RTBlintAlpha

Blog · CTV

European programmatic CTV can grow 44 percent while open auction samples stay STB-heavy, because most programmatic views clear as guaranteed deals.

· RTBlint

Trade coverage this week repeats the same headline: connected TV now accounts for 53 percent of premium video ad views in Europe, programmatic views grew 44 percent year on year, and more advertisers bought through automation. Those numbers come from a biannual ad-server report on first-half 2026 delivery across the United States and fourteen European countries. They describe impressions that already cleared inside one monetization stack. They do not, by themselves, describe the open RTB request your bidder sees when you buy Europe from a US-trained integration.

Programmatic growth is not one auction shape

The underlying report compares January through June 2026 with the same months of 2025. It defines an ad view as each time an ad is displayed, aggregated from the vendor platform. Total streaming ad views rose 12 percent in the US and 20 percent in Europe. Programmatic ad views rose faster: 25 percent in the US and 44 percent in Europe, roughly twice the pace of total delivery in each region. Unique advertisers delivering programmatically rose 25 percent across both regions combined, while ad views from those advertisers rose 24 percent.

Programmatic still represents a minority of all ad views on that stack: 34 percent in the US and 22 percent in Europe. The European programmatic market is smaller as a share of delivery but expanding faster in rate terms. That combination is easy to misread as deepening open competition. The same report splits programmatic delivery by deal type, and the split is where the bidstream story diverges from the press release.

Of programmatic ad views, 27 percent in the US were guaranteed deals. In Europe, 71 percent were guaranteed. The glossary classifies a deal as guaranteed when its deal type includes programmatic direct: a transaction priority that reserves supply by guaranteeing purchase of a set number of impressions. Everything else in programmatic counts as non-guaranteed. Europe programmatic growth is therefore dominated by a contract shape that often never touches the anonymous open auction your QA environment samples.

What guaranteed means on the wire

OpenRTB still carries guaranteed inventory as structured deals, not as a mystery flag on a generic impression. A buyer expecting open exchange traffic looks for an impression without a private marketplace block, or with an empty deals array. Guaranteed programmatic direct shows up as imp.pmp.deals[] entries with identifiers, floors, and auction types negotiated ahead of time. The bidder that wins is the one entitled to the deal id, not the highest unconstrained bid in a public pool.

"imp": [{
  "id": "1",
  "video": { "mimes": ["video/mp4"], "minduration": 5, "maxduration": 30 },
  "pmp": {
    "private_auction": 1,
    "deals": [{
      "id": "FW-EU-PG-88421",
      "bidfloor": 18.50,
      "bidfloorcur": "EUR",
      "at": 3
    }]
  }
}]

A request like this is programmatic in the ad-server report sense because automation connects buyer and seller. It is not the same diagnostic object as an open imp with no pmp block. When European broadcasters expand programmatic pipes, press examples include UK broadcasters opening video-on-demand inventory to multiple DSPs with programmatic guaranteed and private marketplace paths. That pattern matches a market where 71 percent of programmatic views clear as reserved supply rather than as leftover open slots.

Buyers who debug Europe using only open-exchange captures from US sellers import the wrong prior. Non-guaranteed programmatic in the US already skews heavily to CTV: 92 percent of US programmatic impressions on the same stack, with mobile at 6 percent and desktop at 2 percent, and set-top box VOD below half a percent. European programmatic on the same definition carries 39 percent on CTV, 28 percent on set-top box VOD, 25 percent on mobile, and 8 percent on desktop. More than a quarter of European programmatic impressions in that dataset run on a device category that barely registers in US programmatic.

Device mix breaks filters before price

Total delivery mix reinforces why a US bidder cannot transpose filters. Connected TV carried 86 percent of all ad views in the US on the platform, unchanged from the prior half-year. Europe recorded CTV at 53 percent of ad views, mobile at 24 percent, set-top box VOD at 16 percent, and desktop at 7 percent. CTV ad views grew 15 percent in the US and 33 percent in Europe, but the European headline share still leaves substantial STB and mobile delivery in the same monetization report that celebrates CTV majority.

OpenRTB expresses device class through device.devicetype and through app metadata on app.bundle. A buyer rule set that assumes devicetype=3 (connected TV) for every premium video imp will mis-route STB VOD and phone streaming that European sellers still sell as premium. The report itself notes European device usage is more evenly distributed. Enriched metadata percentages inside customized key-value pairs are computed only on the enriched slice of supply; Genre leads US enriched video keys at 66 percent, while Europe ranks ContentTitle and ContentLanguage above Genre at 50 and 48 percent versus 38 percent for Genre.

"device": { "devicetype": 7, "ua": "…", "geo": { "country": "DEU" } },
"app": { "bundle": "com.operator.stb.vod", "name": "…" },
"content": { "language": "de", "title": "…" }

A German STB VOD imp may validate as well-formed video while failing a buyer segment built for Samsung or Roku CTV bundles. That failure surfaces as no-bid or low participation long before anyone argues about creative. External analyses cited in the same coverage note that only about 40 percent of CTV bid requests in the wider market carry usable program-level content signals, which is a different denominator from enriched-field percentages inside the vendor slice. Both facts push the same direction: European programmatic growth can accelerate while the open requests you log still look sparse or unfamiliar compared with US CTV training data.

What completion rates do and do not prove

The report publishes average completion rates for pre-roll, mid-roll, and post-roll at 94, 98, and 91 percent across US and Europe combined, measured on delivery to Roku, Samsung TV, and Amazon Fire TV as the three most popular CTV sub-devices. Sellers read that as premium engagement. Skeptical buyers note that non-skippable in-stream formats push completion toward structural highs regardless of bidstream quality. The report does not define skipability or publish denominators for the completion sub-group.

Completion is a post-auction playback statistic on served ads. It cannot rescue a buyer who filtered valid European STB or mobile imps before bidding. Conversely, high completion on guaranteed deals does not tell an open-exchange buyer why their European line item under-delivers: the guaranteed 71 percent share may be clearing through deal ids your seat never received because the integration path was open-only.

RTBlint is independent and not affiliated with Comcast, FreeWheel, or the publishers cited in trade coverage. The linter checks whether a captured request matches OpenRTB and AdCOM expectations. It does not score fraud, viewability, or completion. It does catch missing pmp blocks where your trafficking brief promised a deal, inconsistent device and app fields, and video objects that will not parse on the bidder side.

Checklist for European CTV demand

  • Split logs by deal class: guaranteed programmatic direct (imp.pmp.deals with negotiated ids) versus non-guaranteed programmatic versus true open imps without pmp.
  • Compare device mix on European programmatic captures to US open CTV samples. Expect STB VOD and mobile shares far above US programmatic norms.
  • Do not treat headline CTV share of all ad views (53 percent in Europe) as the device mix inside programmatic or inside open auction subsets.
  • When metadata enrichment is partial, log whether targeting filters require Genre, series, or language fields that appear on less than half of enriched European video keys in vendor benchmarks.
  • Align sales promises on guaranteed reach with the deal ids actually emitted on imp.pmp.deals; open-exchange filters will not see reserved European inventory.
  • Where BidResponse.nbr is returned, bucket no-bids separately for invalid request versus unsupported device before blaming weak European demand.

Validate a recent European capture in the OpenRTB validator or read no-bid reason triage if participation dropped after expanding EU sellers. The fix is often deal and device segmentation, not another round of US CTV prospecting.

Sources