AI agents in advertising

In this episode of Datalicious, Carsten Sander, David Porzelt (Teavaro), and Markus Caspari (Dentsu/iProspect) discuss agentic advertising – the use of AI agents in digital media buying. They examine two competing approaches: The Ad Context Protocol (AdCP) as a revolutionary new start with a focus on transparency, and the IAB Tech Lab Agentic RTB Framework as an evolution of existing standards. This episode highlights the technical, economic, and cultural challenges facing the industry – and why it is worth acting now, despite the high risks.

Topics discussed in this episode:

1. The basics of agentic advertising and Ad Context Protocol (AdCP): What does using AI agents in media buying actually entail, how does AdCP work as a new standard, and what role do the MCP and A2A protocols play in this?
2. IAB Tech Lab vs. AdCP – evolution or revolution: What different approaches are pursued by the IAB Tech Lab with its Agentic RTB Framework and the AdCP initiative? Are they really in competition or do they complement each other?
3. Market power, adoption, and the open-web dilemma: Why are major players missing from AdCP, what economic interests are behind this, and how can the open web compete with walled gardens?
4. Practical implementation and challenges: What initial use cases and tests already exist, what technical and operational hurdles need to be overcome, and how can publishers, agencies, and advertisers get started?

Takeaways from this episode

1. Agentic advertising is more than automation: While AI agents enable more efficient media buying, they also create new forms of interaction between publishers, agencies, and advertisers.
2. The two standards take different approaches: AdCP focuses on transparency, human-in-the-loop, and a fresh start, while IAB Tech Lab is an evolution of existing standards with 80% latency reduction. They can coexist.
3. Market power determines success: Despite the technical advantages of open-source approaches such as AdCP, major players (The Trade Desk, Amazon) decisively influence the adoption of new standards due to their market dominance.
4. Act now – don’t wait: Even if some of the projects will fail, those who do not engage with agentic AI risk significantly higher opportunity costs. Early adopters actively help shape the future.

Chapters

00:00 – Introduction: Agentic advertising and transformation of the industry

03:45 – Presenting the guests: David Porzelt (Teavaro) and Markus Caspari (Dentsu/iProspect)

09:43 – The relevance of agentic AI within agencies and for customers

13:10 – Gartner prediction: 40% of all projects will fail – why invest anyway?

18:01 – Ad Context Protocol (AdCP) in a nutshell: technology, objectives, and founding members

22:16 – Transparency and auditability: the key promise of AdCP

26:33 – Google’s Universal Commerce Protocol and the role of Gemini

28:10 – Carsten Sander’s case report: first experiences with AdCP implementation

34:23 – The chicken-or-egg problem: adoption by publishers and media agencies

42:54 – Latency reduction and server-to-server communication: the IAB approach

48:30 – Evolution vs. revolution: comparing the approaches

55:45 – Time machine: outlook on the world of digital advertising in three years