Butler/Till and Pubmatic Test Agentic AI to Slash Media Supply Chain Costs
Key Takeaways
- Independent agency Butler/Till and supply-side platform Pubmatic have successfully completed initial tests of agentic media buying, demonstrating significant reductions in both media and supply chain costs.
- The shift from rules-based programmatic to autonomous, goal-oriented AI agents marks a pivotal evolution in digital advertising efficiency.
Key Intelligence
Key Facts
- 1Initial tests by Butler/Till and Pubmatic showed significant reductions in media and supply chain costs.
- 2The shift moves media buying from rules-based programmatic to goal-oriented autonomous AI agents.
- 3Pubmatic reported a $0.13 EPS beat in Q4 2025, driven by accelerating agentic AI adoption.
- 4The agency aims to reduce the 'ad-tech tax' by using agents to find the most efficient supply paths.
- 5Pubmatic is targeting double-digit revenue growth in H2 2026 through its agentic AI strategy.
| Feature | ||
|---|---|---|
| Decision Logic | Manual rules & triggers | Autonomous, goal-based reasoning |
| Optimization | Human-led periodic adjustments | Real-time, continuous self-correction |
| Supply Path | Often fragmented/multi-layered | Direct, cost-optimized routing |
| Primary Benefit | Automation of manual tasks | Strategic efficiency & cost reduction |
Who's Affected
Analysis
The digital advertising landscape is witnessing a fundamental shift from automated to autonomous operations. Independent agency Butler/Till, in partnership with Pubmatic, has released results from its first 'agentic' media buying tests, signaling a departure from the traditional programmatic model that has dominated the industry for over a decade. While programmatic advertising relies on pre-defined rules and manual optimization, agentic AI utilizes autonomous agents capable of making real-time decisions to meet high-level campaign objectives. The initial results are stark: a measurable reduction in media costs and a streamlined supply chain that eliminates traditional inefficiencies.
This development comes at a critical time for the ad-tech industry, which has faced increasing scrutiny over 'made-for-advertising' (MFA) sites and opaque supply chains that siphon off significant portions of advertiser budgets. By deploying agentic AI, Butler/Till is essentially giving the machine a goal—such as maximizing reach within a specific audience segment at the lowest possible cost—and allowing it to navigate the complex web of SSPs and DSPs autonomously. This reduces the 'ad-tech tax' by identifying the most direct and cost-effective paths to premium inventory, effectively bypassing the bloated middle layers of the digital supply chain.
In February 2026, Pubmatic reported earnings that beat expectations by $0.13 EPS, largely attributed to its early adoption of AI-driven optimization tools.
Pubmatic’s involvement is particularly noteworthy. The company has recently signaled a total commitment to agentic AI, viewing it as the next growth engine for its supply-side platform. In February 2026, Pubmatic reported earnings that beat expectations by $0.13 EPS, largely attributed to its early adoption of AI-driven optimization tools. By integrating agentic capabilities directly into the supply side, Pubmatic is positioning itself to capture double-digit revenue growth in the latter half of 2026. This partnership with Butler/Till serves as a real-world proof of concept that agentic AI is not merely a theoretical upgrade but a functional tool for margin improvement.
What to Watch
However, the success of these tests brings a new set of challenges for the industry. As AI agents take over the decision-making process, questions regarding transparency, brand safety, and human oversight become paramount. If an agent is optimized solely for cost reduction, it may inadvertently prioritize low-quality inventory unless strict guardrails are in place. Furthermore, the shift toward agentic buying threatens the traditional roles of media planners and traders, who must now transition from manual operators to 'agent orchestrators.' The industry must now grapple with how to maintain accountability when the 'black box' of AI is making the final call on where millions of dollars in ad spend are allocated.
Looking forward, the success of the Butler/Till and Pubmatic tests is likely to trigger a wave of similar adoptions across the agency landscape. We expect to see a rapid evolution of 'Agentic Operating Systems' within the ad-tech stack, where multiple specialized agents—one for pricing, one for inventory quality, and one for creative optimization—work in concert. For advertisers, the promise is clear: higher ROI through reduced waste. For the broader market, this represents the beginning of a consolidation phase where platforms that cannot support agentic workflows risk becoming obsolete in an increasingly autonomous ecosystem.