Inside NBCUniversal’s test to use AI agents to sell ads against a live NFL game

“It works. It is a functioning technical proof-of-concept that accurately represents what the buyer wants to buy and what the seller has to sell,” said Ryan McConville, chief product officer and evp of ad products and solutions at NBCUniversal on the latest Digiday Podcast.

Despite the successful test, NBCU isn’t about to outsource its entire ad sales process to AI agents anytime soon. “We are a ways away from having this fully productionalized where multiple agencies are using this day in and day out to replace current workflows,” he said.

That being said, traditional TV ad buying and selling workflows have been in need of an upgrade. “There’s still a lot sold via direct IO,” McConville said. So NBCU, RPA, Newton Research and FreeWheel enlisted the AI agents to handle some of the more menial, manual tasks, like gathering inventory availability, pricing and impression forecasts to package up an ad sales package to present to RPA’s buy-side AI agent.

That agent-to-agent interaction actually isn’t so far off from the human-to-human interaction of traditional ad sales or the human-to-web dashboard interaction of programmatic ad sales. And there is a human ultimately at either end of the interaction. They’re just interacting with the AI agents.

McConville described it as “the emerging [user interface] experience where the buy side may just be in an agent and asking these questions and the answers are coming to them through the agent versus through a phone call, an email, fax or even a dashboard, which is the more modernized version of it.”

Here are a few highlights from the conversation, which have been edited for length and clarity.

‘AI is the new UI’

One of the taglines we’ve been kicking around is “AI is the new UI.” What I mean by that is that we’re actually not creating something net-new under the hood. The identity graph, the forecasting model, the product catalog, all of those things are built into traditional applications. What we’re essentially doing is making those APIs available to these MCP servers so that agents can just ask the questions of those applications.

‘Is agentic buying going to replace programmatic?’

When we look at agentic AI, we see the opportunity to automate the full TV buying process. All-encompassing. That can include agents that negotiate deal IDs and then push those deal IDs to DSPs and SSPs. I get the question a lot: “Is agentic buying going to replace programmatic? And my answer to that is always that it gives you the optionality. But no, not necessarily.

‘Garbage in, garbage out’

The place where the most amount of work needs to be done is in the data foundations. You need clean data. Your product catalog needs to be well-organized. You need to have your data in order. You need to have these services available. Otherwise, it’s garbage in, garbage out.

‘Premium automation’

Agentic AI and the ability to orchestrate these AI agents offer what I was calling premium automation. OpenRTB can streamline and automate the buying of certain things, but understanding bespoke ad formats, different sponsorship integrations – OpenRTB doesn’t really do that. But you can ask those questions of an [AI] agent and get those answers and start to automate some of that higher-end part of the market.