Interactive Advertising Bureau
22 July 2025

Driving Media Decarbonisation Through Precision: GMSF v1.2 and the Power of Contributed Data

In this blog post, Andrei Baragan, CTO, Cedara, and IAB Europe’s Data Analyst & Sustainability Lead, Dimitris Beis, explore how the latest version of the Global Media Sustainability Framework (GMSF v1.2) marks a significant step forward in the advertising industry’s journey towards meaningful climate action.

Backed by a benchmarking study conducted by Cedara and leading media organisations, they highlight why data fidelity is essential for a sustainable media future.

As the issue of environmental sustainability matures in the advertising industry, the spotlight is increasingly shifting from intentions to implementation, especially when it comes to quantifying and reducing carbon emissions across the digital media supply chain. Transparent and robust carbon measurement is no longer optional; it is foundational to building a sustainable media ecosystem.

To meet this critical need, the newly released Global Media Sustainability Framework (GMSF) v1.2 provides a robust and standardised methodology for measuring carbon emissions in programmatic media. This version introduces a hierarchical approach to emissions data quality, explicitly prioritising contributed data—actual, supplier-provided metrics—as the gold standard for emissions reporting. By structuring the methodology around data fidelity, the GMSF empowers advertisers, publishers, and ad tech providers to make better-informed decisions that drive meaningful emissions reductions.

A Hierarchy of Emissions Data: Transparency Through Structure

At the core of GMSF v1.2 is a clear and pragmatic data hierarchy that categorises the input data, i.e., the data required to obtain emissions figures using the GMSF, into levels. For example, the component of the GMSF methodology that addresses the volume of programmatic activity behind an impression features the following data levels:

  • Contributed Data (Level 2)
    This is actual, supplier-generated data. For example, precise bidding volumes or real-time log files submitted by publishers, SSPs, or platforms. It enables granular, high-precision emissions calculations and best reflects the carbon footprint of media transactions. As the framework's preferred source, contributed data forms the basis of the most accurate, transparent, and actionable measurement possible.
  • Ads.txt-Based Proxies (Level 1)
    When contributed data is not available, the framework recommends using publicly accessible ads.txt files. While not as precise, the length of an ads.txt file serves as a structural proxy for estimating emissions at the publisher level, helping identify the relative complexity and potential carbon intensity of ad delivery systems.
  • Standardised Defaults (Level 0)
    In scenarios where neither contributed data nor ads.txt files are available, the framework supplies baseline default values derived from aggregated industry benchmarks. While these values enable coverage across all digital media activity, they provide the lowest level of fidelity and should be viewed as a last resort.

This tiered model offers a scalable path for the industry to progress from relying on generalised assumptions towards creating a shared foundation for operational carbon reduction that is reliably quantified.

Proving the Value: Benchmarking Emissions with Cedara

To assess the real-world impact of this data hierarchy, four leading media and advertising organisations: DPG Media, IQ digital, GumGum, and Adnuntius partnered with Cedara to conduct a comprehensive benchmarking study. The goal was to quantify carbon emissions intensity (grams CO₂e per impression) across all three GMSF data tiers.

The results were both compelling and consistent. Across the board, the use of contributed data yielded dramatically lower emissions intensity figures compared to estimation-based methods.

CompanyContributed Data
Ads.txt Proxy
Standard Defaults
Estimate Improvement*
Adnuntius0.10 g CO₂e0.47 g CO₂e0.79 g CO₂e4X+
DPG Media0.13 g CO₂e0.43 g CO₂e0.82 g CO₂e3X+
GumGum0.16 g CO₂e0.33 g CO₂e0.81 g CO₂e2X+
iq digital0.09 g CO₂e0.45 g CO₂e0.77 g CO₂e5X+

* reduction in emissions estimate arising from using contributed data over the ads.txt proxy

These findings validate a central design principle of GMSF v1.2: the closer the data source is to the actual transaction, the more reliable, and often lower, the resulting emissions estimate will be. By enabling advertisers and publishers to move beyond proxies and defaults, contributed data offers a path toward more accountable, defensible, and climate-conscious media buying.

What This Means for the Industry

The implications of these results are profound. For brands under pressure to decarbonise their marketing operations, sourcing emissions figures based on contributed data is not just a technical preference; it’s a strategic imperative. Accurate emissions data empowers advertisers to prioritise lower-carbon inventory, influence partner selection, and drive competitive advantage through credible sustainability claims.

For publishers and ad tech platforms, the message is clear: transparency pays off. By enabling access to real data, they can better position themselves as low-emission media partners and strengthen trust with sustainability-minded buyers.

Furthermore, by standardising expectations around data fidelity, GMSF v1.2 lays the groundwork for a cleaner media supply chain, one where carbon is no longer an invisible cost of digital reach.

Conclusion: From Framework to Action

As media decarbonisation gains momentum, frameworks like GMSF v1.2 represent a critical leap forward in how the industry measures, manages, and ultimately reduces its environmental impact. The evidence from Cedara’s benchmarking study reinforces that data quality matters, and that contributed data is not just more accurate, it’s transformational.

For an industry long reliant on estimates and assumptions, the path to net zero now begins with real numbers. And the sooner stakeholders embrace them, the faster the media ecosystem can realise a more sustainable, transparent, and accountable future.

Appendix: Study Methodology

The benchmarking analysis examined one month of advertising data, focusing on the top 1,000 publishers by impression volume for each participating vendor. The study calculated Ad Selection emissions—encompassing both server processing and network transmission—using the three data levels defined in GMSF v1.2. 

For contributed data calculations, actual bid logs provided precise bid volumes and request counts for the vendors, which were then extended to estimate rest of supply chain activity using GMSF multipliers.

The ads.txt proxy method applied standardised coefficients (1.412 for server emissions, 1.464 for network emissions) to the number of lines in each publisher's ads.txt file, with a default assumption of 3,000 lines where ads.txt files were unavailable.

To isolate the impact of data quality on Ad Selection emissions specifically, other emission components such as Creative Transmission and User Device were held constant across all three measurement approaches.

Our Latest Posts

Sign up for our newsletter
linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram