Scope 3 Data Maturity Guide
Sustainability teams face pressure to enhance the accuracy of their scope 3 footprint, which is key to unlocking real decarbonization.
This guide sets out principles for advancing data maturity, informed by scope 3 inventories across hundreds of enterprises.
Spend as a foundation – and its limits
Practitioners reach for industry-average spend factors such as USEEIO or EXIOBASE because they are easy to assign. The input is a purchase amount, which finance already holds and which usually arrives as an export of vendor, description, category and amount. Each line maps to an economic sector, each sector carries an emission factor in kilograms of CO₂e per dollar, and the multiplication covers the whole ledger. Two adjustments belong in that arithmetic and are often skipped: BEA industry-specific inflation adjustment, so spend and factors sit in the same dollar year, and currency conversion on foreign purchases.
Spend-based analysis is a solid foundation, and typically only takes a few minutes with modern software. The limitation shows up the moment the inventory is asked to support a decision. A sector factor describes an average producer selling at an average price, so it responds to the amount paid and to nothing else. Lower-carbon materials usually cost more per unit, which means buying them registers as an increase in emissions.
A developer switches two builds to a mix with 40% slag cement. Volume poured is unchanged. Each pair of bars is the concrete category total for the year before and after that switch, in tCO₂e, calculated first from spend and then from the plants' own declarations.
CONCRETE CATEGORY TOTAL, tCO₂e PER YEAR · COMMON SCALEThe GHG Protocol treats spend-based calculation as a legitimate method, and for most companies it is how the first inventory gets built. Coverage and scale are reliable. Anything that changes carbon without changing money stays invisible, which is a reporting inconvenience and an action-blocker.
Start here to evaluate hotspots, but work toward enhanced accuracy for material categories and those where you would like to register decarbonization.
The factor maturity quadrant
But what, precisely, does it mean to get more accurate? Maturity diagrams usually draw a single line from industry-average spend factors to supplier-specific data. Two separate things are changing along that line. One is whether the factor is applied to money or to a physical quantity. The other is whether it describes a sector or a particular supplier. Keeping them apart is useful, because the cost, the data source and the payoff differ on each axis.
Click any quadrant to see what sits in it. The general direction of travel is toward the top right, and categories legitimately sit in different quadrants at the same time.
Start in the bottom left to find the hot spots, then move up and to the right on the categories that warrant it. Be tactical about which ones those are, since every category does not need to end up in the same quadrant.
When in doubt, decompose
A Category 1 line item reading "concrete, 88,200 tCO₂e" is not granular enough to act on. Making it actionable means breaking the emissions into some combination of supplier, SKU, geography and project: calculate everything from spend to see where the footprint sits, then split the largest categories by supplier, match factors to each supplier's region and sub-sector, and push the biggest suppliers down to product level with quantities attached.
Sourcing product declarations and running targeted data requests comes after that, and it is the point where the inventory starts driving purchasing decisions rather than describing them.
The walkthrough below runs the sequence on one inventory. Use the arrows under the graphic to step through. Figures are composited from several portfolios.
Not every hot spot is worth the calories
Decomposition forces a choice about where to spend the next month. The instinct is to start with the largest category. The more useful question is whether a different purchase is available: another supplier, another product, or another specification with a genuinely different emissions profile. Where a substitute exists, better data tells you which option is lower and proves the switch afterwards. Where none exists, better data produces a more precise number and the same emissions.
That is why ranking by size alone disappoints. A large category can be fragmented across many suppliers, none individually large enough to justify a request or to answer one, or it can be one where the specification is set by a regulator, a tenant, or a commodity market. Two questions sort the list faster than a scoring model: can the data realistically be obtained, and is there a version of this purchase with a different footprint. Skip the second one and you end up with precise numbers attached to purchases nobody can change.
Substitution is also why supplier outreach pays for itself where it works. Asking what a product’s footprint is tends to reveal which of a supplier’s sites or mixes can already do better, so the request that produces the factor is often the start of the switch.
Spend your calories where a substitute exists. Not every category needs to reach the top right, and the ones left on modeled factors deserve a line in the disclosure rather than a permanent place on the backlog.
Leverage public sources before outreach
A useful share of supplier-specific data already exists in public form. Sweeping it first shrinks the outreach list to what genuinely cannot be found, which matters more than it sounds. Suppliers to large buyers field a version of the same questionnaire from every customer they have, and supplier fatigue is what we hear about when response rates drop. Every request you send that a public database could have answered spends goodwill you will want later, on the categories where you actually need supplier engagement.
Sourcing is only half the problem. A declaration arrives as a document, with a declared unit that may not match the purchase unit, a scope that may or may not include transport to site, a validity window, and a verifier. Turning it into a factor means reading it, checking it applies to the product actually bought, converting units, recording provenance, and attaching it to the purchase lines it covers. Done by hand, that is the bottleneck for the whole program.
Where outreach is genuinely needed, response rates rise with fewer and more specific requests, with multiple response formats rather than one template, with something the supplier values attached, such as preferred supplier status or procurement scoring that credits data provision, and with guidance for smaller suppliers who have no carbon accounting of their own. Tracking the share of the footprint on primary data, against a stated coverage target, keeps the program legible to whoever funds it.
Run the public sweep before the survey is designed, then size the outreach against what is left. A short list of well-chosen requests protects the relationships you will need when the conversation moves from data to decarbonization.
Conclusions
Recommended tooling
Many Scope 3 teams find it works best to have these capabilities in one integrated platform.
Put these principles to work
Gravity builds Scope 3 inventories for hundreds of enterprises — decomposing the categories that change decisions and leaving the rest at spend. See what that looks like for yours.