How to measure a supply chain carbon footprint and improve the evidence
Four people examine a large printed map. Papers, material samples and a laptop sit on the table, and warehouse shelving is visible beyond the windows.
Diagram text description
A supply chain carbon footprint estimates greenhouse gas emissions linked to the goods, services and other value-chain activities a company relies on. Depending on the purpose and boundary, the work can also cover what happens after a product is sold, including its use and disposal. The result is usually expressed in carbon dioxide equivalent, or CO₂e, so gases can be reported on a common basis.
Start with the company and reporting period, then map the activities that belong in the assessment. An activity can only be assigned to the right source when the boundary is clear. Use a screening estimate to identify likely large sources and record where the calculation rests on assumptions. The confidence in each estimate helps you decide whether to request better evidence, test a change with a supplier or move to the next source.
Keep the calculation and the evidence register together. The total tells you where to look. The register tells you whether the number can support the decision you are making. The worked example below shows how to improve one purchased-material estimate while keeping the part that still relies on a proxy visible.
What a supply chain carbon footprint includes
A corporate inventory starts with the reporting entity. Scope 1 covers direct emissions from sources the company owns or controls under its chosen boundary. Scope 2 covers emissions from generation of purchased energy the company consumes, such as electricity or heat. Scope 3 covers other indirect emissions in the company’s value chain. In everyday use, “supply chain footprint” sometimes means only the upstream part. Specify whether your assessment also covers downstream activities.
| Record | Object and coverage |
|---|---|
| Upstream corporate view | Purchases and other upstream activities associated with the reporting entity and period |
| Full corporate value-chain view | Applicable upstream and downstream scope 3 categories, including use and end of life where relevant |
| Product life cycle footprint | Greenhouse gas emissions for a specified product, reference unit and defined life cycle boundary |
Table 1, shown one row at a time. Every field is retained.
Row 1
- Record
- Upstream corporate view
- Object and coverage
- Purchases and other upstream activities associated with the reporting entity and period
Row 2
- Record
- Full corporate value-chain view
- Object and coverage
- Applicable upstream and downstream scope 3 categories, including use and end of life where relevant
Row 3
- Record
- Product life cycle footprint
- Object and coverage
- Greenhouse gas emissions for a specified product, reference unit and defined life cycle boundary
For a manufacturer, production of sheet material at a supplier is typically part of the purchased goods category. Fuel burned in the manufacturer’s own controlled equipment generally belongs in scope 1. The fuel and electricity a customer uses to operate sold equipment can belong in downstream scope 3. The party operating the source and the company’s relationship to that activity affect the classification.
A product footprint may contribute an input to the company inventory when its unit and boundary fit. To reuse it, identify which part relates to the purchases in the reporting period and where any stages beyond the intended purchase boundary are already covered.
A reporting-entity inventory assigns purchases, transport and relevant downstream sources to applicable categories and the reporting period. A product result takes only attributable activity for the defined unit and selected stages. Screen material, supplier production, freight, use and end of life against the chosen scope.
Diagram text description
Map emission sources before requesting numbers
Map activities before emailing suppliers. Start with an entity list, the reporting period and a copy of the purchase, freight and sales records that cover it. Agree with finance which business units and locations the records represent. Procurement can then group lines by material, component or service and flag bundles that need to be split.
| Candidate source | Record that starts the map | Initial owner |
|---|---|---|
| Purchased materials, components and services | Purchase orders, goods receipts, bills of materials and service descriptions | Procurement |
| Freight and distribution | Freight invoices, routes, shipment mass, transport mode and who paid | Logistics |
| Third-party processing or storage | Contracts, quantities processed, warehouse activities and delivery terms | Operations |
| Use of sold products | Sales quantities, product specifications and expected use by market | Product team |
| End of life | Material composition, market distribution and plausible treatment routes | Product or waste specialist |
Table 2, shown one row at a time. Every field is retained.
