Tier 2 Supplier Carbon Data: What to Request and How to Use It
Retained synthetic and unbranded same-article featured image. It does not show a real customer site or an emissions result.
Diagram text description
Tier 2 supplier carbon data is useful when it explains an upstream input that your current purchase estimate represents poorly. Its value depends on where it belongs in the calculation. Receiving another emissions figure does not create another emissions category, and attaching a supplier's name does not establish that a number fits your product.
Consider a buyer purchasing an assembled metal component. The direct supplier makes the component; an indirect supplier produces the metal. A material record from that indirect supplier might improve the component model, but the existing cradle-to-gate estimate may already include metal production. The task is to find the overlapping assumption, establish whether the new evidence can replace it, and retain the parts of the model that still apply.
This guide follows that task from purchase selection to a replacement ledger. It distinguishes an unanswered request from measured data, a plant total from a product footprint, and a change in evidence from a reduction in physical emissions. You can use the record below with one purchase before expanding the approach across a supply chain. The aim is a calculation whose inputs can be explained, rather than the largest possible collection of supplier responses.
What Tier 2 means for a product and for Scope 3
Tier describes the supplier's relationship to the reporting company. Your Tier 1 supplier sells directly to you; its supplier is your Tier 2 supplier for that chain. A material producer can occupy a different tier in another customer's chain. The label describes distance through commercial relationships, rather than an inherent property of a factory or material.
For a component purchase, the chain might run from material producer to component supplier to reporting company. Materials flow towards the buyer. The carbon evidence needed to explain those materials can travel through the direct supplier, with enough product identity to connect the response to the purchase. Neither direction establishes ownership of the upstream facility or permission to obtain its confidential operating records.
Material moves from the indirect material producer to the direct component supplier and then to the reporting company. Category classification is a separate question described in the section.
Diagram text description
Scope 3 categories answer a different question: what activity is the reporting company accounting for? Purchased goods and services generally belong in Category 1 unless another upstream category applies. Category 2 concerns capital goods; it does not mean second-tier suppliers. For a production component, the relevant upstream production can extend through several supplier tiers while remaining within the same purchased-goods boundary. Conversely, knowing a supplier's tier does not settle how a capital purchase should be classified.
Keep the relationship map and accounting map as separate fields. The relationship map identifies who may hold evidence and who can request it. The accounting map identifies which production stages the purchased item requires. This separation also prevents a common freight mistake: the GHG Protocol Category 1 guidance treats transport from the direct supplier to the reporting company separately in Category 4 where its stated conditions apply. A supplier's full delivery figure therefore needs examination before it is used as a production factor.
Do not add Tier 2 emissions on top of an existing footprint
A cradle-to-gate purchased-component estimate can already contain the material producer's emissions. Adding a new material figure to that whole estimate would count the same upstream production twice if the boundaries overlap. The risk persists even when the material response is detailed, current and supplied directly by the producer. Better provenance does not cancel a duplicated boundary.
A container labelled existing component footprint encloses upstream material and other upstream production. Invisible model relations record containment; no arrow adds a separate supplier total.
Diagram text description
There are two different replacement cases:
- In a disaggregated component model, the generic material contribution is identifiable. A suitable Tier 2 record may replace that contribution while the remaining production stages stay in place.
- In an aggregate component factor, the material contribution cannot be separated reliably. Subtracting an invented material share and inserting the new figure would create a calculation with an unsupported remainder. Obtain a breakdown, rebuild an appropriate model, or retain the new record as supporting evidence until reconciliation is possible.
Ask what the new number includes before deciding which case you have. A material intensity might cover production only up to the material supplier's gate. Another response might include processing, packaging or transport already represented elsewhere. A corporate total might cover many products and operations. Similar units in a spreadsheet are insufficient: two values expressed as kilograms of carbon dioxide equivalent can refer to very different quantities and activities.
