Most supplier data surveys fail in a predictable way.
You send a request to four hundred suppliers. Ninety reply. Of those ninety, a third send a corporate emissions figure instead of a product one. Several attach an environmental product declaration for a similar but different product. A few send a PDF with a number in it and no statement of what that number covers. Two send a spreadsheet in a format nobody can open.
Then you spend six weeks chasing clarifications and end up estimating anyway.
The instinct is to blame supplier maturity. Usually the survey is the problem. A well designed request gets usable product carbon footprint data from suppliers who have never produced it before. A vague one gets noise from suppliers who genuinely tried.
Here is how to build one that works.
First, Decide Which Artefact You Actually Need
Three different things get requested interchangeably, and they are not substitutes.
A product carbon footprint (PCF) covers greenhouse gas emissions for a specific product, per declared unit. For business to business exchange the standard scope is cradle to gate, everything from raw material extraction to the point the product leaves the factory. The relevant methodologies are the GHG Protocol Product Standard, ISO 14067, and the PACT Methodology.
An environmental product declaration (EPD) is a third party verified declaration covering multiple environmental impact categories, not just carbon. It follows ISO 14025, and in construction EN 15804+A2, and is built on a product category rule that defines how products in that category must be assessed. EPDs are comparable within a category, verified, and slower and more expensive to produce.
A life cycle assessment (LCA) is the underlying study, following ISO 14040 and 14044. It may cover the full life cycle rather than cradle to gate, may cover many impact categories, and may or may not be third party verified.
For Scope 3 Category 1 reporting, a cradle to gate PCF per declared unit is usually what you need. Asking a supplier for “your carbon data” gets you whichever of the three they happen to have, which is why response quality collapses.
Decide first. Then ask for that.
Step One: Segment Before You Send
A single survey issued to every supplier is the most common design error, and it produces the worst results in both directions. Large suppliers find it too crude. Small ones find it impossible.
Rank suppliers by contribution to your Scope 3 emissions, not by spend. They are different lists, and a small supplier of something carbon intensive matters more than a large supplier of low impact goods. Our explainer on Scope 3 emissions measurement covers how to run that screening exercise.
Then build three tiers.
- Tier one, your top twenty or so by emissions. These get a conversation, not a form. A named contact, a call, and a specific request tied to specific part numbers.
- Tier two, the next hundred or so. These get a structured template with pre-filled product details and a clear specification.
- Tier three, the long tail. These get either nothing or a very short request. Spend based estimates are adequate here until the category becomes material.
One legal point worth knowing if you or your parent report under CSRD. Omnibus I introduced a value chain cap limiting what in scope companies can require from value chain partners with 1,000 employees or fewer, and it applies to non-EU partners too. Gross Scope 1, 2 and 3 greenhouse gas emissions are carved out of that cap, so emissions data remains a legitimate ask. Broader environmental and social questions are not. Keep them in a clearly separate, clearly voluntary section.
Step Two: Specify The Boundary In The Request, Not The Response
If you do not state the boundary, every supplier picks their own, and none of the answers are comparable.
Your request should state, explicitly:
- System boundary. Cradle to gate unless you have a specific reason otherwise. This is the standard scope for business to business exchange.
- Declared unit. Per kilogram, per item, per litre, per square metre. Be specific per product line, because “per unit” means different things to different suppliers.
- Reference period. Which twelve months the data covers.
- Methodology and version. PACT Methodology V3 became the version for new PCF calculations on its release in April 2025, superseding V2. Ask which version was used, because V2 footprints calculated before withdrawal remain valid for a defined period and you need to track which is which.
- Global warming potential set. Specify it rather than assuming. Mixed GWP sets across a supplier base quietly corrupt aggregation.
- Biogenic carbon reported separately from fossil, if relevant to your materials.
- Allocation approach where a site produces multiple products.
This looks like a lot to put in a supplier request. It is roughly eight lines, and it removes almost all of the clarification cycles that otherwise consume your autumn.
Step Three: Draw The Boundary Before You Count Anything
Two boundaries matter, and getting them wrong is the most expensive mistake available.
The organisational boundary. The revised ESRS are more flexible here. Financial control remains the starting point, in line with the GHG Protocol, but operational control or equity share can be used instead. That flexibility exists for a practical reason: it lets you align your ESRS emissions boundary with the one you already use for group reporting or IFRS sustainability disclosures.
Take the option. One boundary across CSRD, SECR, UK SRS and any SBTi submission means one set of numbers to defend. Different boundaries for different frameworks means an annual reconciliation exercise nobody enjoys and auditors always notice.
The value chain boundary. Which entities, tiers and geographies sit inside your Scope 3 inventory. Document the logic, because it will be tested.
Step Four: Map The 15 Categories Against Materiality, Not Ambition
Under the GHG Protocol Corporate Value Chain Standard, Scope 3 splits into 15 categories, eight upstream and seven downstream. Under CSRD, what you disclose is gated by your double materiality assessment rather than by completeness for its own sake.
That distinction matters. Work through every category, record a materiality conclusion for each, and keep the reasoning. The revised standards also add a relief allowing activities to be excluded from metric calculations where they are not a significant driver of the relevant impacts, risks or opportunities and exclusion would not impair faithful representation. If you use it, you disclose that you used it.
One caution worth building for now. The GHG Protocol is separately revising the Scope 3 Standard, and its March 2026 progress update proposes a completeness floor requiring at least 95% of required Scope 3 emissions to be accounted for, with exclusions quantified and justified. That is draft, the 2011 standard still applies, and a final version is targeted for late 2027. But the two frameworks are pulling in slightly different directions, one toward materiality gating and the other toward completeness. Record your emissions and your exclusions in a way that can serve either, and you will not be rebuilding in 2028.
