A recent project found a 60%-plus discrepancy between data from a major currency measurement company and an ACR-ad supported measurement provider, raising urgent questions for the TV industry.
I’m currently consulting on a project that involves comparing data from one of the major TV currency measurement companies with data from an ACR-supported ad measurement provider that is not an industry currency and has no interest in being one. Specifically, we looked at estimated audience impressions from a major U.S. broadcast group, aggregated across their markets to reflect total 2024 impressions.
What we found was stunning: There was upwards of 60% discrepancy between the two sources. While it’s understandable that an audience measurement company and an ad intelligence platform — one estimating viewership, the other counting ad occurrences — would show some variation, a gap of this magnitude raises serious questions. How can one source report less than half the impressions of the other and still be considered credible?
Should The Industry Be Concerned?
Imagine you’re a broadcaster selling a show based on an estimated audience of 400,000 impressions, when in reality, the show is delivering closer to 1 million. That discrepancy means broadcasters are leaving money on the table, while advertisers are getting a tremendous deal.
Wouldn’t it be a game-changer if broadcasters could definitively prove their programming is delivering double the impressions currently credited by traditional measurement? You’d think the buy-side would pay more for that kind of verified performance. Yet network executives tell me that buyers won’t adjust pricing, even with definitive third-party proof of a larger audience.
This logic seems inconsistent: If a measurement company reports an audience doubling week-over-week, buyers adjust their spend accordingly. But when a competing data set shows that the actual audience is consistently 2x what the currency reports, it’s dismissed?
Why Traditional Measurement Alone Isn’t Enough
Traditional TV measurement is based on panels, small samples used to extrapolate national viewing behavior. These panels offer critical demographic details — age, gender, household size, income, ethnicity — and help identify who’s in the room watching. And now, with the integration of ACR (automatic content recognition) data from smart TVs, measurement companies are blending panel data with large-scale, passive viewing data.
Panels also help represent hard-to-measure audiences: rural homes, OTA viewers, households with non-smart TVs and DVR users. This hybrid model — panels plus Big Data — is intended to improve accuracy. But with all this data, shouldn’t measurement be getting better, not more fragmented? Or has the fragmentation of devices and platforms — OTA, FAST channels, HDMI streaming sticks, native OS apps, set-top boxes — made modern TV viewing harder to capture than ever?
Is 60%-Plus Discrepancy Acceptable?
A number somewhere near a 60% difference in estimated impressions between a traditional measurement provider and an ad counting company whose technology captures over 1 billion TV ad occurrences per day using fingerprinting is just too large to ignore.
ACR-based impression counts are based on a bottom-up approach of detecting ads from a catalog of known videos on participating panel televisions, then using information about markets and TV ownership to extrapolate those screen impressions to the whole population. This is very different from measuring reported program viewership in various demographics and then using complex modeling to assess how many people watched each segment.
The fingerprinting is a 1:1 counting method. It works across platforms, apps and inputs. Shouldn’t a technology that recognizes all ad occurrences, regardless of content source, be closer to the truth?
And yet, I keep hearing from executives that even if this more comprehensive data proved a significantly larger audience, it still wouldn’t change how much buyers are willing to pay. That’s a fundamental disconnect.
A Path Forward: Counting The Whole Audience
If broadcasters could accurately count all impressions across all devices — and offer audience-based buying across OTA, FAST, HDMI, native apps and set-top boxes — we could finally move beyond buying based on age/gender ratings within a specific show. We could start selling like YouTube, Hulu, Netflix, Fire TV and other platforms that are rapidly capturing market share.
Streaming is sold on an impression basis, not tied to specific shows. As the TV ad market shifts toward audience-based, addressable buying, broadcast must evolve with it.
The Technology Exists
Watermarking, coupled with the IAB Tech Lab’s Ad Creative Identity Framework (ACIF) known in the U.S. by Ad-ID is a promising solution. The A334 and A335 watermark specs standardized by the ATSC can be implemented by TV OEMs, HDMI devices and streaming apps. When inserted in ads these watermarks can be read by devices (if enabled by manufacturers), enabling countable and addressable advertising, even in linear broadcast.
I’m currently working with a team of incredibly smart people to roll out a proof of concept with CIMM (Coalition for Innovative Media Measurement). We’re collaborating with participating broadcasters, device OEMs and advertisers. The result will be a white paper documenting the implementation, the results and the real-world implications. Best of all — it’s relatively easy to deploy.
The Stakes Are High
In 2024, digital video ad spend officially surpassed broadcast TV ad spend ($63 billon vs. $59 billion, according to the IAB), largely because it’s easier to buy digital at scale. If we continue to rely on legacy measurement and delivery models, broadcasters risk being left behind.
Imagine a future where a broadcaster’s entire audience — across all devices — is counted, addressable and measurable, one where unduplicated reach and frequency can be quantified across a campaign flight, giving buyers the same transparency and flexibility they expect from digital platforms.
It’s Time To Act
Broadcasters must come together to fix these foundational issues. By enabling accurate, consistent measurement and targeting across platforms, we can make it easier for buyers to plan, place and optimize campaigns — and ultimately, increase their investment in broadcast TV.
The tools are available. The demand is there. The only question is will we move fast enough to meet it?
