The Internet Changed Everything. Now AI Will Change Everything Again.
The programmatic ecosystem has always thrived on standardization, with shared pipes and protocols enabling a global marketplace for digital advertising that today supports a $770 billion industry. But as the industry scales in complexity, the ways we collaborate have not kept pace. Too often, critical workflows break down due to outdated modes of communication. The result? Missed revenue, higher operating costs, and unnecessary friction.
But just as the internet has transformed how businesses communicate, collaborate, and transact, generative artificial intelligence is fundamentally reshaping how systems interact, make decisions, and solve problems at scale.
In programmatic advertising, this transformation isn’t coming. It’s here.
At PubMatic, we don’t plan to just adapt to the new reality, we plan to create it by ushering in an entirely new operating system for digital advertising. Today, we’re publishing the first in a series of open draft specifications for AI agent-to-agent communication protocols in programmatic advertising, starting with deal management to drive conversation and feedback on the best way forward to improve and drive meaningful adoption.
From Dashboards to Conversations
The programmatic advertising ecosystem processes billions of dollars of transactions monthly, yet it still runs on email threads, manual troubleshooting, siloed systems that can’t talk to each other, and static APIs that break when anything changes. Plus, platform proliferation and UI lock-ins require buyers and publishers to continuously context switch between multiple dashboards and interfaces.
The result? A massive tax on the entire industry:
- Days to resolve issues that should take minutes
- Millions in lost revenue from broken workflows
- Countless hours of human talent wasted on problems machines should solve
- Cognitive overhead from constantly switching between platforms and interfaces
That is where agentic AI comes in.
Agent-to-Agent Communication: The Breakthrough Enabler
Instead of waiting on humans to connect the dots, agents can query, respond, and act instantly. Unlike rigid, rule-based systems that can only automate what they’ve been explicitly programmed to do, AI agents can learn, adapt and evolve based on real-world data. They explore different strategies, measure outcomes, and refine their approach iteratively, becoming smarter and more efficient with every interaction. At scale, these agents can run thousands of decisions simultaneously, optimizing across deals, campaigns or partners in ways no human team or static integration layer could ever match.
Not only does this increase the efficiency of the teams engaged in a programmatic transaction, but it also reduces the learning curve required to onboard new platforms and technologies, making sophisticated programmatic advertising more accessible to brands and publishers. And ultimately, it will mean less friction, reduced operational overhead, more transparency, and stronger outcomes for players across the digital advertising supply chain.
Emerging standards like Model Context Protocol (MCP) and Agent to Agent Protocol (A2A) establish a framework for AI-powered agents to exchange structured, machine-readable information in real time. And because these protocols are compatible with widely adopted AI tools like ChatGPT or Claude – or through a company’s internal AI tools – industry stakeholders can interact directly through familiar interfaces.
Our newly published specification makes it possible for any participant – publisher, DSP, or buyer – in the programmatic ecosystem to plug into PubMatic’s server and accelerate their workflows:
We can finally move beyond static, manual communication and into a world where agents collaborate directly. The benefits are immediate and represent a shift toward a truly adaptable, scalable and interoperable supply chain.
From Reactive to Predictive: The Deal Management Revolution
Deal management touches every transaction in programmatic advertising, yet it remains one of our industry’s biggest pain points. The challenge is clear: while the majority of programmatic ad spend now flows through private marketplace deals and programmatic guaranteed arrangements, the systems managing these high-value transactions remain stubbornly manual. Billions of dollars in media investment still depend on email chains and spreadsheet coordination. We’re addressing that first.
The Old Way: Human Bottlenecks at Machine Scale
Current Workflow: A buyer sees an underdelivering PMP deal. They reach out to their DSP contact. The DSP runs checks, then escalates to PubMatic. After multiple log checks and message swaps, the root cause might be identified, often days after the original issue.
Timeline: Days. Success rate: Inconsistent. Cost: Enormous.
The New Way: AI Agents Collaborating at Light Speed
Agentic Workflow: A buyer’s AI agent (or a buyer working directly within their favorite AI tool like ChatGPT or Claude) asks a DSP’s AI agent why the deal isn’t delivering. The DSP agent automatically queries PubMatic’s MCP Server. Within seconds, root cause analysis is returned – pending creative approval, advertiser blocks, pacing constraints – along with specific resolution steps.
Timeline: Seconds. Success rate: Improving continuously. Cost: Minimal.
The issue is fixed in minutes, not days. The campaign delivers. Revenue flows. Trust builds.
The Network Effect: When Everyone Wins
The true power emerges when this communication layer reaches critical mass. As more participants adopt these protocols, each new participant makes the network more valuable for everyone:
- Publishers get faster issue resolution and higher yield optimization
- DSPs reduce operational overhead while improving client satisfaction
- Buyers ensure their campaigns deliver on pace and on budget with dramatically less effort
- The entire ecosystem benefits from increased transparency and trust
Our open-source deal management specification is publicly available for review and contribution here.
We invite every publisher, DSP, buyer, and technology partner to:
- Review our deal management specification
- Contribute feedback and improvements
- Help us extend these protocols to inventory forecasting, reporting, and beyond
- Shape the future of programmatic collaboration
And this is just the beginning. The framework we’ve built can extend to other critical workflows, from inventory forecasting to reporting and more.
Now it’s time to make it truly intelligent.
The Future is Agent-Native
The programmatic ecosystem has always been about automation and scale. We’re not just making it more efficient. We’re making it more intelligent, more accessible, and more collaborative at the structural level.
Ready to help build the future of programmatic collaboration? Review our open specifications and join the transformation. The next generation of digital advertising won’t wait.
