Scope 3 is not one number. It is fifteen separate accounting problems, defined by the GHG Protocol's Corporate Value Chain (Scope 3) Standard, each with its own boundary, its own data sources, and its own failure modes. Companies that treat it as a single line item almost always miss the categories that matter most to their actual business.
The fifteen categories split into two groups: eight upstream categories (emissions from what a company buys and how it gets to them) and seven downstream categories (emissions from what happens after a product leaves the gate).
Upstream categories
- Purchased goods and services — the embodied emissions in everything a company buys, from raw materials to office supplies. For most companies, this is the single largest Scope 3 category.
- Capital goods — the embodied emissions in equipment, machinery, buildings, and vehicles a company purchases, amortized over the asset's use.
- Fuel- and energy-related activities — the upstream emissions of fuels and electricity a company uses, not already counted in Scope 1 or 2 (extraction, refining, transmission losses).
- Upstream transportation and distribution — emissions from third-party logistics moving goods to the company, and moving the company's products between its own facilities.
- Waste generated in operations — emissions from the treatment and disposal of waste produced at the company's own sites.
- Business travel — emissions from employee travel for business purposes, in vehicles not owned by the company (flights, trains, rental cars, hotels).
- Employee commuting — emissions from employees traveling between home and work.
- Upstream leased assets — emissions from operating assets a company leases but does not report under Scope 1/2 (relevant mainly for lessees using the operational control approach).
Downstream categories
- Downstream transportation and distribution — emissions from transporting sold products to customers, when not paid for by the reporting company.
- Processing of sold products — emissions when a company sells an intermediate product that a downstream party processes further before it reaches an end user.
- Use of sold products — emissions from customers using the products a company sells over their lifetime. For any manufacturer of something that consumes energy or fuel, this is frequently the single largest category in the entire inventory.
- End-of-life treatment of sold products — emissions from disposing of or recycling products at the end of their useful life.
- Downstream leased assets — emissions from assets a company owns and leases to others, not already in Scope 1/2.
- Franchises — emissions from the operations of franchisees, for companies that operate a franchise model.
- Investments — emissions associated with a company's investments, relevant primarily to financial institutions, private equity, and holding companies.
Why the split matters in practice
A retailer's inventory looks nothing like a carmaker's. For a retailer, Category 1 (purchased goods) usually dominates — sometimes over 80% of total footprint. For an automaker, Category 11 (use of sold products) usually dominates, because the fuel burned over a vehicle's lifetime dwarfs the emissions of building it. A bank's inventory is almost entirely Category 15 (investments and financed emissions), with everything else being a rounding error.
This is why a generic "Scope 3 percentage" benchmark across industries is close to meaningless. The right first move in any Scope 3 program is not to estimate all fifteen categories with equal effort — it's to identify the two or three categories that plausibly hold 80% of the footprint, and put the measurement budget there first.
Screening before you measure
The GHG Protocol's own guidance recommends a relevance screening step before full measurement: for each category, assess size, influence, risk, stakeholder expectations, and outsourcing before deciding how much precision to invest. Categories that are clearly immaterial can be estimated coarsely and documented as such; categories that are both large and within the company's influence deserve primary supplier data, not spend-based proxies.
Getting this triage right — before a single supplier survey goes out — is usually the difference between a Scope 3 inventory that takes three months and one that takes eighteen.