Prior to the digital age, marketers who invested in paid media strategies were oftentimes very “data poor.” The market was very reliant on traditional research methods, which relied on panel-based data that lacked a lot of granularity & speed. As a result, the value of media, especially across partners or channels, was hard to determine. Traditional research methods still exist today, and are important part of any marketers strategy, but now we exist in a world where there is a plethora of granular digital data that can drastically help us better understand who our customers are, where are the best places to reach them, and how can we persuade them to transact. This is where the newly introduced Ad Context Protocol (AdCP) can shine. Imagine marketers being able to explore rich audience definitions, browse contextual signals aligned with targets, or request custom audience packages from the world’s leading digital publishers all before actually setting up a campaign.
The use of most granular digital data faces one fundamental challenge. For most marketers the primary method for accessing and utilizing this data has been through participating in impression-based auctions. In most digital environments this is done through programmatic advertising, which by some estimates makes up 96% of digital advertising. This is also true for major platforms such as Meta and Google who have dominated digital advertising over the last decade.
The challenge with this dynamic is that it makes it extremely hard to truly value media. For example, a marketer wanting to reach “frequent grocery shoppers in the Northeast” must currently bid blind into auctions for impressions or use opaque 3rd party data, instead of querying sellers for their specific first party segments, retail signals, or custom packages beforehand. Marketers oftentimes have to participate in the media buying process before really gaining access to any insights which ties the customer segmentation & planning process too closely to media activation and removes a key part of the process. As a result, many marketers have become “addicted” to the free insights given by these major platforms or struggle to navigate the complexities of developing an impression-bidding strategy that can help them effectively value the plethora of data across the rest of the digital ecosystem.
However, this is changing and the driving factor is the birth of agentic advertising, powered by AI-infrastructure like LLMs and MCP servers. More specifically, a recently developed protocol call for agentic advertising, AdCP was recently introduced to help both media buyers and advertisers enhance their existing programmatic strategies through the use of AI and bring a new wave of innovation into the programmatic ecosystem. This protocol promises to act as a “universal API” for the advertising ecosystem and we see it helping to advance the already strong foundation built by standards such as OpenRTB, Prebid, and others. In the future marketers will be able to “ask” for what they want across our entire ecosystem (e.g. “I need to reach shoppers of EV vehicles who are in-market and have a household income over X”) and publishers and media sellers can respond with segments, packages, or outcomes they can deliver.
AdCP is well positioned to help solve the “everything, everywhere, all at once” problem facing many modern marketers. It is often challenging for marketers to truly leverage the rich digital data sets of their media partners because they are too complex to utilize through the 100ms auction process. This simplicity is what has allowed platforms like Meta and Google to capture the majority of digital advertising growth.
AdCP creates value through collaborative insights and ad product discovery between marketers and their media partners. AI is uniquely suited to support this process, enabling marketers to work with dozens or even hundreds of media sellers, parse their very different digital data sets and ad product catalogs, and find the right partners to not only build a media plan but also share insights that enhance their overall understanding of their customers. This approach, known as Agentic Collaboration, allows AI agents to scale campaign activity across the ecosystem on tasks such as identifying “audiences who showed purchase intent for home improvement products in the last 30 days” to sourcing “all available high-impact placements around basketball content during Q4”.
The future points to deeper collaboration between existing programmatic companies and emerging agentic advertising solutions to create more effective ad products. The industry has not seen a wave of innovation quite like the one driven by AI and agentic advertising in many years, and progress across dozens of companies is already demonstrating the potential to deliver stronger results for brands and agencies.
Original publication source and date Dec 11, 2025: https://www.ana.net/miccontent/show/id/ii-2025-12-auctions-agents
.png)
.png)