The safe conceptual operation is to replace a matching represented stage, not to add a supplier tier. That does not guarantee a lower revised result. A more representative input may increase the estimate, reduce it, or leave it broadly similar. The defensible benefit is improved representation of the purchased item. Any claim that the supplier reduced emissions needs evidence of a physical or operational change, distinguished from changes in coverage, allocation and calculation method.
Choose the purchase where better data could change a decision
Start with a purchase where an uncertain upstream stage matters to a real decision. A long supplier list is a poor starting point because it says little about which missing input could alter a product choice, sourcing discussion or inventory estimate. The World Economic Forum's supplier segmentation guidance supports using a baseline to identify priorities; the baseline is a screening tool, not proof that its largest estimated line is accurately measured.
Three questions help narrow the first request. Is the material or process plausibly significant within the purchased item? Could the available evidence change the choice under consideration? Can the direct supplier connect a response to the actual product, source and period? A consequential but untraceable record may be less useful today than a smaller, well-defined stage that the team can reconcile properly.
For example, a buyer considering two metal routes may need material-production evidence before asking for more detail about minor packaging inputs. That is a prioritisation hypothesis, not a universal ranking of metal and packaging impacts. The product's quantities, processes and existing model determine the next investigation. Procurement can help identify the source and specification; the carbon analyst can explain which model assumption has enough uncertainty to affect the decision.
Set the scope in a short request statement: the purchased part and revision, the material or process in question, the period of supply, and the intended use of the evidence. Include what the request will not settle. Material-production data alone will not establish the footprint of the finished component or prove that two complete products are comparable. This makes it possible for a supplier to return a focused answer instead of a generic sustainability presentation.
If the decision is insensitive to the missing detail, document that finding and move to a more consequential gap. Primary-data collection has a cost for both parties. Its purpose is to improve an identified decision or accounting need, rather than to replace every secondary input merely because a direct response sounds preferable.
Ask your Tier 1 supplier to open the evidence chain
Your direct supplier is usually the practical bridge to an indirect supplier. It can identify the material source, connect it with the component specification and explain the commercial arrangements for sharing evidence. A buyer should not assume that a procurement relationship with Tier 1 grants access to every Tier 2 facility's energy bills, production volumes or other confidential records.
Send the defined request to the Tier 1 contact and agree who will carry it forward. Explain whether you need a product footprint, an allocated process result or activity information for a model. Ask how the returned record will identify the relevant product and how its permitted use will be communicated. A supplier may be able to share a bounded summary without disclosing commercially sensitive plant details; the summary still needs enough methodological explanation to be evaluated.
Permission to share and fitness for calculation are separate decisions. An authorised response can still have the wrong boundary. A technically detailed response may have restrictions that prevent onward disclosure. Keep both facts visible rather than treating possession of a file as unrestricted permission to use or publish its contents.
Two visibly separate branches from Tier 1 permission lead either to a permissioned scoped response or to a restricted-data gap. The gap is not presented as an emissions result.
Diagram text description
Where permission is unavailable, record the restriction as a data gap. It is not evidence of high emissions, low emissions or poor environmental performance. Nor should the team fill the silence with a supposed supplier-specific result. The request can instead seek a suitable aggregated record, a clarification of the relevant stage, or confirmation that the existing estimate must remain for now.
Agree a return route as well as a request route. The carbon analyst may have questions about allocation that the procurement contact cannot answer, while the supplier may need the buyer to clarify the purchased specification. Identifying a technical owner and a commercial contact avoids repeatedly forwarding the same spreadsheet without resolving the mismatch.
What to request from the Tier 2 supplier
Request a record that identifies what the value represents before requesting more decimal places. At minimum, it should connect the material or product to a plant or stated origin, a production period, a quantity basis and a defined production boundary. A result without these attributes may be useful background, but it cannot automatically replace a purchased-product input.
