Yesterday we published a piece about Omnicom Media Group's "two maps" problem. Megan Pagliuca building one future. Joanna O'Connell researching another. Ships passing in the night. Very dramatic. Great narrative tension.

OMG's comms team had notes. Oh, did they ever.

And look, when a holding company pushes back on coverage, the response is usually some flavor of "we disagree with the characterization" followed by three paragraphs of nothing. This was different. They sent Digiday clips. They sent product documentation. They sent a polite but pointed question about whether we'd actually looked at how the CES announcements were developed.

Here's the thing. The sources we spoke to, people who've worked inside OMG's vendor ecosystem for years, described a tension. "Megan built the best version of the current machine. The question is whether the current machine is the right machine." That's a direct quote from someone who knows Pagliuca personally and respects her work.

That tension is real. We didn't invent it.

But tension and dysfunction aren't the same thing. And this is where we may have painted with too broad a brush.

The CES coverage tells a more complete story. Digiday, reporting on the announcements as they happened, noted explicitly: "Much of Omnicom Media's moves announced at CES this week were informed by research conducted by Joanna O'Connell."

Not "inspired by." Not "aligned with." Informed by.

That's a different relationship than two teams working on parallel tracks. That's research feeding product. Which is, you know, how it's supposed to work. And how it almost never actually works at holding companies.

Let's trace the lines, because OMG was kind enough to draw them for us.

The Google Consumer Prompt Insights Tool. This is an agentic search product that decodes consumer intent from conversational queries, not just keywords. Where did it come from? O'Connell's "Future of Search" work and the Brand Influence research on how consumers are asking questions differently in an AI-native world. Megan's team built it. Joanna's research scoped it. That's not ships passing. That's a handoff.

The Walmart/Meta Influencer Discovery Agent. This one connects Walmart purchase data to Meta's creator graph to identify which influencers actually move product, not just which ones have followers. The insight driving it? O'Connell's finding that 42% of influencer-driven purchases are spontaneous. If almost half of creator-influenced buying is impulse, you need tools that connect real commerce signals to creator selection. Megan's team built that tool. Joanna's research told them why it mattered.

The Pinterest Shoppable Boards. This collaboration brings commerce into creator environments on a platform where users show up with intent. The thesis underneath? O'Connell's work on trust shifting from institutions to people. Digiday quoted her directly: "You have this kind of trust thing that's happening that's so powerful and so important in these environments." The product team operationalized that insight into shoppable boards. Research became roadmap.

So. Three major CES announcements. Three clear lines back to the intelligence team's work. That's not decoration. That's not thought leadership as sales collateral. That's research driving product.

Let's be precise about what we're saying here.

We're not saying there's no friction. Organizations where research and product agree on everything aren't doing real research. The tension our sources described is probably real. Megan is building for a world that still runs on impressions and clicks because that world is still paying the bills. Joanna is documenting a world where those signals are shrinking. Those perspectives create friction. They should.

What we're saying is that friction appears to be productive. The CES announcements aren't parallel-track product development that happens to vaguely align with research themes. They're specific tools built in response to specific findings. That's the difference between "our research informs our strategy" (which everyone says) and "here are three products we built because of these three research conclusions" (which almost nobody can demonstrate).

We're also not saying OMG has solved the problem. The zero-click economy is still growing. GenAI search still has essentially no ads. The impression-based infrastructure Megan built is still optimized for a world that O'Connell's research says is shrinking. Whether OMG's product roadmap evolves fast enough to match what the intelligence team is finding remains an open question.

But they're at least having the conversation. And they're building things that come out of that conversation. Which puts them ahead of most of their competitors, who are still publishing decks and calling it strategy.

Here's what we got right: the industry-wide problem is real. Most holding companies keep research and product on separate tracks. Most thought leadership is decoration. Most "innovation" announcements are partnerships dressed up as products. The structural critique holds.

Here's what we got wrong: we picked OMG as Exhibit A of the dysfunction when the evidence suggests they might be Exhibit A of the solution. Or at least an early, imperfect attempt at one.

That's actually a more interesting story. It's easier to find agencies that exemplify the research-product disconnect. It's harder to find ones that are genuinely trying to close it. OMG appears to be trying. Whether they succeed is a different question, and one we'll keep watching.

The question now isn't whether OMG's teams are talking to each other. They clearly are, at least more than our initial reporting suggested.

The questions that remain:

Is this structural or personal? Does the research-to-product pipeline survive if O'Connell or Pagliuca leave? Is it built into how OMG operates, or is it a function of two specific executives who happen to work well together?

How deep does it go? Three CES announcements is meaningful. But OMG ships a lot of product. Is research informing the whole roadmap, or just the tentpole announcements that get Digiday coverage?

Does it scale? Building tools in response to research findings is one thing. Changing how billions of dollars in media actually gets planned and bought is another. The products exist. Are clients using them? Is the research actually changing outcomes, or just changing what gets built?

In Parts 2 and 3, we'll dig into the mechanism. How does research actually get into the product roadmap at OMG? What does the organizational structure look like? And what can other agencies learn from it, assuming they want to learn anything at all?

Because the holding company that figures out how to connect research to product, really connect it, not just in press releases but in how money moves, has a meaningful advantage. OMG might be further along that path than we gave them credit for.

Might.

We'll keep looking.