Row 1
- Candidate source
- Purchased materials, components and services
- Record that starts the map
- Purchase orders, goods receipts, bills of materials and service descriptions
- Initial owner
- Procurement
Row 2
- Candidate source
- Freight and distribution
- Record that starts the map
- Freight invoices, routes, shipment mass, transport mode and who paid
- Initial owner
- Logistics
Row 3
- Candidate source
- Third-party processing or storage
- Record that starts the map
- Contracts, quantities processed, warehouse activities and delivery terms
- Initial owner
- Operations
Row 4
- Candidate source
- Use of sold products
- Record that starts the map
- Sales quantities, product specifications and expected use by market
- Initial owner
- Product team
Row 5
- Candidate source
- End of life
- Record that starts the map
- Material composition, market distribution and plausible treatment routes
- Initial owner
- Product or waste specialist
Mark each category as relevant, still to be assessed or excluded with a documented reason. An outsourced coating operation may be recorded as a purchased service, while a separate freight invoice covers the trip to the coater. Follow the actual activities across invoices and contracts. The accounting label on an invoice alone may hide the activities bundled into it.
Use purchasing data to check whether all buying teams are represented. Use the logistics record to catch shipments booked outside the main purchasing system. For downstream sources, agree which sales and product lines are represented and flag any missing markets or use profiles. Give every unresolved activity a named owner before starting the calculation. A blank source map cell should trigger a question about coverage.
Start with a screening estimate, then improve what matters
A screen is a first estimate for finding sources worth more effort. Choose the best available activity information for each source. Where all you have is expenditure, multiply it by a relevant spend factor expressed in emissions per currency unit and record the factor’s sector, geography, price year and currency. Use it to rank sources with comparable limitations. High prices or changing exchange rates can move a spend estimate even when the activity stays the same.
Rank each source twice, first by expected footprint size and then by uncertainty that could change the decision. A large estimate with a poor proxy gets early attention. A large source with already reliable data may be ready for an operational trial. A small source with high uncertainty can stay in the register until an improved estimate would change its priority. Where the ranking is close, test a reasonable alternative proxy to see whether the order changes.
Rank each source by expected footprint and by uncertainty that could affect the decision. Improve a material source with a poor proxy first. A large source backed by strong evidence may already support action. A small source can wait unless its uncertainty matters to the decision.
Diagram text description
Improve the most material gaps with a better physical quantity, a factor for the relevant geography or technology, or supplier-specific evidence. Keep a version of the original calculation so the team can see why the priority moved. If supplier information is unavailable, retain the best documented proxy for that coverage and give the next evidence request an owner and date.
Choosing activity, supplier and secondary data
Activity data describe what happened, such as kilograms of resin bought or tonne-kilometres travelled. A factor converts the activity to emissions for a defined process and boundary. An average factor or spend factor comes from secondary data. A supplier-specific factor may combine measurements for the supplier’s own operations with secondary datasets for upstream inputs. Ask which parts are specific before relying on it.
| Method | What it needs | When to question the result |
|---|---|---|
| Spend screening | Category spend and a factor matched to the sector, currency and price year | A premium price, mixed purchases or currency changes may dominate |
| Physical activity with an average factor | Quantity or distance and a matching factor for the material, mode or process | The factor may represent another geography, technology or boundary |
| Supplier-specific product factor | Purchased quantity and a factor for the actual product, plant, period and boundary | A plant average, allocation assumption or missing upstream input may be material |
| Hybrid estimate | Supplier evidence for defined processes plus suitable secondary factors for named gaps | Overlapping stages or omitted material inputs may distort the estimate |
Table 3, shown one row at a time. Every field is retained.
