Spend-based vs activity-based emissions: which method to use, and when to switch

Spend-based emissions estimate a footprint by multiplying how much you spent by an emission factor per dollar; activity-based emissions multiply a physical quantity, like kilowatt-hours or tonne-kilometers, by a factor per unit. Spend-based is fast and complete because you always have the spend, but coarse. Activity-based is far more accurate but needs primary data you often do not hold. The right answer for most companies is not one or the other: use spend-based to screen the whole footprint, then graduate your biggest lines to activity-based data.

That trade-off is the whole decision, and getting it right is what separates an inventory that costs a quarter to build and still gets challenged from one that is defensible and proportionate. This guide explains how each method works, what auditors expect, and a practical rule for when to switch.

How the spend-based method works

The spend-based method takes each line of your accounts payable, groups it by category, and multiplies the dollar amount by a spend-based emission factor expressed in kilograms of CO2-equivalent per dollar. Those factors come from environmentally extended input-output models such as the EPA's USEEIO or EXIOBASE, which estimate the average emissions intensity of a whole sector's output. Spend $100,000 with a steel supplier, apply the factor for primary metal manufacturing, and you get an estimate without ever asking the supplier a question. The full procedure is in our guide to calculating Scope 3 emissions from spend.

Its strength is coverage. Spend is the one dataset that spans your entire value chain, so the spend-based method can put a number on all fifteen Scope 3 categories at once. Its weakness is resolution. Two suppliers in the same sector get the same factor even if one runs on hydropower and the other on coal, so spending more looks like emitting more, and genuine efficiency is invisible. That makes spend-based data good for finding hotspots and terrible for showing progress.

How the activity-based method works

The activity-based method uses physical activity data: the actual kilowatt-hours a supplier consumed, the tonne-kilometers a shipment traveled, the kilograms of a specific material you bought. You multiply that quantity by an emission factor per physical unit, ideally a supplier-specific one. Because it reflects what physically happened rather than what you paid, it captures real differences between suppliers and rewards genuine reductions.

The cost is data. Activity data usually lives in someone else's system, so you have to collect it: send supplier surveys, gather bills of materials, pull freight weights and distances. Much of it starts as a PDF or a scanned document, so teams building activity inventories often need to extract the line items from supplier invoices and documents before the numbers are usable. That collection effort is why you do not want to do it for every line, only the ones that matter.

Spend-based vs activity-based: the trade-off in one table

	Spend-based	Activity-based
Input	Dollars spent per category	Physical quantity (kWh, kg, tonne-km)
Factor source	Input-output models (EPA USEEIO, EXIOBASE)	Process or supplier-specific factors
Coverage	Complete: every line you paid	Only where you have activity data
Accuracy	Coarse, sector-average	High, supplier-specific
Effort	Low: you already have the spend	High: collect primary data
Shows real reductions	No, tracks spending	Yes, tracks physical change
Best for	Screening, boundary-setting, hotspots	Managing and reporting your largest sources

What auditors and standards expect

The GHG Protocol explicitly permits both methods and expects a hybrid in practice: screen with spend-based data, then improve the material categories with activity data. Assurance providers do not demand activity-based data everywhere. What they demand is that your method is documented, consistent, and proportionate to the size of each source, and that every figure records the factor and version behind it so it can be reproduced. A spend-based estimate on a small category is fine and defensible; a spend-based estimate on the category that is 60 percent of your footprint, with no plan to improve it, is the finding an auditor writes up.

The practical rule: when to switch from spend to activity

Do not agonize over the choice line by line. Run a spend-based screen across everything first, rank the categories by estimated emissions, and then graduate only the top of that list to activity data. A common pattern is that three or four purchased-materials categories drive the majority of Scope 3, so those are where supplier-specific data earns its collection cost. Everything in the long tail can stay spend-based until it grows or a customer asks. The sequence is: measure everything roughly, then measure the few things that matter precisely.

This is exactly how our carbon accounting software is built to work. It classifies your accounts payable file into scopes and categories with a spend-based estimate immediately, shows you which lines dominate, and lets you replace those estimates with activity data as you collect it, keeping the factor and the source document attached to every figure. You get a complete number on day one and a more accurate one over time, without rebuilding the inventory each step. If your reporting trigger is California, the same inventory feeds SB 253 reporting directly.

You can run the spend-based classification on your own accounts payable lines in the demo, before any email address changes hands, and see where your hotspots sit.