The quantity basis deserves particular care. An intensity such as kg CO2e per kg of material differs from a total in kg CO2e for a batch. Both differ from electricity consumption in kWh, which is activity information requiring an appropriate emissions calculation. Ask whether the quantity refers to input material, saleable output or another basis, and whether losses are already included. Otherwise the same loss can enter once in the supplier's allocation and again in the buyer's material requirement.
Request the method and its supporting references. For an intensity, obtain the factor source or calculation version, the included processes and the treatment of shared production. Where a plant makes several products, the allocation explanation should show how the reported share relates to the supplied material. A facility's annual emissions divided by an unsuitable output quantity does not become a product footprint simply because the division produces a convenient unit.
The evidence package should also state the reporting period, the data owner, revision date and any assurance scope actually applicable to the result. If an assurance statement covers a corporate inventory, do not extend it to a product result without support. If some inputs are measured and others modelled, identify them separately. A useful response can contain both; calling the entire record measured would conceal the assumptions that matter when comparing it with the old estimate.
Finally, ask what changed since the previous version. A different plant, product mix, electricity assumption or boundary can explain a revised value without demonstrating an improvement in the same activity. The owner should be able to say which purchases the record can represent and what event would make it obsolete. That applicability statement is often more useful to the buyer than an unsupported claim that the data is universally accurate.
How to fill a gap without pretending the data is measured
A missing Tier 2 response does not prevent a transparent Category 1 estimate. The GHG Protocol describes supplier-specific, hybrid, average-data and spend-based methods. Their names describe how evidence enters the calculation; they are not a simple quality league table. Its guidance explicitly cautions that more specific supplier data may still be less accurate if its sources, granularity or allocation are weak.
Use the following sequence to choose a defensible treatment for the selected purchase:
- Check whether a suitable product-level cradle-to-gate inventory is available from the supplier. Examine its coverage and quality before adopting a supplier-specific approach.
- If usable supplier activity information covers only part of production, consider a hybrid model. Identify the directly supported portion and the upstream gaps filled with secondary data; make the pieces mutually consistent.
- Where physical purchase quantities and representative secondary factors are available, consider an average-data estimate. Document how the material, technology, geography and period correspond to the purchase.
- Where only economic purchase information can support the estimate, assess an appropriate spend-based approach. Keep its currency and price basis explicit, and avoid presenting price movements as physical efficiency gains.
These are evaluation steps, not an instruction to force every purchase through an automatic descending ladder. Different purchases can justify different methods. The practical question is whether the chosen evidence is good enough for the inventory purpose and whether an available alternative would represent the activity more reliably.
| Evidence available | Possible treatment | Gap to keep visible |
|---|---|---|
| Suitable product-level cradle-to-gate inventory | Evaluate supplier-specific method | Applicability and quality of the supplied result |
| Relevant supplier activity plus incomplete upstream detail | Evaluate hybrid method | Stages represented by secondary inputs |
| Physical quantity with representative secondary factor | Evaluate average-data method | Difference between the factor and actual production |
| Purchase expenditure with suitable economic factor | Evaluate spend-based method | Currency, price basis and broad sector representation |
Table 1, shown one row at a time. Every field is retained.
Row 1
- Evidence available
- Suitable product-level cradle-to-gate inventory
- Possible treatment
- Evaluate supplier-specific method
- Gap to keep visible
- Applicability and quality of the supplied result
Row 2
- Evidence available
- Relevant supplier activity plus incomplete upstream detail
- Possible treatment
- Evaluate hybrid method
- Gap to keep visible
- Stages represented by secondary inputs
Row 3
- Evidence available
- Physical quantity with representative secondary factor
- Possible treatment
- Evaluate average-data method
- Gap to keep visible
- Difference between the factor and actual production
Row 4
- Evidence available
- Purchase expenditure with suitable economic factor
- Possible treatment
- Evaluate spend-based method
- Gap to keep visible
- Currency, price basis and broad sector representation
Do not merge the labels measured, supplier-provided and verified. They answer different questions about an input. A supplier may provide an estimate, a measured activity may require modelling, and verification may cover only a stated subject. Preserve that distinction in the record so that the next analyst can improve the weakest assumption rather than repeat a request already answered.
