How We Calculate Your Carbon Emissions
We believe in full transparency. This page explains exactly how your Carbon Draft report is generated — from the data you upload to the final emissions figure in your PDF. No black boxes.
The GHG Protocol Corporate Standard
Every Carbon Draft report is aligned with the GHG Protocol Corporate Accounting and Reporting Standard (Revised Edition, 2015), the most widely used international standard for corporate greenhouse gas accounting. Over 92% of Fortune 500 companies that report emissions use this framework.
The GHG Protocol was developed by the World Resources Institute (WRI) and the World Business Council for Sustainable Development (WBCSD). It provides the accounting framework that underpins nearly every major climate disclosure program, including CDP (formerly the Carbon Disclosure Project), the Science Based Targets initiative (SBTi), and the EU's Corporate Sustainability Reporting Directive (CSRD).
We follow the GHG Protocol's classification of emissions into three scopes:
Scope 1 — Direct Emissions
Emissions from sources your company owns or controls directly. This includes on-site fuel combustion (natural gas boilers, diesel generators), company-owned vehicle fleets, and fugitive emissions such as refrigerant leaks from HVAC systems.
Scope 2 — Indirect Energy
Emissions from the generation of purchased electricity, steam, heating, and cooling consumed by your company. While you don't burn the fuel yourself, your energy demand drives these emissions. We use location-based emission factors from the EPA eGRID database for US-based companies.
Scope 3 — Value Chain
All other indirect emissions in your value chain — both upstream (purchased goods, business travel, employee commuting, waste) and downstream (use of sold products, end-of-life treatment). Scope 3 typically accounts for 70–90% of a company's total emissions footprint.
Spend-Based Accounting: How Your Data Becomes Emissions
Our Carbon Draft uses the spend-based method, one of the five calculation approaches recommended by the GHG Protocol's Scope 3 Technical Guidance. This method is particularly well-suited for small and medium businesses because it works with data you already have — your financial records.
The Formula
For every line item in your spend CSV, we determine the appropriate industry sector and multiply the dollar amount by the corresponding Environmentally-Extended Input-Output (EEIO) emission factor. This produces a kilogram-equivalent of CO₂ emissions for that line item. We then sum all line items across spending categories to produce your total footprint.
Step-by-Step Process
You Upload Your Spend CSV
Your CSV should contain at minimum a vendor/description column and a spend amount column. We accept exports from QuickBooks, Xero, FreshBooks, Sage, and most general ledger systems. Multi-currency CSVs are automatically converted to USD using prevailing exchange rates.
AI Classifies Each Line Item
Our classification engine analyzes each vendor name and transaction description, mapping it to one of the EPA's EEIO economic sectors. For example, "Delta Airlines — SFO to JFK" maps to the "Air transportation" sector, while "ConEd — Monthly Electric" maps to "Electric power generation." Known vendors from our curated database receive higher confidence scores.
Emission Factors Are Applied
Each classified sector has a corresponding emission factor from the EPA's EEIO database (measured in kg CO₂e per dollar of economic output). This factor represents the average carbon intensity of that sector across the US economy. We apply the factor to the USD spend amount to calculate the estimated emissions.
Report Generation
Your results are compiled into a professional PDF showing: total emissions (in metric tonnes CO₂e), breakdown by scope, breakdown by spending category, top emission sources, data quality scores per category, and methodology notes. Enhanced tier reports include a full transaction-level ledger with the classification rationale visible for every line item.
Emission Factors: Where the Numbers Come From
Our emission factors come from the US EPA's Supply Chain Greenhouse Gas Emission Factors for US Industries and Commodities, commonly known as the EPA EEIO (Environmentally-Extended Input-Output) database. This dataset provides emission factors expressed in kg CO₂e per dollar for over 1,000 US economic sectors.
