Publishers Face Agentic AI Shift Without New Integration Costs
Agentic AI may change how media is bought, with buyers already paying for necessary integrations, according to AdExchanger. Publishers risk missing auctions if their inventory is not evaluated as worthy of a bid.
Original reporting · adexchanger.com
Original: “Publishers: Agentic AI Isn’t About Pricing, But Whether You Get Any Bids At All”
Read “Publishers: Agentic AI Isn’t About Pricing, But Whether You Get Any Bids At All” at AdExchanger →
How this was written. CTCV wrote this article with a language model from AdExchanger’s reporting (model: @cf/meta/llama-3.3-70b-instruct-fp8-fast); it describes that reporting rather than adding to it. The original, linked above, is the full story.
Key points from AdExchanger’s report
- Agentic AI doesn’t require new integration or industrywide standards, as buyers have already paid for the necessary infrastructure.
- Publishers risk missing auctions if their inventory is not evaluated as worthy of a bid.
- The work to prepare for agentic AI is the same as it has always been, with steps including creating stable identifiers and structuring first-party attributes.
Anthony Katsur, who runs IAB Tech Lab, expects several false starts and years of experimentation with agentic AI, AdExchanger reported. The mood around agentic AI is considered inevitable but incremental. Digiday’s research found that 54% of brand and agency professionals say their companies don’t use agentic AI in their workflows.
Agentic AI is different from previous industry shifts, as it doesn’t require new integration or industrywide standards, and buyers have already paid for the necessary infrastructure. This means that agentic AI can start evaluating inventory and informing bidding decisions without new pipes or standards.
Money is already moving, with Microsoft ending its relationship with Dentsu and moving its global media planning and buying to Publicis, with a price tag of between $700 million and $1.2 billion. Publicis also agreed to acquire LiveRamp, which will give it more control over the identity layer that its agents use to reason over.
The industry’s experience with bid shading shows how quickly new buying logic can reshape auctions, with exchanges moving to first-price auctions in 2017 and bid shading spreading across SSPs and DSPs within months. Publishers were left to understand bid shading’s impact and adjust their pricing floors accordingly.
Agentic logic is divided into two halves, with AI-powered decision-making and autonomous execution, according to IAB research. While current adoption is centered on decision support, 63% of video buyers see a role for agentic AI in programmatic and RTB, and 28% are using AI-based deal execution and negotiation for video campaigns.
Publishers risk losing opportunities if their inventory is not evaluated as worthy of a bid, and the work to prepare for agentic AI is the same as it has always been, with steps including creating stable identifiers, structuring first-party attributes, and authorizing data access explicitly.
Questions and answers
What is agentic AI and how does it differ from previous industry shifts?
Agentic AI is different from previous industry shifts, as it doesn’t require new integration or industrywide standards, and buyers have already paid for the necessary infrastructure.
How can publishers prepare for agentic AI?
Publishers can prepare for agentic AI by creating stable identifiers, structuring first-party attributes, and authorizing data access explicitly.
What is the risk for publishers if they don’t prepare for agentic AI?
Publishers risk losing opportunities if their inventory is not evaluated as worthy of a bid, and may miss auctions entirely.
