Scope 3 (the indirect emissions scattered across a company's value chain) usually makes up 70 to 90 percent of a company's total carbon footprint. It's also, without much competition, the hardest and most resource-intensive part to measure.

Why the Manual Approach Falls Apart

The traditional playbook looks something like this: sustainability teams work through invoices line by line, look up emission factors in published databases one at a time, run calculations in a spreadsheet, and chase suppliers for primary data whenever they can. It works, in the sense that reports eventually get produced. But it's slow, it's error-prone, and it simply doesn't scale as a company grows or as reporting requirements tighten.

What a Real Automation Pipeline Looks Like

A more systematic approach breaks down into four stages:

1. Invoice parsing: pulling line items out of both structured formats (UBL, CII) and the messier unstructured ones
2. Emission factor resolution: matching each item against verified databases (EU, Defra, EPA) using product classification codes, rather than a human searching one by one
3. Supplier outreach: automatically requesting Product Carbon Footprint data where it's available, instead of relying on manual email chains
4. Fallback logic: falling back on published factors when supplier data isn't available, with a clear data quality score attached so nobody mistakes an estimate for verified data

Why This Isn't Optional Under CSRD

ESRS E1 requires Scope 3 to be reported category by category, with data provenance that holds up under mandatory assurance. A manual process built on spreadsheets and email threads simply can't deliver the consistency or the audit trail that requirement demands, neither reliably nor at scale.

That gap between what CSRD asks for and what a manual process can actually deliver is exactly why we built BUME to automate the full pipeline, from the moment an invoice comes in through to a category-level CO₂e figure ready for the report.