EEIO models work by tracing the flow of money through the economy and mapping it to the environmental impacts associated with each step in the supply chain. When you spend $1,000 at a concrete supplier, the EEIO model accounts not only for the emissions from manufacturing the concrete itself, but also the emissions from mining the raw materials, trucking them to the factory, and generating the electricity used in production.
Example Emission Factors (kg CO₂e per USD)
| Sector | Factor | Meaning |
|---|---|---|
| Electric Power Generation | 1.42 | $1,000 spent ≈ 1,420 kg CO₂e |
| Air Transportation | 0.78 | $1,000 spent ≈ 780 kg CO₂e |
| Professional Services | 0.19 | $1,000 spent ≈ 190 kg CO₂e |
| Software Publishing | 0.09 | $1,000 spent ≈ 90 kg CO₂e |
Source: US EPA Supply Chain GHG Emission Factors v1.2. Factors shown are illustrative and simplified. Actual factors used in reports include more granular sub-sector classifications.
AI-Powered Spend Classification
The accuracy of a spend-based carbon estimate depends heavily on how well each transaction is classified into the correct emission sector. A flight classified as "Professional Services" instead of "Air Transportation" would dramatically underestimate emissions. Our classification system uses a multi-layered approach to maximize accuracy:
Known Vendor Database
We maintain a curated database of major vendors with pre-assigned sector classifications. When your CSV contains "Amazon Web Services," "United Airlines," or "Duke Energy," the classification is instant and high-confidence. This database covers thousands of common business vendors and is continuously expanded.
AI Fallback Classification
For vendors not in our database, we use large language model (LLM) analysis to interpret the vendor name and any additional description text. The AI considers contextual clues — "Joe's Trucking LLC" clearly maps to freight transportation, while "Smith & Associates, CPA" maps to accounting services.
Confidence Scoring
Every classification receives a confidence score. Known vendors from our database receive high confidence (85–100%). AI-classified vendors receive a dynamic score based on the clarity of the vendor name and description. Low-confidence classifications are flagged in your report so you can review them.
Data Quality Scoring
In line with the GHG Protocol's 2024 Scope 3 guidance on data quality, your report includes an overall Data Quality Score. This aggregates classification confidence, emission factor specificity, and the proportion of spend covered by known vendors — giving you and your stakeholders a clear picture of estimate reliability.
Limitations & What This Is Not
We believe in being upfront about the limitations of spend-based carbon accounting. Understanding what a Carbon Draft can and cannot do is essential for using it appropriately.
Estimates, Not Measured Data
Spend-based accounting uses industry averages to estimate emissions. It does not measure your actual fuel consumption, electricity usage, or specific supplier emissions. Two companies spending the same amount on electricity will receive the same emissions estimate, even if one uses 100% renewable energy. For precision on your highest-impact categories, activity-based data (kWh, liters, tonnes) is recommended.
Not a Substitute for Formal Audit
A Carbon Draft is a screening tool — it provides a directionally accurate estimate suitable for initial supplier disclosure requests, internal strategy planning, and identifying emission hotspots. For regulatory filings, formal verification under standards like ISO 14064-3, or Science Based Targets validation, you will need a full activity-based audit conducted by a qualified third-party verifier.
Price Sensitivity
Spend-based estimates are affected by pricing changes. If your electricity bill increases by 20% due to rate hikes (not increased usage), the spend-based estimate will show a 20% increase in emissions even though your actual energy consumption didn't change. We recommend year-over-year consistency in methodology to maintain comparability.
When to Use a Carbon Draft vs. a Full Audit
| Scenario | Carbon Draft | Full Audit |
|---|---|---|
| Enterprise customer requests your carbon data | — | |
| Internal strategy and hotspot identification | — | |
| First-year sustainability baseline | — | |
| CDP or CSRD formal disclosure | Screening only | |
| Science Based Targets validation | — | |
| Year-over-year reduction tracking | Directional |
Need a qualified auditor for a full engagement? Browse our Verified Auditor Directory to find certified sustainability professionals in your area.