---

"Scope 3" is the most-discussed and least-evidenced part of most corporate carbon inventories. Boards approve the headline number. Auditors then look behind it and find that 70–90% of total emissions are concentrated in a half-dozen categories that were estimated, screened or skipped.

This is a working primer: what Scope 3 actually covers, where inventories typically miss, and what a defensible baseline looks like.

The three scopes — quickly

The GHG Protocol defines three scopes:

• Scope 1 — direct emissions from owned or controlled sources (boilers, vehicles, refrigerants)
• Scope 2 — indirect emissions from purchased electricity, steam, heat or cooling
• Scope 3 — all other indirect emissions in the value chain, upstream and downstream

For most companies, Scope 3 dwarfs Scopes 1 and 2 combined. Often 70–95% of total emissions sit in Scope 3.

The fifteen Scope 3 categories

The GHG Protocol splits Scope 3 into fifteen categories — eight upstream and seven downstream.

Upstream

1. Purchased goods and services — emissions embodied in everything you buy
2. Capital goods — emissions embodied in capex
3. Fuel- and energy-related activities — Scope 2 you don't already count (transmission losses, well-to-tank)
4. Upstream transportation and distribution — supplier logistics into your operations
5. Waste generated in operations — disposal of operational waste
6. Business travel — air, rail, hotel
7. Employee commuting — including homeworking energy use
8. Upstream leased assets — assets you lease as the lessee

Downstream

9. Downstream transportation and distribution — getting your product to customers
10. Processing of sold products — emissions from third-party processing
11. Use of sold products — what your product emits in customer hands
12. End-of-life treatment of sold products — disposal of what you sold
13. Downstream leased assets — assets you lease out as the lessor
14. Franchises — emissions from franchisees
15. Investments — financed emissions (especially material for financial services)

Where corporate inventories miss

Five recurring patterns.

1. Category 1 estimated by spend, then forgotten

The single largest Scope 3 category is usually purchased goods and services — and most baselines build it by multiplying spend by an emissions factor. This is acceptable as a screening exercise. It is not acceptable as a static answer. Spend-based estimates carry uncertainty bands of ±50% on a good day, and they don't tell you which suppliers are decarbonising.

2. Category 11 — use of sold products — ignored entirely

If you make heating equipment, vehicles, IT hardware or anything that consumes energy in use, Category 11 is usually your largest single Scope 3 line. Skipping it because it is "downstream" misrepresents the business.

3. Category 15 — financed emissions — out of scope by reflex

For any institution with a balance sheet of investments, loans or insurance — financed emissions dwarf operational emissions by orders of magnitude. PCAF (Partnership for Carbon Accounting Financials) provides the methodology; pretending Category 15 is "too hard" no longer passes scrutiny.

4. Boundary confusion between leased and owned assets

GHG Protocol gives you two approaches: equity share and operational control. Switching between them — even unintentionally, year on year — creates apparent emissions movements that have nothing to do with decarbonisation. Lock the boundary on day one.

5. No data refresh cadence

Scope 3 is built and then quietly inherited. The right cadence is annual data refresh for the top 80% of emissions by category, refreshed methodology every 3–4 years.

What a defensible baseline looks like

Five characteristics.

• Categories screened, not skipped. Every category is assessed for materiality, including those you ultimately exclude. The exclusion logic is documented.
• Spend-based estimates flagged as such. Each emissions number is tagged with its data quality — supplier-specific activity, industry-average activity, spend-based — and reported transparently.
• Top vendors are activity-based. The top 20–50 suppliers by emissions report activity data directly, not spend.
• A consistent boundary that survives mergers. Equity share or operational control — pick once, document why, restate prior years on any major change.
• Independent assurance pathway. The baseline is built to limited-assurance standards from the start, not retrofitted at year three.

Closing

Scope 3 is the area where the most carbon — and the most disclosure risk — sits. Most companies are still in screening mode. The competitive advantage in the next five years will sit with the organisations whose Scope 3 inventory is accurate enough to make decisions on, not just accurate enough to report.

That is a different calibre of work. And it starts with naming what's missing.