If there is no credible uncertainty range, describe the limitation in words rather than inventing a percentage. Identify the uncertain stage, why the proxy may differ and the decision it could affect. This is less precise numerically but more honest than an apparently scientific interval with no basis.
Reconcile the new figure before replacing an estimate
The following illustrative example uses no emissions values or factors. A buyer purchases a metal housing from an assembler. The existing model separately represents a generic material-production input and the assembler's other upstream production stages. The material producer returns a record for the alloy and production route used in that housing. This example assumes that the connection to the purchased part can be established; it does not describe a real supplier or measured saving.
First, compare the old material input and the returned record at their common quantity basis. Establish whether both cover the same production stage, whether the material specification and period fit, and whether either includes losses or further processing already represented elsewhere. The returned file's title is insufficient. The explanation of the boundary determines whether the two inputs overlap.
In this example the generic material proxy is replaced by the suitable Tier 2 material input; the separately modelled assembly stages remain. No additional Tier 2 total is appended to the component result. The analyst retains the original model as revision A, creates revision B with the replacement and records the affected purchase population. The change note identifies the substituted input, its source version and the reason it represents those purchases better.
Before and after columns align the generic material proxy with the Tier 2 evidence that replaces it. The lower row explicitly keeps other stages unchanged in this example. Equal box size does not express emissions quantities.
Diagram text description
Now consider the exception. Suppose the old model contains only one aggregate component factor and the supplier cannot explain its material share. The analyst cannot subtract an assumed fraction merely to accommodate the new record. Revision B remains a proposed change while the team seeks a compatible breakdown or rebuilds a supported model. The existing estimate and its limitation remain visible. That outcome is a resolved accounting decision even though it has not produced a new reported number.
A second exception arises if the returned material record includes processing also present in the assembly stage. The replacement needs another boundary reconciliation before use. Retaining all assembly stages unchanged would then preserve an overlap. The diagram illustrates only the first, matched case; it is not permission to keep every remaining input without examining its scope.
When the revision is accepted, describe the difference as a method or evidence update unless separate evidence establishes a physical reduction. Preserve enough history to distinguish later supplier improvements from this initial change in representation. If the update affects a previously reported inventory, apply the organisation's relevant recalculation and reporting policy rather than silently overwriting the historical result.
A decision record for procurement and carbon reviewers
Use one ledger for each proposed replacement, with references to the underlying records rather than copied files that can drift apart. The template below is an organisational aid, not a certification form. Bracketed fields are intended to be completed; a blank or unresolved field is a visible action item, not implied approval.
| Ledger field | Entry to complete | Decision supported |
|---|---|---|
| Purchase identity | [Part, revision, Tier 1 supplier, purchase period] | Which purchases the decision can affect |
| Upstream identity | [Material, specification, Tier 2 source, plant or origin] | Whether the evidence matches the input |
| Old representation | [Model revision, old method, factor reference, represented stage] | What is being retained as the baseline |
| Quantity and units | [Purchase quantity, input or output basis, intensity denominator, conversion] | Whether quantities can be reconciled |
| New evidence | [Document reference, version, period, method and allocation] | What supports the proposed substitute |
| Boundary comparison | [Included stages, exclusions, losses and overlapping processes] | What must be removed or retained |
| Replacement decision | [Accept, hold or reject; affected model input and reason] | Whether a calculation change is justified |
| Evidence owner | [Supplier technical contact and internal decision owner] | Who can resolve questions and revisions |
| Uncertainty | [Measured and modelled portions, limitations and unresolved gaps] | How much reliance the intended use permits |
| Change history | [Old revision, new revision, decision date and rationale] | How the revised result differs from history |
| Review trigger | [New source, plant, specification, method or reporting-period change] | When this decision must be reconsidered |
Table 2, shown one row at a time. Every field is retained.
