CIMM East was full of insights highlighting the complexities of media measurement’s future. From currency fragmentation to AI workflows and data transparency, here are six takeaways for the sell side to keep top of mind as the ground keeps shifting.
Let’s be honest—if you attended CIMM (Coalition for Innovative Media Measurement) East looking for clarity, you probably left with more questions. Especially about currencies and clean rooms.
But, the beauty of this moment in media measurement is that it’s messy, in motion, and offers an opportunity to reshape how value gets defined, delivered, and measured.
From panels on the economics of attention to fireside chats about measurement independence, CIMM East was packed with insights, contradictions, and a few truth bombs that hit harder than expected.
Here’s what stood out—and why the sell side should pay close attention.
1. The Currency Question Isn’t Settled—And Maybe That’s a Good Thing
We’ve officially entered the multicurrency era. And now, the narrative is shifting from “Which one wins?” to “Can we just agree on what these numbers mean?”
This debate is no longer about displacing Nielsen. Today’s conversation focuses on deciphering the complex math underlying competing measurement models.
“A single currency is not sacred; it’s just inconvenient,” said Kelly Abcarian, Chief Strategy Officer at Matter More Media, during her industry fireside chat with Jon Watts, Managing Director at CIMM.
That sentiment echoed across other sessions as well.
In his fireside chat, Mediaocean’s Ramsey McGrory emphasized that while his platform won’t define currency, it’s built to support whatever buyers and sellers agree to transact on.
“We’re the system of record,” he said. “Currency-grade measurement has to sit at the campaign and even line-item level—because that’s where deals get reconciled.”
Buyers and sellers need flexible, accurate, and transparent options. But, the industry’s still caught between panel math that doesn’t reflect modern viewing and a patchwork of alternatives that aren’t all interoperable.
The old days of blindly trusting a currency just because it’s MRC-audited are over. As Abcarian pointed out, we need to understand how measurement models differ, not just which logo is stamped on the deck.
2. CTV Is the Problem Child of Measurement—And Everyone’s Trying to Parent It Differently
Sure, CTV has promise, but it also has a lot of problems.
CTV is going through growing pains. Just look at the messiness of measurement, the lack of attention on creative, and the influx of incompatible clean rooms.
“We’re in the gangly teen phase,” explained CIMM’s Evan Cohen during a group preview of the coalition’s study on best practices for CTV.
You’ve got buyers logging into a dozen different clean rooms and struggling to connect fragmented data sets. And creative strategy—arguably the most important piece—is still being bolted on at the end.
“We got a great quote from Danny Reuben, Associate Vice President, CTV at Kargo, one of our sponsors,” said Bhavana Smith. “He said, ‘You know, creative for CTV is an afterthought, but it’s the only thing the consumer sees.’”
It was a pointed reminder, given all the talk about data, targeting, and tech stacks, that creative still rarely gets a seat at the planning table.
And yet, despite all the fragmentation and noise, we’re still treating creative like a downstream deliverable. Like it’s something to be slotted in after the targeting is set and the tech is stitched together.
Measurement is already complex across the industry, but especially in CTV. That complexity is compounded by the lack of standardization across platforms and a disconnect between strategy and execution.
AI might offer hope here, but until systems can communicate with one another, even the best data can’t deliver the best results.
3. Data Quality Isn’t Just a Tech Issue—It’s a Trust Issue
When it comes to data quality, Zach Lain, Director of Global Data Partnerships at PepsiCo, discussed the ground rules for data quality. He said he wants data that’s decision-ready, not perfect, but transparent.
“Tell me where your data is the weakest,” he said. “Transparently flawed will always be better than aspirationally perfect.”
Whether it’s device graphs, ACR data, or inferred household identities, marketers want partners to admit where the cracks are. And they want tools to plan around those cracks, instead of pretending they don’t exist.
It’s about knowing which signals can actually be trusted across planning, targeting, and measurement. And that means publishers—and vendors—need to get honest about what their data can and can’t do.
Perfect data is no longer the goal. Business-grade, actionable data is the new bar, according to Lain. And publishers that can supply it have a leg up.
“Can you tell me where your data is the weakest? Can you tell me where it works, where it doesn’t work?… Maybe transparently flawed will always be better than aspirationally perfect.” — Zach Lain, Pepsico
4. AI in Ad Workflows? Still More Hype Than Help—For Now
Everyone’s talking about AI. But few are actually using it to make meaningful media decisions. According to IAB’s Angelina Eng, 70% of organizations don’t have an AI roadmap yet.
“AI isn’t here to take your job. It’s here to take your spreadsheet,” said Henry Innis, CoFounder of Mutinex.
If AI gives people more time to think and fewer hours lost chasing pivot tables, it’s a win. Yet, most AI use cases are still stuck in the crawling phase, focusing on tasks like RFP support, meeting notes, and some early-stage forecasting.
AI should not be used just for automating internal processes. The tech should be used to serve clients faster and more effectively. For example, Ampersand’s Rachel Herbstman is using machine learning to run predictive simulations for addressable activation and local campaigns.
Is AI transforming media? Not yet. But it’s quietly changing how decisions are made and how fast teams can move.
5. Live TV Isn’t Dead. It’s Getting a Programmatic Makeover
In case your TikTok For You Page had you thinking otherwise, sports, news, and live events are still important.
Now, the focus is on supercharging them with automation, dynamic ad insertion (DAI), and more intelligent targeting.
“Ten years ago, you couldn’t stream live sports reliably. Now we’re talking about frame-accurate, programmatic ad insertion,” said Matt McLeggon, SVP, Advanced Solutions at Magnite.
It’s also time to democratize access to live inventory, making it available via programmatic pipes to brands with smaller budgets, said Dan Rosenfeld, SVP, Advertising, Analytics & Insights at DIRECTV. It’s not just the Super Bowl advertisers that should benefit.
Still, limitations remain. Pacing tools, system readiness, and creative availability are all friction points that prevent programmatic from being a full-fledged player in live. But the trajectory is clear.
6. Audience Intelligence Isn’t Optional—It’s the Backbone of Smarter CTV Planning
To achieve better CTV performance, audience insights must guide the entire planning process. The data exists, but it’s not being integrated early enough to drive more innovative planning or more effective creative.
Audience data has to be integrated across the full campaign lifecycle
Many campaigns are still being built on data that’s big but not accurate, warned Fariba Zamaniyan, VP of Global Data Monetization at TiVo. This leads to targeting gaps that ripple downstream. She also emphasized the importance of incorporating deterministic data into the process and providing real-time feedback to ensure optimal outcomes.
Meanwhile, Wendi Dunlap, EVP of Business Intelligence and Audience Science at Mediahub, pushed for a balance between granularity and scale. She noted that overly narrow targeting can limit reach and inflate costs without improving performance.
According to Alex Groysman, VP of Ad Product Development at Spectrum Reach, planning must account for not only who the audience is, but also where they’re watching. Context is as important as identity, he said. Unfortunately, many platforms still can’t deliver reliable content-level signals.
From the publisher’s perspective, Sara Grimaldi, VP of Audience Impact and Intelligence at Paramount, called for stronger measurement loops to inform mid-flight decisions and post-campaign analysis. She also explained the importance of tying audience insights directly to content adjacency strategies.
But, it was Stephen Fugedy, Lead Solutions Marketing Manager, LiveRamp, who brought home the current operational reality. Identity, segmentation, and campaign optimization are only as strong as the data pipes that support them.
Collecting data isn’t the end goal. You have to activate it across the entire buying workflow. Without interoperable systems connecting identity to activation, segmentation has no impact.