Row 1
- Method
- Spend screening
- What it needs
- Category spend and a factor matched to the sector, currency and price year
- When to question the result
- A premium price, mixed purchases or currency changes may dominate
Row 2
- Method
- Physical activity with an average factor
- What it needs
- Quantity or distance and a matching factor for the material, mode or process
- When to question the result
- The factor may represent another geography, technology or boundary
Row 3
- Method
- Supplier-specific product factor
- What it needs
- Purchased quantity and a factor for the actual product, plant, period and boundary
- When to question the result
- A plant average, allocation assumption or missing upstream input may be material
Row 4
- Method
- Hybrid estimate
- What it needs
- Supplier evidence for defined processes plus suitable secondary factors for named gaps
- When to question the result
- Overlapping stages or omitted material inputs may distort the estimate
Check that the denominator matches the purchasing record. A factor per kilogram of dry material cannot be applied directly to a delivered mass that includes moisture and packaging. Reconcile the masses first and retain the conversion. Ask the supplier how shared processes were allocated among products, and whether the reported plant actually supplied the purchase.
Prefer the evidence that best represents the activity for this decision and can be reviewed. A relevant published average can be more defensible than a supplier’s undocumented company-wide intensity. Revisit the choice when the supplier can demonstrate better coverage.
Engage suppliers with a request they can answer
Ask for data in the form a supplier can provide and a reviewer can use. Name the product, site and purchase period in the request. Explain the buyer’s intended use and who can see the response. Agree whether the supplier can send a product factor, supporting activity data or a statement of what remains unavailable.
| Request field | What the supplier can send |
|---|---|
| Product and production site | Item code, grade or revision and the site or sites that supplied it |
| Period and amount | Period represented and quantity or activity covered |
| Boundary and exclusions | Processes covered, treatment of upstream inputs and transport, and explicit exclusions |
| Value and unit | Factor with numerator and denominator, or activity quantities with units |
| Method and support | Calculation basis, allocation where used, source references and review or assurance scope |
Table 4, shown one row at a time. Every field is retained.
Row 1
- Request field
- Product and production site
- What the supplier can send
- Item code, grade or revision and the site or sites that supplied it
Row 2
- Request field
- Period and amount
- What the supplier can send
- Period represented and quantity or activity covered
Row 3
- Request field
- Boundary and exclusions
- What the supplier can send
- Processes covered, treatment of upstream inputs and transport, and explicit exclusions
Row 4
- Request field
- Value and unit
- What the supplier can send
- Factor with numerator and denominator, or activity quantities with units
Row 5
- Request field
- Method and support
- What the supplier can send
- Calculation basis, allocation where used, source references and review or assurance scope
Send a short example for each field and allow suppliers to mark unknowns. If a smaller supplier has utility bills and production totals but no allocated product footprint, ask for the relevant quantities and what they cover. Agree how shared processes could be allocated before using the data for a specific product. A named contact can help interpret the request or accept a more suitable record.
Give the supplier one place to send corrections. Keep versions and log which value the buyer actually used. A useful first response can narrow a missing-data problem even if a factor cannot yet be approved.
Build a data-quality register beside the total
Create one register entry per estimate or clearly separated portion of an estimate. Use a stable reference to link it to the calculation, source file and approval record. Copy these fields for each new entry. The sample column illustrates the level of detail a reviewer should be able to find.
| Field to enter | Example entry |
|---|---|
| Source and version | Receiving extract M-42, revision 2, and factor sheet A-17 |
| Reporting and data period | Reporting year 2025. Factor represents production in 2024 |
| Boundary and coverage | Purchased sheet at the supplier gate. Delivery to buyer recorded separately |
| Activity and factor units | Kilograms of sheet and kilograms CO₂e per kilogram of sheet |
| Estimate method | Physical purchase mass multiplied by the regional average factor |
| Quality and uncertainty | The factor’s production route is known. Actual supplier route is unresolved |
| Validation and owner | Buyer reconciles mass to receipts. Category analyst reviews the factor and boundary |
| Improvement action and due date | Request plant and route evidence before the next quarterly refresh |
Table 5, shown one row at a time. Every field is retained.
