Scope 3 Estimation When Your Suppliers Will Not Answer

Supplier emissions data is sparse, inconsistent, and often simply absent. Here is how to build a defensible Scope 3 estimate anyway, using proxy models, spend-based fallbacks, and supplier engagement tactics that actually move the needle.

Why Scope 3 Category 1 Is the Hardest Emissions Problem You Will Face

Purchased goods and services (Category 1) typically account for 40–80% of a manufacturer's total carbon footprint, yet the data quality is the worst of all fifteen Scope 3 categories. The GHG Protocol's Corporate Value Chain (Scope 3) Accounting and Reporting Standard explicitly acknowledges this and allows for estimated data — but "estimated" still needs to be methodologically sound if you are preparing for CSRD disclosure, CDP submission, or SBTi target validation.

The core problem is structural. Your Tier 1 suppliers have their own Tier 2 suppliers, who have their own suppliers, and almost none of them have a verified product-level carbon footprint (PCF) ready to hand you. The suppliers who do respond to data requests often send spreadsheets that are internally inconsistent, conflate market-based and location-based electricity figures, or simply report Scope 1 and 2 without any allocation to your specific purchase volumes.

You need an estimation architecture that works even when the inbox stays empty.

What Are Your Fallback Data Sources When Suppliers Go Silent?

When primary data is unavailable, the GHG Protocol allows three classes of secondary data, in rough descending order of preference:

Spend-based emission factors map a spend value in a commodity category to a CO₂e figure. Exiobase 3.8 and the US EPA's USEEIO v2.0 model are the two most commonly used multi-regional input-output (MRIO) databases. USEEIO gives you kg CO₂e per USD of output by NAICS code. Exiobase gives you similar figures at country-of-origin resolution for 163 sectors across 49 regions. If you know what you bought, from which country, and what you paid, you can produce a number.

Physical intensity factors are preferable where you can get them. The UK DESNZ/BEIS conversion factor tables, Ecoinvent 3.10, and the International Energy Agency's emissions factors by country are the primary sources. If a supplier makes injection-moulded polypropylene parts and will not give you a PCF, you can estimate from kg of material shipped multiplied by an Ecoinvent process factor for PP injection moulding (roughly 2.8–3.2 kg CO₂e per kg for a European average grid, higher for coal-heavy grids).

Industry average PCFs from databases like the Responsible Business Alliance's materials tool, the EcoInvent unit processes, or sector-specific LCA studies published by trade associations. These are coarser but better than a spend-based fallback when you have weight or volume data.

The practical decision tree:

Data available from supplier	Recommended method
Verified PCF with allocation methodology	Use primary data directly
Scope 1+2 emissions + revenue	Attribute pro-rata by your spend share
Material type and weight only	Physical intensity factor (Ecoinvent or BEIS)
Spend value and country only	MRIO spend-based (USEEIO or Exiobase)
Nothing	Spend-based with a worst-case sensitivity range

Apply the most granular method you can for each line item. Do not use a single method across all categories; the error compounds badly.

Building a Proxy Model That Survives an Auditor's Scrutiny

An estimate is auditable if the assumptions are explicit, the data sources are cited, and the uncertainty is quantified. That is the entire standard.

Practically, this means your estimation model should:

• Record which method was used per supplier or commodity, so reviewers can see where primary data ends and proxy data begins.
• Carry a confidence tier (primary / secondary-physical / secondary-spend) for each line item.
• Document the emission factor version and publication date, because BEIS factors update annually and Ecoinvent 3.10 differs materially from 3.8 for some processes.
• Produce a sensitivity table showing what happens to your Category 1 total if the proxy factors are 20% higher than assumed. If your total Scope 3 does not move by more than 5–10% under that shock, your model is structurally sound.

The tool stack here is not exotic. A well-structured PostgreSQL schema with one row per procurement line item, joined to a reference table of emission factors, will outperform a sprawling Excel workbook for anything above a few hundred supplier relationships. You can query uncertainty ranges, group by commodity or geography, and pipe results into whatever reporting layer you need without manual aggregation.

How Do You Actually Get Suppliers to Respond?

Engagement tactics matter, but the framing matters more. Suppliers do not respond to sustainability questionnaires because they perceive them as low-priority administrative burden with no commercial consequence.

The tactics that change that calculus:

Make it a commercial signal. Procurement teams that tie supplier evaluation scores to carbon data response rates see meaningfully higher engagement. This does not require a carbon tax; it requires the supplier to believe that non-response affects their relationship with you.

