‘Rapidly becoming table stakes’: The state of agentic advertising
When hearing from industry leaders, one thing is clear: Agentic advertising is here to stay.
As Josh Lee, chief technical officer and head of product at EDO, put it: “Agentic AI delivery of ad intelligence is rapidly becoming table stakes across the buying and selling of media.”
Looking at the ecosystem, that may not be hyperbole. Every player — from DSPs to SSPs to agencies — seems to be developing or discussing some sort of agentic tool or strategy.
What’s less clear is how exactly the emerging tech is currently impacting advertising workflows.
“Right now, we’re mostly seeing companies in their very early tests of agentic, which will determine what their next steps will look like,” said Travis Clinger, chief connectivity and ecosystem officer and general manager at LiveRamp. “Companies need to feel comfortable learning how agents work and understanding the guardrails that exist, before they invest more in them.”
It’s been six months since the launch of Ad Context Protocol (AdCP) set in motion debates over the inevitability of agentic advertising; the arrival of IAB Tech Lab’s Agentic RTB Framework (ARTF) and then its agentic road map accelerated discussions.
While those initiatives have framed seemingly dueling ideologies, the reality is more nuanced. I heard from nearly a dozen executives across the advertising landscape about their observations and experiences with agentic advertising ahead of the Possible conference next week, where the topic will undoubtedly reach a fever pitch.
“Agentic advertising is still very much in its infancy, but the efficiency gains are already becoming clear,” said Karim Rayes, chief product officer at Nexxen.
The communication gap
First, there’s an elephant in the room: agentic advertising may be inevitable, but for many marketers, it still feels abstract.
Anthony Katsur, CEO of IAB Tech Lab, said that there’s a “tendency to overcomplicate” agentic advertising definitions, and that it’s just “systems executing and interacting across workflows.”
If it’s so simple, some would argue that the industry hasn’t done a great job explaining it.
“Most of the market still does not understand how agentic advertising works in practice,” wrote TeqBlaze CEO Nikita Bansal in an op-ed for The Current. “If the goal is faster adoption, then the industry needs to reduce recruitment-style messaging and increase practical, step-by-step explanations.”
Ali Manning, founder of Chalice, echoed a similar sentiment in a separate piece.
The conversation is “noisy because the industry is squabbling over standards, nearly every company started claiming agentic capabilities in a very short period of time without credible specifics about what’s changed and how it works, and most critically, barely anyone is putting it in terms of better consumer connections.”
What agents could solve — and what they shouldn’t
With that out of the way, what do industry leaders think agentic advertising could help solve? Removing friction. Accelerating workflows. Closing efficiency gaps. Interoperability. These were some of the common buzz phrases from those I asked.
“Eliminating inefficiency in the supply chain is the No. 1 problem to solve,” said Bill McLaughlin, senior vice president of advertiser solutions at PubMatic. “There’s still too much redundancy, too many hops and too much guesswork in how impressions are valued and transacted. Agentic systems can collapse that by identifying the most efficient path between demand and supply in real time, essentially making SPO truly dynamic and outcome-based.”
Others point to execution as the immediate win. Frans Vermeulen, president at Swivel, said that execution automation is where agentic’s value is right now, citing a case study with TelevisaUnivision that claimed to drive “significant year-over-year revenue” (it did not, though, provide any specific numbers in that regard).
“Teams across the industry know what they need to do, but execution speed and scale across fragmented platforms have become the bottleneck,” Vermeulen said.
Katsur, though, sees the most immediate value of agentic advertising even earlier in the process, removing the manual work before execution.
“The problem is everything that happens before execution: discovery, negotiation, setup and order management,” he said. “That’s where most of the friction is. Agentic should address that through shared protocols.”
The Tech Lab has started to realize its agentic goals for the industry after launching Agentic Advertising Management Protocols, the umbrella that its agentic initiatives live under; for instance, earlier this month PMG was one of the first agencies to integrate the Tech Lab’s existing protocols into its Alli Marketplace to build an agentic media buying model.
But, Katsur added, “what agentic should not rely on is probabilistic behavior or undefined logic. Advertising already has standards. This needs to build on them so outcomes are consistent.”
EDO’s Lee noted that agents also shouldn’t be relied upon to make brand-level judgment calls.
“Agentic systems optimize for the signals you give them,” he said. “If your inputs are wrong, your outputs will be wrong at scale and at speed.”
Quality data is the North Star
Most agreed that none of this works without strong data.
“There’s a painful gap between when a campaign goes live, when you get performance data and when you can actually act on it,” Lee said. “Agentic systems can close that loop in near real time … but only if the underlying data is great. Workflow automation without investment-grade data inputs is just moving fast in the wrong direction.”
Others concurred: Data is essential.
“The companies that win will be those that unify data, decisioning and execution, rather than layering AI on top of disconnected systems,” said Nexxen’s Rayes.
And that intelligence needs an anchor.
“If agents aren’t grounded in first-party data and brand context, it doesn’t matter how interoperable they are; they won’t produce meaningful results,” said Ian Maier, general manager and ad tech lead at Hightouch.
Subscribe to The Current
Early approaches give us an idea. EDO’s agentic AI interface, ChatEDO, aims to more quickly surface insights on its TV outcomes data for advertisers, and U.S. agencies have free access to it starting this week ahead of the TV upfronts. The Trade Desk announced Koa Agents in alpha this week, which Stagwell plans to pilot.
“AI is only as good as the data beneath it. Inside large-scale, objective platforms, AI can evaluate the full context of media buying — audiences, inventory, performance and measurement — against a marketer’s goals,” Aravind Chandrasekharan, senior vice president of engineering at The Trade Desk, wrote in an op-ed for The Current.
“An objective platform, operating across the open internet, can evaluate every opportunity on its merits. In a world where AI drives decision-making, that difference becomes structural, not philosophical.”
‘Scale with proof’
Of course, with great data there must also come great trust.
“If agents are grounded in a company’s first-party data, brand context and real performance signals, then marketers should be comfortable giving them a high degree of autonomy,” Maier said. “If they’re not, then that’s where the hesitation should be. That’s the standard the industry needs to adopt: Trust the system only as much as you trust the data underneath it.”
The industry is trying to build that trust through standards, protocols and governance frameworks. “We are now in a critical period where we can collectively build the standards that will ensure governance, privacy and trust are built into agentic from the ground up, helping agentic activity to scale faster and to allow us to deliver on AI’s transformative potential,” LiveRamp’s Clinger said.
Beyond that, trust in agents can be based on one question, according to Lee: What combination of data structure and agent communication actually moves the brand’s business?
“Every marketer’s comfort should scale with proof,” Lee said. “If an agent can’t show its reasoning, you shouldn’t be handing it autonomous control. That line, for advertisers, is accountability that comes from immediate, investment-grade outcomes signals.
“Trust is earned incrementally, not granted upfront.”
The Current is owned and operated by The Trade Desk, Inc.