Row 1
- Field to enter
- Source and version
- Example entry
- Receiving extract M-42, revision 2, and factor sheet A-17
Row 2
- Field to enter
- Reporting and data period
- Example entry
- Reporting year 2025. Factor represents production in 2024
Row 3
- Field to enter
- Boundary and coverage
- Example entry
- Purchased sheet at the supplier gate. Delivery to buyer recorded separately
Row 4
- Field to enter
- Activity and factor units
- Example entry
- Kilograms of sheet and kilograms CO₂e per kilogram of sheet
Row 5
- Field to enter
- Estimate method
- Example entry
- Physical purchase mass multiplied by the regional average factor
Row 6
- Field to enter
- Quality and uncertainty
- Example entry
- The factor’s production route is known. Actual supplier route is unresolved
Row 7
- Field to enter
- Validation and owner
- Example entry
- Buyer reconciles mass to receipts. Category analyst reviews the factor and boundary
Row 8
- Field to enter
- Improvement action and due date
- Example entry
- Request plant and route evidence before the next quarterly refresh
Use separate fields for measured, modelled and missing inputs if an estimate mixes them. Record coverage, time, geography and specificity where each affects confidence. Add a status such as accepted for screening, accepted for reporting within scope, or hold. Attach the review decision and the evidence that could change it.
Request the item, site, period, boundary, units and evidence. Choose physical activity with a matched factor, a reviewed supplier result, or a documented representative estimate. Record coverage, period, geography, specificity, validation and gaps beside the value.
Diagram text description
When you replace a value, create a new version with the reason for the change. Keep the earlier one available for comparisons and for anyone who needs to reconstruct a reported total.
Prevent gaps and double counting
Reconcile the source map against the organisation and reporting period before adding category totals. Start with the population of records, then verify that each included activity has a single place in the calculation. This is a check on the reporting company’s inventory. Different companies can legitimately report emissions associated with the same activity in their respective scopes.
| Reconciliation check | Action |
|---|---|
| Supplier boundary and inbound freight | Identify where the factor ends and where the freight leg starts. Separate any duplicated leg before adding the freight estimate |
| Intercompany movements | Check whether both entities fall inside the reporting boundary and account for the underlying source in its appropriate scope |
| Purchases and third-party processing | Match service and purchase references so a bundled operation is included once |
| Mass, distance and currency | Normalise units and document conversions before multiplying or summing |
| Missing records | Trace gaps to the entity, month or activity affected and disclose the estimate or exclusion |
Table 6, shown one row at a time. Every field is retained.
Row 1
- Reconciliation check
- Supplier boundary and inbound freight
- Action
- Identify where the factor ends and where the freight leg starts. Separate any duplicated leg before adding the freight estimate
Row 2
- Reconciliation check
- Intercompany movements
- Action
- Check whether both entities fall inside the reporting boundary and account for the underlying source in its appropriate scope
Row 3
- Reconciliation check
- Purchases and third-party processing
- Action
- Match service and purchase references so a bundled operation is included once
Row 4
- Reconciliation check
- Mass, distance and currency
- Action
- Normalise units and document conversions before multiplying or summing
Row 5
- Reconciliation check
- Missing records
- Action
- Trace gaps to the entity, month or activity affected and disclose the estimate or exclusion
For example, a factor supplied at the buyer’s gate may include delivery that is also recorded in the carrier data. Ask the supplier to separate the delivery leg or provide an upstream factor ending at the supplier gate. If the boundary remains uncertain, document the overlap and hold the affected calculation for review or use a transparent estimate with a clear boundary.
Compare the aggregate purchase and freight activity against the totals in the underlying systems. Investigate missing routes, credits and returns, and record the treatment of items spanning two periods. Before reporting, assign one person to sign off on the coverage and another, where feasible, to trace a sample back from the calculated result to the original records.
Turn the inventory into a reduction plan
Use the inventory to choose changes within the team’s influence. A source can be large but hard to alter in the next purchasing cycle. Give priority to actions with a material expected effect, an owner who can change the activity and a way to measure the result. Estimate feasibility and evidence quality separately.
| Material source | Feasible action | Measure of progress |
|---|---|---|
| High-emission purchased material | Test a qualified lower-emission specification with the supplier | Compare equivalent output and track the quantity bought at the new specification |
| Frequent part-load shipments | Consolidate eligible orders or change the delivery schedule | Track load utilisation, transport activity and emissions for the affected routes |
| Supplier processing energy | Agree an efficiency project and define the process to measure | Compare energy per equivalent output using consistent boundaries and periods |
Table 7, shown one row at a time. Every field is retained.