Reduce the friction of responding. A request for a full lifecycle assessment is ignored. A request for three figures (total Scope 1+2 emissions last year, total revenue last year, and the NAICS or HS code for what they sell you) gets answered more often. You can compute the rest.

Use CDP's supply chain programme. 280+ companies with a combined purchasing power of over USD 6.4 trillion use CDP's supply chain module to request emissions data from suppliers through a standardised questionnaire. If you are a large enough buyer, your suppliers already have CDP accounts and the request lands in a familiar interface rather than a cold email.

Offer your proxy model as a starting point. Send the supplier your estimate of their emissions share in your value chain and ask them to correct it. People correct wrong numbers. They rarely fill in blank forms.

Realistically, you will get primary data from 20–40% of your spend-weighted supplier base after a focused engagement campaign. The remaining 60–80% stays on proxies. That is an acceptable outcome for most disclosure frameworks, as long as you disclose the coverage rate.

Handling Multi-Tier Supplier Opacity

Most Scope 3 guidance stops at Tier 1. The GHG Protocol's Scope 3 standard does not require Tier 2 mapping, but SBTi's FLAG guidance and the EU's Corporate Sustainability Due Diligence Directive (CS3D) are beginning to push expectations further upstream, particularly for high-deforestation-risk commodities.

For multi-tier problems, the practical options are:

• Sectoral decarbonisation approaches (SDA): Use sector-level benchmarks to set targets without resolving individual supplier chains. SBTi publishes SDAs for cement, aluminium, iron and steel, chemicals, and pulp and paper.
• Spend-based MRIO with origin country resolution: Exiobase's country-by-sector granularity can approximate Tier 2 effects implicitly, because the input-output tables model inter-industry flows within and between economies.
• Supplier audit requirements in contracts: For high-risk commodity categories (palm oil, soy, cattle products, timber), the deforestation risk is the emissions risk. Contractual requirements for certification (RSPO, FSC, RTRS) are a proxy for forest carbon integrity.

Do not try to model Tier 2 from scratch without a dedicated data engineering effort. The marginal accuracy gain rarely justifies the cost unless you are a large consumer goods company with regulatory exposure to deforestation-linked commodities.

Conclusion

The absence of supplier data does not excuse absence of an estimate. Pick the highest-resolution proxy method your data supports, document your assumptions, quantify your uncertainty, and disclose your coverage rate. That is what CSRD's ESRS E1-6 and the GHG Protocol both ask for.

The next concrete step: audit your procurement data now. If you can extract material type, weight, and country of origin for your top 20 spend categories, you can build a defensible Category 1 estimate in four to six weeks. If you cannot extract that data cleanly, the data infrastructure problem is upstream of the emissions problem and needs to be solved first.

FAQ

Can I use spend-based emission factors for my entire Scope 3 Category 1 if I have no supplier data at all?

Yes, and many companies do for their first disclosure cycle. Spend-based estimates using USEEIO or Exiobase are explicitly permitted by the GHG Protocol. The requirement is that you disclose the method and its limitations. Auditors will expect you to migrate toward physical intensity or primary data over subsequent reporting periods.

What is a reasonable coverage rate for primary Scope 3 data?

CDP and SBTi do not mandate a specific percentage, but in practice, companies aiming for SBTi validation target primary or verified secondary data for at least 67% of spend-weighted Category 1 emissions. Coverage below 50% will attract questions from assurance providers, particularly under CSRD's limited assurance requirement.

How often do I need to update my emission factor reference tables?

BEIS/DESNZ conversion factors update annually, usually in June. Ecoinvent releases major versions roughly every 18–24 months. IEA country electricity factors update annually. Build your reference table as a versioned dataset, not a static lookup, so historical disclosures remain reproducible when factors change.

Is a product-level carbon footprint (PCF) from a supplier automatically trustworthy?

No. A PCF is only as reliable as its methodology and boundaries. Check whether it follows ISO 14067:2018 or the PACT Pathfinder Framework v2.0, whether system boundaries include raw material extraction, and whether it has been third-party verified. Unverified, self-declared PCFs should be treated as indicative, not primary data.

What is the difference between a Scope 3 estimate and a Scope 3 calculation?

Nothing meaningful in practice. Every Scope 3 figure involves assumptions about allocation, system boundaries, and emission factors. The GHG Protocol uses "calculation" broadly to include both measured and estimated inputs. The distinction that matters is whether your assumptions are documented and your uncertainty is disclosed, not whether you called it an estimate or a calculation.

Have a project in mind? Contact Sodio Technologies to discuss your requirements and explore the right technology solution for your business.