Automation in digital ad sales is supposed to cut out middlemen. But agentic advertising may simply rebuild the same layers of complexity and bias that publishers faced before.
Publishers looking to streamline digital ad sales with agentic advertising are confronting a familiar dilemma. While the technology promises to automate planning, packaging, forecasting, proposals, and campaign execution, the reality is more complicated. Instead of removing intermediaries, agentic advertising may be reconstructing the same layers of complexity and bias that defined the programmatic era.
Historically, publishers adopted dozens of SSPs, each offering incremental demand but also introducing duplication, operational headaches, and margin pressure. Many publishers now work with more than 25 SSPs, often losing sight of the cumulative costs and inefficiencies. The rise of agentic selling is showing similar patterns. Direct publisher-managed sales agents, SSP-powered agents, and audience-driven agents from data platforms all promise value, but together they risk recreating the very inefficiencies they were meant to solve.
Buyer incentives play a central role in this cycle. Publishers did not add multiple SSPs by accident; buyers rewarded this behavior by bidding across many paths, driving incremental revenue for each new integration. In agentic environments, buyer agents are programmed to optimize for campaign performance, efficiency, or platform economics. These agents may favor supply paths that align with their objectives, often prioritizing their own first-party sales agents over third-party options. This can reintroduce bias, with spend flowing to the most incentivized supply rather than the highest-performing inventory. As a result, publishers may feel compelled to deploy numerous agents to maximize revenue, echoing the SSP sprawl of the past.
Not all sales agents serve the same function. Some extend distribution, while others shape how inventory is packaged and valued. For example, a custom content agent might focus on matching publishers with content and sponsorships, while a programmatic agent could handle pricing, yield, and scalable executions. A more strategic approach involves defining a primary sales agent to manage inventory, audiences, pricing, and packaging, then building a few specialized agents for key execution types. Standardizing payment paths and consolidating clearing house partners can help maintain clarity and leverage, while selective participation in external ecosystems ensures incremental value without redundant representation.
The economics of agentic advertising are shifting as companies seek to control both buyer and sales agents within their ecosystems. Control over settlement, reporting, and payment flows may ultimately matter more than who builds the sales agent. Platforms such as Scope3, PubMatic, Magnite, and Triplelift are positioning themselves as clearinghouses, collecting spend from buyers and distributing payments to publishers. Publishers are encouraged to ask these platforms how they plan to support third-party sales agents and prevent supply path bias.
After a decade of expanding distribution through multiple intermediaries, the industry now faces a reset. Agentic advertising can deliver on its promise only if buyers and sellers avoid duplicating the same layers of bias and complexity. Publishers should deploy specialized sales agents based on product and buying needs, not for every SSP or integration. Agencies and brands are advised to build their own buyer agents to avoid redundant supply paths. The real test is whether agentic advertising can truly eliminate the duplication and intermediation it was designed to solve.
For more on how publishers are experimenting with AI-powered ad formats and agentic media, see this report on publishers confronting AI revenue questions at Cannes.