Row 1
- Material source
- High-emission purchased material
- Feasible action
- Test a qualified lower-emission specification with the supplier
- Measure of progress
- Compare equivalent output and track the quantity bought at the new specification
Row 2
- Material source
- Frequent part-load shipments
- Feasible action
- Consolidate eligible orders or change the delivery schedule
- Measure of progress
- Track load utilisation, transport activity and emissions for the affected routes
Row 3
- Material source
- Supplier processing energy
- Feasible action
- Agree an efficiency project and define the process to measure
- Measure of progress
- Compare energy per equivalent output using consistent boundaries and periods
Set a baseline for each action with the source, period and method recorded. Agree who will track the operational change and who will recalculate its effect. Where a material or process changes, check that the relevant performance requirements still hold and account for other stages affected by the decision.
Set the category boundary and map sources. Screen each source, retain gaps and improve evidence for the material ones. Agree a supplier or specification change with its owner. Compare the later physical activity and emissions on an aligned basis before describing a reduction.
Diagram text description
When a total falls, separate operational changes from better data, changes in volume and changes in the calculation method. If you cannot compare like with like, show the estimated change with its limitations and agree what additional records are needed to measure progress.
A worked supplier-data decision
Consider this fictional example. Fennel Works bought 120,000 kg of a specified steel sheet during 2025. The purchase calculation covers production to the supplier gate. Freight from the supplier gate to Fennel’s receiving site sits in a separate transport record. The volumes, factors and supporting files are invented for this example.
The initial screen uses one average factor of 1.80 kg CO₂e per kg for the full purchase. After screening, one supplier provides a plant and grade-specific factor with a matching production period and an explanation of how shared plant energy was allocated. The second supplier provides a company-wide factor without the upstream steel production boundary. The diagram shows the decision applied to each portion.
Compare the product, site, period, units and boundary. In the fictional case A’s evidence fits 80,000 kg of steel and B’s upstream boundary is missing for 40,000 kg. Keep the screening factor for B and request the missing evidence. A lower revised estimate reflects the factor change.
Diagram text description
| Estimate or purchase portion | Activity and factor | Calculated emissions |
|---|---|---|
| Initial screen | 120,000 kg × 1.80 kg CO₂e per kg | 216,000 kg CO₂e |
| Supplier A, evidence accepted for this purchase | 80,000 kg × 1.50 kg CO₂e per kg | 120,000 kg CO₂e |
| Supplier B, screening factor retained | 40,000 kg × 1.80 kg CO₂e per kg | 72,000 kg CO₂e |
| Revised category estimate | 80,000 kg plus 40,000 kg | 192,000 kg CO₂e |
Table 8, shown one row at a time. Every field is retained.
Row 1
- Estimate or purchase portion
- Initial screen
- Activity and factor
- 120,000 kg × 1.80 kg CO₂e per kg
- Calculated emissions
- 216,000 kg CO₂e
Row 2
- Estimate or purchase portion
- Supplier A, evidence accepted for this purchase
- Activity and factor
- 80,000 kg × 1.50 kg CO₂e per kg
- Calculated emissions
- 120,000 kg CO₂e
Row 3
- Estimate or purchase portion
- Supplier B, screening factor retained
- Activity and factor
- 40,000 kg × 1.80 kg CO₂e per kg
- Calculated emissions
- 72,000 kg CO₂e
Row 4
- Estimate or purchase portion
- Revised category estimate
- Activity and factor
- 80,000 kg plus 40,000 kg
- Calculated emissions
- 192,000 kg CO₂e
The buyer accepts A’s factor for the 80,000 kg after checking the site, grade, allocation and purchased quantity. For B, procurement records the missing upstream boundary and requests a production route and a factor covering the purchased grade to the supplier gate. The register shows 40,000 kg still modelled with a secondary factor. Both lines retain their original purchase references.