Step Five: Ask For The Metadata, Not Just The Number
A figure without provenance is unusable downstream, whether for assurance, for a customer audit, or for your own recalculation next year.
Collect these fields alongside every product carbon footprint data point:
| Field | Why it matters |
|---|---|
| Declared unit and quantity | Aggregation is impossible without it |
| System boundary | Tells you what is and is not included |
| Reference period | Determines when it expires |
| Methodology and version | Comparability and validity tracking |
| Sector rulebook or PCR applied | Sector rules add constraints, so downstream users need to know which applied |
| Primary data share | The proportion based on measured supplier data rather than secondary factors |
| Verification status and verifier | Whether it is self declared, verified, or partially verified |
| Validity or recalculation date | When you need to ask again |
That primary data share field is becoming the important one. The GHG Protocol’s March 2026 progress update on its Scope 3 Standard revision proposes disaggregating reported emissions by data type and disclosing verification status. It is a working draft, the 2011 standard still applies, and a final version is targeted for late 2027. But if you start collecting the provenance now, you will not be reissuing four hundred surveys later.
Step Six: Design For Systems, Not PDFs
A verified EPD as a PDF is a fine document and a poor data source. Someone has to read it and retype the numbers, and that person will make mistakes.
Collect structured fields wherever you can. PACT publishes technical specifications and a data model for exchanging product carbon footprints between systems, with a conformance programme for software providers, which means machine to machine exchange is realistic rather than aspirational for larger suppliers. Most organisations access it through a conformant platform rather than building an integration.
For everyone else, a locked spreadsheet template with validated fields beats a free text form, and both beat an email asking for “your PCF”. Accept the PDF as supporting evidence, attached to the structured record, not instead of it.
Step Seven: Make It Easy To Answer
Response rate is mostly a function of effort required.
- Pre-fill everything you already know. Product name, your part number, annual volume purchased. The supplier should only be adding what you cannot supply yourself.
- One product per row. Do not ask a supplier to describe a range in a paragraph.
- Include a worked example. A completed row for a fictional product removes most misunderstanding.
- Send it from a person. A named contact who can answer a question beats a no-reply address by a wide margin.
- Explain what happens to the data, who sees it, and whether it is confidential. Suppliers withhold data they think will be shared with competitors.
- Give a route for suppliers who have nothing. Most small suppliers have never calculated a footprint. Tell them what an acceptable interim answer looks like, such as activity data you can convert yourself, rather than leaving them to guess or ignore you.
- Set a realistic deadline. Four weeks for tier two, longer if you are asking for a first ever calculation.
Step Eight: Validate On Arrival, Not In Month Three
Build checks that run as responses come in.
Flag unit mismatches, figures that are orders of magnitude away from category norms, missing boundary statements, expired validity dates, and the very common case of a corporate footprint submitted as a product one. A supplier who gave you the wrong thing in week two will usually fix it in week three. The same supplier contacted in month four has moved on.
Step Nine: Give Suppliers A Reason
This is the step that determines everything else, and it belongs to procurement rather than sustainability.
Suppliers respond when the request carries commercial weight. That means embedding product carbon footprint data into supplier onboarding, RFQs and tender scoring rather than sending an annual standalone survey. It means feeding something back: a benchmark against category peers, or recognition in preferred supplier status.
An annual email from a sustainability inbox gets a low response rate almost regardless of design quality. The same request, referenced in a contract renewal, gets answered.
Why This Is Getting Urgent
Product level data is moving from a reporting nicety to a condition of sale.
The EU Digital Product Passport registry went live in July 2026. Battery passports apply from February 2027 under the Batteries Regulation. Under the ESPR working plan, iron and steel is the first product group for a delegated act, with textiles, tyres and aluminium following, then furniture and mattresses later in the decade. Construction products run on a separate timetable under the revised Construction Products Regulation. Add CBAM for relevant materials, and customers with CSRD obligations pushing for granular Scope 3 figures, and the direction is clear.
If you sell physical products into the EU, or supply someone who does, the question is no longer whether you will need product level environmental data but which of your product lines needs it first.
Common Mistakes
- One survey for every supplier. Segment, or you optimise for nobody.
- Asking for “carbon data”. Name the artefact, the boundary and the unit.
- Collecting numbers without provenance. Unusable for assurance and impossible to update.
- PDF only collection. Guarantees manual rekeying and errors.
- No procurement involvement. The single biggest driver of response rate sits outside the sustainability team.
- Treating it as an annual campaign. It should be a permanent process attached to onboarding and renewals.
How Enistic Helps
We produce EPDs, life cycle assessments, TM65 assessments and product carbon footprints for clients across manufacturing and construction, and we run the supplier data collection that sits underneath them. Our AI enabled platform is built and maintained in house, which cuts data collection effort by up to 80% and keeps the boundary, methodology version, factor set, source document and verification status attached to every figure, so the data holds up when a customer or an assurance provider starts asking questions.
- 5,000+ compliant reports delivered.
- 100% compliance record, with no failures.
- 11.7% audit rate, against an industry average of 33%.
- 98% client renewal rate.
Our carbon consultants and auditors are in house, and one is named to your account. They will run supplier engagement alongside your procurement team rather than handing you a portal and a deadline, all covered by one simple monthly payment with unlimited expert support.
A good supplier survey is not a form. It is a process that gets better data every year with less effort, and it is worth building once, properly.
Book a demo and we will review your current supplier request, show you where the response quality is being lost, and help you rebuild it around the products that actually matter.