• by Cory Treffiletti , Featured Contributor, April 15, 2026

I remember the first media plan I ever put together. It was a super-detailed PowerPoint deck accompanied by a highly detailed spreadsheet that truly impressed my boss, but which no one would recognize today.

It had rows of placements across different channels and formats, columns of circulation counted in audience and impressions, GRPs, CPMs, reach and frequency estimates cobbled together from rate cards and gut instinct. It took days to build and days more to get approved.

It was the plan, and the plan was sacred in the world of media planning and buying.

Here's a question I've been sitting with lately: What happens to that plan when an AI agent can do all of that in just a few minutes?

That's the conversation for where we are today. Right now, platforms like Yahoo DSP, PubMatic, and others are embedding agentic AI directly into the buying workflow, creating systems that can take a campaign objective written in plain language and autonomously plan, execute, pace, and optimize a media buy in real time, without a human touching a single line item.

The Interactive Advertising Bureau Tech Lab has already released a roadmap for agentic buying. NBCUniversal is testing agentic systems that execute buys across its entire portfolio automatically.

The machinery of the media plan, the targeting parameters, the bid logic, the pacing rules, the optimization levers, are all being absorbed into software that doesn't sleep and doesn't need a PowerPoint to explain its rationale.

The media planning document, as we knew it the last 30 years, is dying.  Maybe that’s OK?

The unspoken truth about the media plan is most of it was simple performance. It was a formal exercise in making assumptions look like certainty. We built elaborate spreadsheets to give clients confidence that we'd thought through every variable, when the reality was that the second a campaign went live, half of those variables went out the window, and we were optimizing on the fly.

The plan was what we called the strategy. Actually, it was the story we told about the strategy. The real strategy came from what we did once things went live.

Agentic media buying makes clear that when the mechanical work disappears, the creative becomes the only real differentiator left. An AI can optimize distribution. It can negotiate placements, shift budget between channels in real time, and identify the best moment to reach a given audience at a given price.

What it cannot do is decide what you say when you get there. It cannot craft the brief that gives the campaign its reason for being. It cannot make the creative decision that turns a campaign into a cultural moment.

We spent so much time on the how and the where that we often shortchanged the why and the what. The best media plan in the world doesn't save bad creative. Now that agents are taking over the how and the where, the why and the what become the whole job.

So, what do media planners do in this new world? I'd argue they become something closer to a strategist and a creative director than they've ever been. They set the objectives that the agent operates within, the approach to the audience, and the tone that no algorithm inherently understands. They read the signals that the agent surfaces and make judgment calls that require context the machines can't yet have (and may never have). They own the brief. The brief becomes sacred the way the plan used to be.

I'll be honest.  I have some nostalgia for the old-school media plan. There was something satisfying about building a comprehensive document that mapped out a campaign from start to finish, even if reality had other ideas.

That being said, nostalgia is a terrible strategy. The media plan was a tool, and like every tool in this industry, it will be replaced by a better one. The best planners I've ever known weren't good because they could build a spreadsheet. They were good because they understood people.

And that part? No agent is coming for that anytime soon.