The revised total is 24,000 kg CO₂e lower than the screen. The activity quantities are identical. The change comes from the factor replacement, so Fennel records it as an estimate revision. To claim an operational reduction later, it would need evidence of a change in the activity or its emissions on a comparable basis.
What the inventory can support
The corporate inventory supports decisions about the sources covered by the reporting entity, method and period. A purchased-material entry can inform procurement or feed a product study. A product result needs the specified product, reference unit, life cycle boundary and data for the included stages. The emissions for Fennel’s steel purchase cannot describe the footprint of a finished machine without the other relevant inputs and processes.
Data traceability helps another person reconstruct the estimate. Assurance depends on the scope and work of the party that carried it out. If a supplier provides a verification statement, read which figures, periods and sites are covered. A check on a supplier’s own inventory may leave the allocation to your purchased grade outside that work.
Match any public statement to the evidence that supports it. A statement about revised estimation coverage needs a record of the data changed and its use. A reduction statement needs a comparable baseline and evidence of an operational change. A claim about the footprint of a product needs evidence for that product’s boundary. Keep the supporting evidence and approval attached to the actual words so the next reviewer can identify what would trigger a revision.
Questions to ask before reporting or buying tools
Agree the method and the records you need before choosing software. Use one category with real data to test whether the proposed process can follow a value from receipt through calculation to review. Give a buyer, an analyst and a reviewer the same sample so each can check what they need to do.
| Question to ask | Evidence to look for |
|---|---|
| Can we identify which entities, activities and dates are covered? | Boundary and category fields and a way to record exclusions |
| Can we reconstruct an estimate? | Activity data, units, factor source and version linked to each calculation |
| Can we see uncertainty and unresolved gaps? | Coverage and status fields, an owner, and an improvement action |
| Can we compare two periods? | Prior versions and a way to separate method changes from activity changes |
| Can we control how supplier data are shared? | Access rules, correction history and an export suitable for review |
Table 9, shown one row at a time. Every field is retained.
Row 1
- Question to ask
- Can we identify which entities, activities and dates are covered?
- Evidence to look for
- Boundary and category fields and a way to record exclusions
Row 2
- Question to ask
- Can we reconstruct an estimate?
- Evidence to look for
- Activity data, units, factor source and version linked to each calculation
Row 3
- Question to ask
- Can we see uncertainty and unresolved gaps?
- Evidence to look for
- Coverage and status fields, an owner, and an improvement action
Row 4
- Question to ask
- Can we compare two periods?
- Evidence to look for
- Prior versions and a way to separate method changes from activity changes
Row 5
- Question to ask
- Can we control how supplier data are shared?
- Evidence to look for
- Access rules, correction history and an export suitable for review
The owner sets the reporting method, category boundaries and required inputs. The reviewer defines acceptance and exception handling. Test one real category with evidence tracing, revisions and an unresolved supplier input before relying on automation.
Diagram text description
Test an incomplete supplier submission. Check whether the buyer can retain an estimate with its rationale and preserve the missing item for follow-up. Ask the reviewer to trace one result back to the source and verify the units without relying on a dashboard label. Record any manual checks that remain and assign them to an owner before scaling up.
Start with one category and a review date
Choose one material category and a reporting period. Ask procurement to pull the purchase records, agree where the category ends and prepare a screening estimate. Fill the register as you calculate, including where an average factor stands in for a supplier record. Use the screen and its uncertainty to select the first supplier request.
Set a review date and choose the decision the improved evidence should inform. For a purchased material that might be whether a factor covers the actual grade and site, or whether there is enough evidence to trial a different specification. Bring the supplier response, the original estimate and the register to that review. If the evidence fits, replace the relevant portion and record the basis. If a material gap remains, keep the proxy visible and agree who will obtain the next record.
Company carbon accounting
