Why CTV Bid Requests Lose Program Data

A CTV bid request that contains no program information is not a malfunction. It is the norm. Peer39 analyzes more than 2.5 billion CTV bid requests daily, and a significant share arrive with nothing beyond a device identifier and an app bundle. No program title. No genre. No content rating. No description of what is actually playing at the moment the impression fires.

Buyers who assume the targeting data in their DSP reflects actual content context are working from an incomplete picture. Understanding why program data gets lost explains both why CTV targeting consistently underperforms expectations and what it takes to fix it.

What a complete CTV bid request should contain

The OpenRTB standard includes fields designed to carry content-level metadata: title, genre, content rating, series, episode, production quality, and live or on-demand status. When these fields are populated, contextual targeting, brand suitability controls, and content-aligned audience strategies can all operate at the level of precision they were designed for.

In a complete bid request, a buyer bidding on a primetime drama can confirm the program title, verify the rating, check the advisory flags, and know whether the stream is live or on-demand before placing a bid. The signal is there. The decision is informed.

In practice, most CTV bid requests do not contain these fields.

Where program data gets stripped from the supply chain

Program data has four opportunities to enter the bid stream and three points where it routinely disappears. Each represents a different kind of failure, with different economic and technical causes.

At the publisher source

The content metadata starts with the publisher. Streaming services know what is playing. The data exists. What does not always exist is a consistent, structured, incentive-aligned process to pass that data into the ad server for every impression.

Smaller publishers and ad-supported streaming services that rely on third-party ad servers often lack the engineering resources to maintain clean, complete content metadata pipelines. Program information may exist in their content management systems but never get mapped to OpenRTB content fields in a usable format. The data is there. It just does not make it to the bid request.

Larger publishers have a different problem: selective disclosure. Content metadata is commercially sensitive. A streaming service that reveals its exact program mix and viewership distribution through bid request metadata is giving supply-side intelligence to buyers who may use it to buy around certain content, target competitors, or negotiate differently. The economic incentive to omit or generalize metadata is real, even for publishers with the technical capability to include it.

At the SSP

Supply-side platforms aggregate inventory from multiple publishers and standardize bid requests before passing them upstream. In that standardization process, content fields that are inconsistently populated across publishers often get dropped rather than preserved.

An SSP that receives program data from some publishers but not others faces a choice: pass through inconsistent metadata that only covers part of the inventory, or normalize to the lowest common denominator and omit it entirely. Many normalize downward. The result is a cleaner bid request technically and an emptier one for the buyer.

SSPs also apply their own bid request compression to reduce latency and file size. Content object fields are frequently omitted in compressed bid requests because they add payload weight and because buyer demand for content-level bidding, until recently, was not strong enough to create market pressure to include them.

At the DSP

By the time a bid request reaches the buying platform, program data that survived the publisher and SSP stages can still be lost. DSPs may not surface all available content fields in their UI, may not pass them through to third-party data integrations, or may apply their own bid request normalization that drops fields not required for their core targeting logic.

Buyers working in DSP interfaces often do not know which content fields are present in the underlying bid request and which are absent. The gap between what the DSP shows and what the bid request contains is real and rarely documented.

Program data is not missing because CTV lacks content. It is missing because the supply chain has not been designed or incentivized to preserve it at every handoff.

Why the signal gap perpetuates

Each layer of the supply chain -- publisher, SSP, DSP -- has historically operated without strong market pressure to fix the metadata problem. Buyers have accepted signal gaps as a feature of CTV buying rather than demanding authentication. Publishers have had limited incentive to expose content intelligence. SSPs have optimized for speed and standardization, not data richness.

The result is a structural gap that cannot be closed by any single participant in the chain. Publishers cannot fix SSP normalization. SSPs cannot fix publisher disclosure incentives. DSPs cannot fix either.

What authenticated signal addresses

Filling the program data gap requires an independent authentication layer that does not rely on what publishers choose to include in bid requests or what SSPs choose to preserve.

Peer39 authenticates program-level content signals independently, operating across the bid stream without depending on publisher-declared metadata. Genre, sub-genre, content rating, advisory flags, and keyword-based content classifications are applied at the impression level before the bid fires, across buying platforms.

The authentication layer sits outside the publisher-SSP-DSP chain. It does not wait for publishers to disclose. It does not depend on SSPs to preserve. It does not rely on DSPs to surface. The signal is there at the bid, regardless of what the bid request itself contains.

Buyers who target CTV with the assumption that bid request metadata is complete are targeting on a fraction of available signal. The program data that makes contextual targeting, brand suitability, and content-aligned buying work in practice requires authentication, not declaration. The supply chain will not fix itself.