Row 1
- Ledger field
- Purchase identity
- Entry to complete
- [Part, revision, Tier 1 supplier, purchase period]
- Decision supported
- Which purchases the decision can affect
Row 2
- Ledger field
- Upstream identity
- Entry to complete
- [Material, specification, Tier 2 source, plant or origin]
- Decision supported
- Whether the evidence matches the input
Row 3
- Ledger field
- Old representation
- Entry to complete
- [Model revision, old method, factor reference, represented stage]
- Decision supported
- What is being retained as the baseline
Row 4
- Ledger field
- Quantity and units
- Entry to complete
- [Purchase quantity, input or output basis, intensity denominator, conversion]
- Decision supported
- Whether quantities can be reconciled
Row 5
- Ledger field
- New evidence
- Entry to complete
- [Document reference, version, period, method and allocation]
- Decision supported
- What supports the proposed substitute
Row 6
- Ledger field
- Boundary comparison
- Entry to complete
- [Included stages, exclusions, losses and overlapping processes]
- Decision supported
- What must be removed or retained
Row 7
- Ledger field
- Replacement decision
- Entry to complete
- [Accept, hold or reject; affected model input and reason]
- Decision supported
- Whether a calculation change is justified
Row 8
- Ledger field
- Evidence owner
- Entry to complete
- [Supplier technical contact and internal decision owner]
- Decision supported
- Who can resolve questions and revisions
Row 9
- Ledger field
- Uncertainty
- Entry to complete
- [Measured and modelled portions, limitations and unresolved gaps]
- Decision supported
- How much reliance the intended use permits
Row 10
- Ledger field
- Change history
- Entry to complete
- [Old revision, new revision, decision date and rationale]
- Decision supported
- How the revised result differs from history
Row 11
- Ledger field
- Review trigger
- Entry to complete
- [New source, plant, specification, method or reporting-period change]
- Decision supported
- When this decision must be reconsidered
Separate an evidence update from an accepted calculation change in the workflow. A new document can be stored while the decision remains on hold. Procurement may confirm that the material came from the stated source, while the carbon specialist still needs an allocation explanation. Those are complementary contributions; neither should disappear behind a single tick labelled data received.
For the housing example, the boundary-comparison entry would identify the generic material stage being displaced and explicitly retain the non-overlapping assembly stages. The uncertainty entry would describe remaining secondary assumptions. For the aggregate-factor exception, the replacement decision would read hold because the overlapping material contribution is not separable. The same template therefore records both an accepted improvement and a reasoned refusal to alter the total.
Choose review triggers that relate to applicability. A supplier changing production location can matter even if the document remains within a routine annual review window. Conversely, a new file date alone may not mean the underlying evidence changed. The ledger should make those distinctions understandable to someone taking over the purchase next quarter.
Keep the record together without treating storage as assurance
A useful evidence chain connects the purchased product revision, the direct supplier, the indirect-supplier response and the calculation that uses it. Each connection needs a meaning: supplied by, describes, replaces or remains unresolved. Keeping documents together can make that chain easier to follow, but storage alone cannot establish a valid boundary, certify a supplier factor or turn an estimate into measured evidence.
Start with the proposed replacement that matters most to an upcoming decision. Give it an owner and complete the ledger until the team can state one of three outcomes: use the new record within a matched boundary, request the missing explanation, or retain the existing estimate with its limitation. All three are more informative than treating every received response as an accepted input.
The lasting result is a traceable decision. A later product revision or supplier change can then reopen the affected assumption without erasing why the earlier model was used. That is how a focused Tier 2 request becomes a repeatable improvement in evidence quality, rather than another unattached number in a growing file collection.
SKU, BOM & supplier data
