How to measure a company's carbon footprint

Carbon accounting is fundamentally a data-collection exercise disguised as an environmental one: the hard part is rarely the arithmetic, it is tracking down activity data — fuel purchased, electricity consumed, freight shipped, goods bought — from systems that were never built to report it. This is an informational, methods-focused guide; whether a specific disclosure is legally required, and under which framework, depends on your jurisdiction, sector, and size, so confirm current requirements with your own compliance or legal function rather than treating anything here as a determination.

The three scopes, and why the boundary matters

The near-universal structure, from the GHG Protocol, is three scopes:

• Scope 1 — direct emissions from sources you own or control: fuel burned in company vehicles, on-site combustion, fugitive refrigerant leaks. This is the easiest scope to measure, because the activity data usually already exists in your own fuel and fleet records.
• Scope 2 — indirect emissions from purchased electricity, steam, heating, or cooling. Measurable two ways: a location-based method using the average emissions intensity of the local grid, and a market-based method that reflects specific contracts, renewable energy certificates, or green tariffs you've purchased. The two can produce meaningfully different numbers for the same company, so report both where your framework asks for it, and state clearly which one is the headline figure.
• Scope 3 — everything else in the value chain: purchased goods and services, business travel, employee commuting, use of sold products, and more, split into fifteen defined categories. For most companies, particularly outside heavy industry, Scope 3 is the majority of total emissions and by far the hardest to measure, because it depends on data your suppliers and customers hold, not you.

Most companies can produce a credible Scope 1 and 2 number within a quarter. A credible Scope 3 number, covering all material categories with primary rather than estimated data, typically takes years of incremental improvement — start with the categories most material to your business, not all fifteen at once.

Where the numbers actually come from

Two ways to calculate an emissions figure for any given activity:

• Spend-based: multiply a dollar amount (money spent on a purchased good, a category of spend) by an average emissions factor for that spend category. Fast, requires no supplier engagement, and imprecise — it captures the category average, not what your specific supplier actually did.
• Activity-based: multiply a physical quantity (liters of fuel, kilowatt-hours, kilometers shipped, units purchased) by an emission factor specific to that activity. More accurate, and requires you to actually track the physical quantity, which is the harder data-collection problem.

The practical path most companies take: start spend-based for Scope 3 categories where you have no better data, and move specific high-materiality suppliers or categories to activity-based, or ideally supplier-provided primary data, over time. A footprint built entirely on spend-based estimates is a reasonable starting inventory, not a number you should present as precise.

Building the inventory, in order

1. Set your organizational boundary — which entities, facilities, and joint ventures count, using either an equity-share or operational-control approach, consistently applied.
2. Collect Scope 1 and 2 activity data from utility bills, fuel records, and fleet logs — usually the most complete data you have.
3. Identify your material Scope 3 categories rather than attempting all fifteen with equal rigor immediately; purchased goods and services, and often either business travel or use of sold products, dominate for most companies.
4. Apply emission factors consistently, documenting the source and vintage of every factor used, because factors are periodically updated and a footprint calculated with mismatched-vintage factors is not comparable across a time series.
5. Set a base year and stick to it, recalculating historical figures when your boundary or methodology changes materially, so year-over-year comparisons stay meaningful.
6. Document everything — the methodology, boundary, factor sources, and estimation assumptions — because whoever reviews or assures the number later needs to reconstruct how you got it.

A shortlist by situation

• You're a developer or platform wanting to embed a carbon calculation into your own product rather than run a standalone tool: Climatiq is built as an emission-factor database and calculation API for exactly that.
• You're a large enterprise with complex, global supply-chain data already sitting in ERP, procurement, and travel systems: Watershed is built to ingest that data directly and produce audit-ready Scope 1–3 inventories.
• You want a free entry point for a single user's annual calculation before scaling up: Persefoni's PRO tier is free for that case, with reporting mapped to CSRD, ISSB, and CDP.
• You want guided, compliance-focused accounting with optional consulting support built in: Normative pairs the calculation engine with advisory services.
• You need to respond to an investor- or customer-driven disclosure request using a standardized, widely recognized format: CDP runs the questionnaire and scoring system many large customers and investors specifically ask suppliers to complete.

Questions to ask a vendor or a consultant

1. Which emission-factor database do you use, how current is it, and can we see the specific factor applied to each activity?
2. Do you calculate Scope 2 both location-based and market-based, and is that distinction visible in the output?
3. How do you handle categories where we have no primary data — pure spend-based estimation, or a documented hybrid approach?
4. Is the audit trail sufficient for third-party assurance, if or when that becomes a requirement for us?
5. What does the tool assume about our organizational boundary, and can we change it without recalculating from scratch?

Common mistakes

• Reporting Scope 2 without specifying location-based or market-based, when the two can tell very different stories.
• Comparing this year's footprint to last year's after a methodology or boundary change, without restating the base year.
• Treating a spend-based Scope 3 estimate as equally precise to a directly measured Scope 1 number.
• Buying carbon credits to claim "carbon neutral" status while treating that as a substitute for measuring and reducing the underlying footprint, rather than a complement to it — and without checking what current disclosure rules in your jurisdiction permit you to claim.
• Losing the documentation trail for emission-factor sources and vintages, making the number impossible to defend or reproduce later.

For the reporting frameworks and disclosure platforms that sit downstream of the inventory itself, see how to choose an ESG reporting tool and every tool in this category.