The broadcasting industry has operated on an uncomfortable compromise for almost a century: measure a small sample of viewers and extrapolate the results to represent tens of millions of households. From Nielsen’s mechanical Audimeter and viewer diaries in the 1950s to the electronic “people meters” of the 1980s, the technology used to collect audience data has evolved considerably. Yet the industry’s underlying methodology has remained remarkably consistent since researchers like Archibald Crossley began calling households in the late 1920s: measuring a relatively small sample and projecting those results to represent a much larger audience. The SES insight paper, “The Evolution of TV Audience Measurement,” follows this journey all the way to the real-time, census-scale analytics now made possible by connected TV. A clear message for both broadcasters and advertisers emerges: the industry is finally bridging the gap between the tools created for a three-channel world and today’s fragmented, multi-platform viewing landscape.
Why Measurement Accuracy Matters
Advertising revenue is allocated or misallocated, through audience measurement. Inaccurate measurement causes advertisers to either pay too much or too little for audiences because the system is unable to fully capture audiences that are not included in the reported numbers. Meanwhile, broadcasters are forced to make decisions about scheduling and commissioning based on incomplete or delayed data, sometimes putting high-value content in slots where it underperforms because the numbers were inaccurate or arrived too late.
According to the insights paper, this equation is completely changed by accurate, real-time measurement. A broadcaster has a fundamentally stronger commercial position than one that relies on estimation if they can show an advertiser exactly how many households in a specific region watched a specific program at a specific time, and they can do so in real-time rather than weeks later.
When the Panel Model Broke
The report highlighted a structural issue: while a panel of a few thousand homes could produce statistically meaningful data when three networks commanded roughly 90 percent of the audience, that same sample size fails in today’s fragmented market. Significant statistical error becomes the standard for all but the largest properties when the same panel size is diluted across hundreds of channels and various platforms.
The problem is made worse by the fact that streaming services were built with measurement in mind from the beginning. Streaming platforms know exactly who is watching, for how long, and for what device. However, as ad-supported tiers have expanded, this data has remained proprietary and only selectively shared with advertisers. As a result, measurement companies have been forced to develop new approaches from Nielsen’s proprietary Big Data + Panel methodology to emerging industry-wide standards.
Connected TV and the HbbTV Opportunity
A key finding of the paper is the measurement potential of Hybrid Broadcast Broadband TV (HbbTV), a European standard that provides connected TVs with a return path to transmit viewing data back to the broadcaster in real time. More than 100 million connected TV households in Europe are HbbTV-compatible, more than 60 percent of connected TVs are HbbTV-enabled, and the number of HbbTV-compliant households is projected to reach 120 million by the end of 2026.
Unlike proprietary smart TV platforms, HbbTV functions across multiple manufacturers and distribution networks, providing a horizontal measurement layer that is not siloed within a single ecosystem. Hybrid approaches like Nielsen’s Big Data + Panel combine return-path data with automatic content recognition, drawing on 45 million households and 75 million devices, with a traditional panel to bridge the gap between scale and demographics. The big data tracks the numbers of devices tuned in, while the panel identifies exactly who in the household is watching.
Real-Time Measurement at Satellite Scale
SES’s solution, Channel Analytics, is an example of how far the industry has advanced. By utilizing an HbbTV application, it measures audiences throughout the SES satellite footprint in real-time, panel-free, and collects data directly from connected devices instead of using delayed snapshots.
With per-minute viewership trends over a 24-hour period, program-level rankings, geographic breakdowns down to postcode level, and demographic enrichment using third-party data, its granularity goes far beyond traditional panels. Within a GDPR-compliant consent framework, broadcasters can get real-time dashboards and overnight reporting.
The Journey Ahead
Just as web analytics replaced a single circulation figure with real-time metrics like unique visitors, broadcast television is going through a similar transition from proxies to reality. Connected TV measurement is the broadcast equivalent of what analytics platforms did for digital publishing.
This transition isn’t complete. It is still difficult to unify linear, streaming, and mobile viewing into a single, coherent picture, and it will take time for markets to agree on common metrics. However, the direction is clear. Almost a century after Crossley’s initial phone calls, the industry is finally positioned to understand who and what is being watched with a level of precision and scale never possible before.
As Michele Gosetti, VP of Sports & Events and New Media at SES, puts it: “As the industry continues to evolve, the question is no longer whether audience measurement needs to change – but how quickly organizations can adapt to a more accurate, responsive way of understanding their viewers.”
