Blog Post 10 Min Read

Scope 3 Emissions

Hai Bao

Data Science Enthusiast • Sep 13, 2026

Why rank companies based on Scope 3 emissions?

Why Scope 3 Rankings Matter for Investment

  • Uncovering Hidden Risks: Scope 3 emissions cover indirect upstream and downstream activities—such as supply chains, transportation, and product use. These often make up over 70% to 90% of a company's total carbon footprint. Ignoring them leaves major financial and regulatory blind spots.
  • Exposing Transition Vulnerabilities: Companies that rely on high-emission suppliers or produce polluting products face future carbon taxes, regulatory penalties, and shifting consumer demands. Ranking highlights which firms are most exposed to these transition costs.
  • Testing Climate Commitments: Many companies make net-zero promises, but only transparent Scope 3 data shows whether they are actively cleaning up their supply chains or just greenwashing their direct operations.
  • Guiding Capital Allocation: Institutional investors use these rankings to direct capital toward industry leaders who innovate and decarbonize their entire ecosystems, lowering long-term portfolio risk.

Rankings

The following rankings are based off data extracted from Annual Reports and Sustainability Reports.

Fig 1.1: Scope 3 Emission Rankings
Click on the bottom right of the graph for a fullscreen view.

Notes

With ESG Data, not every company releases their data at the same time so we might not always have the data for all companies.

Methodology

All companies are ranked across the different 15 Scope 3 Categories.
If the company does not disclose a figure, it is assigned 1 below the lowest score. For example, if the lowest ranking for Category 1 is 9, then non-disclosures are assigned a score of 10.

Takeaways

Most companies are close in score. Porsche has a poor ranking and that is to be expected as performance parts don't enjoy economies of scale. Mazda and Jaguar Land Rover have poor rankings due to non-disclosures. The figures have been normalized through tco2e/vehicle since brands like VW are extremely high volume manufacturers (approx 9 million vehicles), whereas Porsche only produces approx 300k vehicles a year.

Databricks

Implementation of Medallion architecture

Databricks

Fig 1.2: Databricks architecture.
Click image for fullscreen view.

Tech Stack

  • Docling for PDF To Vector conversion
  • PGVector for storing Vectors
  • Claude for Table Extraction from PDF
  • Postgres for storing extracted table data
  • Databricks for ETL using Medallion architecture
  • Power BI for final dashboard deliverable