The Other Proofs: Signals, Sellers, and Supply Paths
The CTV measurement problem does not end when a creative has a persistent identity. This chapter compares the other signals in the CTV supply chain—content classification, seller authorization, supply-path transparency, device authenticity, render evidence, authentication, and disclosure—by the question each one can answer. The practical goal is to stop asking ACIF to prove where an ad was sold or asking a seller file to prove that an ad rendered.
Different Identifiers Answer Different Questions
A CTV transaction contains several kinds of identity, and they are easy to confuse because they may appear in the same bid request, ad response, or report. The first is creative identity: ACIF and the Universal AdID are meant to identify the ad asset and connect it to registry metadata. The second is content identity: a genre or content-taxonomy signal describes the program environment in which an ad opportunity appears. The third concerns the selling path: authorization files and supply-chain records describe which businesses may sell or resell the inventory. Device signals address whether the device information itself can be trusted. Measurement signals address what happened at playback.
These are not competing answers to one question. They are answers to different questions about the same opportunity:
| Signal or mechanism | Primary question it answers | What it does not prove |
|---|---|---|
| ACIF / Universal AdID | Which registered creative is this? | That the ad rendered or the traffic was valid |
| Content taxonomy or genre identifier | What kind of program environment is this? | That the advertiser’s creative was suitable or viewable |
ads.txt / app-ads.txt and INVENTORYPARTNERDOMAIN |
Which sellers are authorized, including an inventory-sharing partner? | That an authorized seller delivered a valid impression |
sellers.json and SupplyChain Object |
Which sellers and intermediaries participated in the path? | That every listed node behaved correctly or that the ad rendered |
| Device attestation | Is the device information authentic? | Which creative was shown or whether a person watched it |
| OM SDK or another render integration | What standardized viewing or impression signals are available? | That the asset has a persistent identity or that all environments are covered |
The table is useful only if its rows remain separate. A buyer who knows the creative ID can consolidate delivery by asset, but still may not know whether the inventory came through an authorized reseller. A buyer who sees a valid supply path may still not know whether the ad began to render. A buyer who receives device information may still lack evidence about the physical television’s power state or the number of people co-viewing.
Content classification illustrates the distinction. The WFA guide points to standardized content genres and extended content identifiers that can be included in a bid request. A publisher can use those signals to describe the program without relying on a title in the bidstream, and a buyer can use the classification when deciding whether the context is suitable for a brand. That is a property of the content environment, not of the ad creative. It complements creative identity because a decision system may need both: which ad is being considered, and alongside what kind of program it may appear.
Seller authorization answers another question. In CTV, multiple entities may have rights to sell ad slots within the same streaming service or application.