Framework use broader definitions that would sweep much of what OpenX labels supply-side targeting back under curation. PPC Land surfaces that tension because it changes whether a product is dismissed as bundling or sold as intelligence. Buyers do not get to pick one dictionary in the SSP contract and another in Prebid schema validation without creating false arguments between ad ops and engineering.

The report also forecast that supply-side targeting spend share would exceed curation-only share by mid-2026, published after that date without 2026 figures to confirm the crossover. PPC Land notes the arithmetic would have required a steeper climb in early 2026 than the roughly two-point quarterly moves visible in late 2025. Forecasts in vendor research are not wire standards; they are sales timing. Your NBR and filter logs still win any dispute about why demand passed.

Sources

Deal share figures, classification rules, and interview descriptions are from PPC Land's November 1, 2026 article summarizing OpenX's State Of Supply-Side Intelligence report, with headline metrics also stated on OpenX's October 15, 2026 press release and research landing page. OpenRTB field names follow IAB Tech Lab conventions. RTBlint is independent and not affiliated with IAB, IAB Tech Lab, or OpenX.

• PPC Land: Supply-side targeting gains 51% of OpenX CTV spend, November 1, 2026
• OpenX: Supply-Side Targeting press release, October 15, 2026
• OpenX: Supply-Side Intelligence Trend Report 2026 landing page
• RTBlint: Bundle ID resolves show-level CTV spend
• Pixellint: Meta on-device conversions are not collector payload