Observed Signal · Sep 29, 2026 · Interview/Analysis · Source: ExchangeWire · Impact: 2/5 · Sentiment: Positive
PubMatic CEO Rajeev Goel on Agentic Buying and Cutting Complexity
In an interview with ExchangeWire, PubMatic CEO Rajeev Goel discussed the company's evolution from an SSP to a broader platform, focusing on reducing ad tech complexity and empowering publishers and advertisers through agentic buying. Goel highlighted three pillars: Activate, for direct buyer-SSP transactions; Agentic OS, a suite of 20+ AI agents; and Decision Fabric, which allows agencies and DSPs to run algorithms within PubMatic's infrastructure. He emphasized that AI agents will surpass human traders in accuracy, with governance and reasoning mechanisms for oversight. Goel predicts that agent-to-agent buying will shift power to publishers and advertisers, the 'principals', rather than technology platforms. The company reported 4,000 publisher-created deals using AI year-to-date through Q2.
While not a product launch or major announcement, the interview provides strategic insight from a major ad tech CEO on the shift towards agentic buying and reducing complexity, which is relevant to the industry's future direction but not a discrete event.
Track PubMatic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- PubMatic CEO Rajeev Goel was interviewed by ExchangeWire at ATS London.
- PubMatic has three pillars: Activate, Agentic OS, and Decision Fabric.
- Activate has been in-market for about four years, with hundreds of advertisers using it.
- Agentic OS includes over 20 AI agents, with 4,000 publisher-created deals using AI year-to-date through Q2.
- Decision Fabric combines data from 300 data partners with containers for agency/DSP algorithms.
Connected Companies & Entities
1 Entity mapped“PubMatic's CEO, Rajeev Goel, spoke to ExchangeWire after headlining ATS London....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
PubMatic Expands Agentic AI Beyond Direct Deals
PubMatic said on its Q1 2026 earnings call that its AgenticOS platform has run more than 30 fully autonomous end-to-end agentic campaigns, in addition to over 1,000 direct publisher deals. The company reported Q1 revenue of $62.6 million, a 13% year‑over‑year increase, and said agentic-enabled deals helped drive an 80% YoY rise in its emerging revenue category, which now represents 14% of total revenue. PubMatic highlighted performance gains from AgenticOS (buyers seeing 30–40% CPM improvements and the ability to buy more inventory), expanded mobile and CTV growth, integrations with major mobile mediation platforms and an Amazon DSP integration that raised publisher CPMs. Management flagged near-term headwinds from a major DSP pullback (attributed to The Trade Desk), a projected Q2 revenue decline, and upcoming executive departures in growth/CRO roles.
PubMatic Deploys AI Agents to Simplify Programmatic Buying
PubMatic is deploying a network of AI agents, via its AgenticOS infrastructure (built with NVIDIA) and its Activate buying platform, to automate planning, execution and optimization of programmatic media campaigns. The rollout is part of a partnership with Untapped Growth, a collective of 20 independent agencies, which allows agencies to input campaign preferences and access negotiated commercial terms. PubMatic says AgenticOS replaces manual DSP/SSP orchestration with specialized bots for tasks such as audience discovery and execution; early tests reportedly showed an 82% reduction in buy-side costs and nearly a 90% reduction in campaign setup time for one agency. Two Untapped Growth agencies are expected to go live shortly, while others are testing the system and exploring direct LLM-driven execution (e.g., via Claude) that can bypass traditional DSP layers.
PubMatic Unveils Fast-Track OS for Agent-Based Advertising
PubMatic has launched an "Agentic OS," a Nvidia-accelerated operating system for agent-based programmatic advertising that aims to enable "Agent-to-Agent" workflows where specialised AI agents plan, execute, optimise and monitor campaigns autonomously. Early partners including WPP Media and MIQ will test agentic campaigns under real market conditions. The three-layer architecture (infrastructure, application/agentic functions, transaction layer) claims up to fivefold faster decisioning, sub-millisecond reaction times and fewer auction timeouts. Agentic functions operate via protocols such as the Ad Context Protocol (AdCP) and the Model Context Protocol (MCP) and automate planning, pacing, optimization, troubleshooting and measurement. PubMatic integrates Agentic OS with existing programmatic buying models (Programmatic Guaranteed, Private Marketplaces) and routes media via its Activate buying platform. PubMatic positions the approach as particularly relevant for complex environments like CTV, while noting adoption will depend on control, transparency and scalable governance in practice.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.