For decades, traditional television has been closely tied to an advertising trading metric: the GRP (Gross Rating Point).
This indicator, based on reach and frequency, has long been the undisputed benchmark in TV advertising. Its calculation comes from multiplying 'reach × average frequency.'
Let’s assume a TV advertising campaign reached 58% of its target audience, and each of those individuals was exposed to the campaign/TV spot 4 times:
GRP = 58% × 4 = 232 GRPs
Based on this premise, it becomes clear that GRPs did not measure 'effectiveness' but rather 'exposure.' Additionally, it is important to highlight a key aspect of this model: “how were/are those audiences measured?” The audience of each program, channel, etc., is based on population estimates obtained through audience panels—that is, representative household samples used to extrapolate the behavior of the whole.
These panels are managed by 'independent' companies specialized in measurement, designated as official measurers and agreed upon by the entire industry in each country.
Thus, the official measurers in recent years have been Nielsen in the U.S., BARB in the UK, Médiamétrie in France, AGF Videoforschung in Germany, Kantar in Spain, ...
After this brief overview, one thing seems clear: all these measurement models, focused on exposure and based on sampling a portion of the audience to then extrapolate to the total population, could hardly be considered 100% accurate. However, the main value of these models was their ability to bring about consensus within the television industry, as they were accepted by Media/Broadcasters, Agencies, and Brands as the trading standard.
In other words, they allowed measuring 'apples to apples,' even if the model was neither ideal nor the most sophisticated.
However, this model shows clear limitations in a fragmented and digitized environment. The rise of Connected TV (CTV), on-demand viewing, and multi-screen consumption has called into question the ability of panels to accurately capture new television consumption habits.
Given the limitations or 'potential shortcomings' of this measurement model, CTV has challenged this consensus since its inception. While there has not been uniformity across all platforms or markets, there has been a certain willingness from the major streaming platforms to be measured under these models (a kind of 'I have to play by the rules'), although not all have chosen to remain under such audits for long.
Without delving into the debate over the validity of one set of metrics versus another (a topic worthy of several opinion pieces), the reality is that today two different languages coexist, based on distinct methodologies: one relying on estimates derived from panels, and the other on the precise counting of impressions served. This diversity has made it difficult for years to establish a truly unified standard, although the industry is now beginning to take firm steps toward that convergence.
In 2025, this scenario is beginning to undergo a definitive transformation. Nielsen, historically the leading TV measurement provider in the U.S., will discontinue its panel-only system, marking a turning point for the entire industry. This decision highlights the need to adopt more flexible and hybrid models, capable of integrating different data sources without losing comparability, and will undoubtedly have short- and medium-term implications for other players in markets worldwide.
As discussed above, the convergence between linear TV and CTV has created an increasingly complex environment in which traditional measurement models no longer seem sufficient. Fragmented consumption across linear channels, streaming services, apps, and connected devices demands new methodologies capable of more accurately reflecting television’s true reach.
In this new landscape, panel-only systems fall short. It is no longer enough to estimate aggregated audiences: it is necessary to measure actual viewing, deduplicate users across platforms, and understand the incremental contribution of each environment. To address these needs, the market is shifting toward hybrid models that integrate large-scale data from set-top boxes, smart TVs, streaming applications, and user-declared information.
This evolution makes it possible to capture actual viewing behaviors rather than relying solely on projections, opening the door to richer, more representative measurement of audiovisual consumption.
The adoption of Big Data alone would not be sufficient to accurately measure Connected TV. Artificial intelligence has become the key tool for processing large volumes of data, performing data cleansing, and turning it into useful and comparable audience metrics.
Unifying Dispersed Data: In CTV, users consume content across different services and platforms. Big Data collects this massive footprint (streaming logs, smart TV data, OTT apps, among others), while AI cleanses, processes, and unifies these signals, reducing duplication and identifying unique audiences.
Identities and True Reach: The absence of cookies or persistent identifiers in television environments requires the use of probabilistic and deterministic models that estimate the number of unique individuals exposed to content and calculate metrics such as true reach and unduplicated frequency.
The growth of CTV and the fragmentation of consumption have highlighted the need to build consistent metrics across environments. While panels remain the foundation for measuring linear TV, CTV and digital data require complementary approaches. The most common solution is to integrate traditional panels with digital platform logs, allowing for the estimation of unique audiences without losing comparability.
This is precisely the approach Nielsen has adopted in the U.S., replacing the 'panel-only' model with a hybrid system that integrates data from tens of millions of devices (smart TVs, set-top boxes, etc.) with the panel sample. This approach, accredited by the Media Rating Council, will allow for the complete phase-out of the traditional panel in 2025 and aims to provide a more accurate, continuous, and representative measurement of television audiences in a fragmented environment.
2025 will be remembered as the year the industry began moving beyond the traditional panel-only model. However, adopting new technologies alone is not enough: for measurement to truly reflect current consumer behavior, it will be necessary to implement hybrid models, promote collaboration among stakeholders, and advocate for unified and open standards across the entire ecosystem.
And what seems increasingly difficult to achieve: industry-wide consensus from all angles: advertising, production, distribution…
The good news is that this change is already underway. Far from losing relevance, television is redefining the way its consumption is measured. In doing so, it is also evolving the way the industry understands and compares its audiences. As discussed throughout this article, Nielsen has replaced the 'panel-only' model with a hybrid system that combines panels and Big Data, maintaining the GRP as a reference but under a new calculation logic better aligned with fragmented consumption.
This means that, while the industry is moving toward more granular and digitally comparable metrics, different measurement languages still coexist. The big question is whether this coexistence will slow the path toward a truly unified metric or, on the contrary, whether market pressure will accelerate the adoption of a common standard.
At tvads we has a professional team able to advise you on this field and and guide you in any area of your streaming advertising business, advising you or even operating it on your behalf if necessary
All author posts