How to Collect Actual Emission Data from Your Top 50 Suppliers Without Annoying Them

Scope 3 Category 1 — purchased goods and services — is the largest single emissions source for most industrial manufacturers, and it is the category where data quality is hardest to improve. The gap between "we have a spend-based estimate" and "we have supplier-verified product carbon footprints for our top 80% of spend" can represent years of supplier engagement work. This article is about shortening that gap without burning through your team's time or damaging supplier relationships in the process.

The honest starting point: mass questionnaire campaigns almost never work well. A general sustainability questionnaire sent to 300 suppliers produces low response rates, inconsistent data formats, and figures that may be calculated differently by every respondent. The approach described here is more surgical — tiered by supplier materiality, supported by automation for the long tail, and designed to actually increase supplier response rates.

Step One: Segment the Supplier Base Before Contacting Anyone

Before initiating any supplier outreach, run a spend analysis stratified by emission intensity. The goal is to identify which suppliers likely represent the top 80% of your Category 1 emissions — and this is not the same as the top 80% of your procurement spend, though it tends to correlate closely for materials-heavy purchasing.

Using spend-based emission factors from DEFRA or the ecoinvent database, you can generate a preliminary Category 1 estimate by supplier category. Apply these factors to your procurement invoice data, sorted by spend and supplier category, and you will typically find that:

• The top 20–30 suppliers by emissions-adjusted spend represent 65–75% of total estimated Category 1 emissions
• The next 50–100 suppliers represent a further 15–20%
• The remaining 200–300 suppliers, however numerous, collectively represent 10–15% — and their individual contribution per supplier is small enough that spend-based estimates are unlikely to introduce material error into the total

This segmentation directly determines your engagement strategy. It makes no sense to invest three months in getting a product carbon footprint from a supplier who represents 0.2% of your Category 1 emissions, while leaving your steel or chemical raw material suppliers on rough spend-based proxies.

Tier 1: Primary Data Requests for the Top 20–30 Suppliers

For your top-tier suppliers by emissions weight, the objective is supplier-specific emission data: either a product carbon footprint (PCF) calculated per ISO 14067 or EN 15804 (for construction products), or the supplier's verified total GHG inventory with sufficient granularity to attribute emissions to the purchased product or service.

A few practicalities that improve response rates dramatically:

• Frame as a commercial conversation, not a compliance burden. The request should come from procurement leadership, not the sustainability team alone. Buyers who have long-standing relationships can introduce the request in the context of supplier performance evaluation and future purchasing decisions — which suppliers take seriously.
• Specify exactly what data format you need. A vague "please send us your carbon footprint data" produces inconsistent responses. A structured request for: (a) total Scope 1 and 2 emissions in tCO₂e for the last reporting year, (b) production volume for the relevant product category, (c) emission factor applied to derive the PCF in kg CO₂e per unit, and (d) whether the data has been third-party verified — produces data you can actually use.
• Set a realistic timeline. Suppliers with mature sustainability programmes can typically respond within four to six weeks. Suppliers building their GHG inventory for the first time may need three to four months. Plan your supplier outreach calendar around these lead times, targeting your top-tier suppliers at the start of Q1 of the reporting year at the latest.

Veltmann Industrials GmbH, an automotive parts manufacturer in Stuttgart, ran their first structured Scope 3 Category 1 collection in 2024. Of their top 25 suppliers by emissions weight — representing approximately 68% of their estimated Category 1 total — they received usable primary data responses from 17 within the first three months. The other 8 required follow-up, and four of those were ultimately able to provide Scope 1+2 intensity data that could be used to derive a supplier-specific Category 1 figure per tonne of purchased material. Two suppliers declined entirely, and spend-based proxies were retained for those positions.

Tier 2: Activity-Based Methods for the Mid-Tier

For the mid-tier suppliers — typically positions 31 to 100 by emissions rank — the most practical approach is activity-based calculation rather than supplier engagement. This means using physical quantity data (kg of steel, litres of solvent, square metres of packaging material) from procurement records, combined with process-level emission factors from ecoinvent or GaBi databases.

Activity-based methods are more accurate than spend-based approaches because they are independent of price fluctuations — the emissions intensity of producing one tonne of primary aluminium does not change if the aluminium market price doubles. For commodity materials (metals, plastics, paper, chemicals), activity-based factors from well-maintained databases like ecoinvent 3.9 or GaBi SP38 typically carry an uncertainty range of ±20–40%, compared to ±60–100% for spend-based methods in volatile commodity sectors. This is a meaningful accuracy improvement without requiring direct supplier engagement.

Tier 3: Automated Spend-Based Estimates for the Long Tail

For the long tail — suppliers 101 and beyond — spend-based methods are entirely appropriate. These suppliers individually represent a small fraction of Category 1 emissions, and the additional accuracy from activity-based or supplier-specific methods would not materially change the total inventory. The key is to automate this calculation rather than running it manually each reporting cycle.

Procurement systems (SAP S/4HANA, Oracle Fusion, Coupa) typically store supplier invoice data with cost object and material group classifications. If those material groups can be mapped to DEFRA or ecoinvent sector codes — which requires a one-time setup of the mapping table — then every invoice processed through the procurement system automatically generates a spend-based emission estimate. The accuracy of this approach depends heavily on how granularly the material groups are defined: a supplier coded as "general manufacturing materials" will produce a less accurate estimate than one coded as "hot-rolled steel strip."

Where the Questionnaire Model Fails — and What to Do Instead

We're not saying that supplier questionnaires have no place in a Scope 3 programme — structured, targeted questionnaires sent to top-tier suppliers with specific data requirements are a legitimate primary data collection method. What we are saying is that generic sustainability questionnaires sent to large supplier populations as a primary strategy consistently fail to produce usable data.

The failure modes are predictable: response rates below 30% for undifferentiated outreach, inconsistent reporting boundaries (some suppliers include Scope 3 in their response, others only report Scope 1), figure formats that cannot be normalised without significant manual effort, and — critically — suppliers who provide whatever figure they expect you want to see rather than a rigorously calculated one. Switching to a tiered model, where only your top 20–30 suppliers receive a detailed primary data request and the rest are handled through automated estimation methods, produces higher-quality Scope 3 data with less manual effort across the reporting cycle.

Documentation for Assurance: What Auditors Will Check

For CSRD limited assurance purposes, your Scope 3 methodology documentation needs to cover: which GHG Protocol categories are included and excluded (with reasons for exclusions); which calculation method was applied for each category and supplier tier (primary data, activity-based, spend-based); which emission factor database was used and its version date; and how supplier-provided data was verified or cross-checked. The tiered approach described here is well-suited to this documentation requirement — it is transparent about the accuracy level applied at each tier and explains why higher-accuracy methods were not applied to the long tail. That transparency is more defensible than a homogeneous approach where a single methodology is claimed for all suppliers but applied inconsistently.