Scope 3 Category 1: Turn Purchasing Records into a Clear Production Footprint
Original original synthetic procurement-workspace photograph. Bright generic analyst comparing physical components, fibre packaging and supplier records; illustrative scene only, not a real measured footprint. Spatial grouping, branches, proportions or directional paths carry the claim with concise labels.
Diagram text description
A supplier invoice can contain materials, a delivery charge and a new machine. Those purchases do not all belong in the same emissions category. Scope 3 Category 1 covers the upstream production of purchased goods and services; capital goods and certain transport activities have separate reporting routes.
The practical task is to turn the purchase book into a set of identifiable activities, then apply evidence that fits each one. A supplier's corporate total is rarely the answer to a specific purchased component. A product factor can be useful, but only when its unit and production boundary match what was bought.
This guide follows that work from classification to calculation. A fictional material purchase shows how a physical quantity becomes a Category 1 subtotal, why a delivered-product factor can overlap freight, and how a price change can move a spend estimate without proving a physical emissions increase. The examples demonstrate recognised calculation structures and explicit assumptions, not a completed company inventory.
What is included in Category 1?
The GHG Protocol defines Category 1 as upstream, cradle-to-gate emissions from production of products purchased or acquired by the reporting company during its reporting year. “Products” includes tangible goods and intangible services. A manufacturer might include bought components, raw materials and contracted design work; a service firm might include outsourced support and office consumables. The question is not whether the item will eventually be sold. It is whether the company acquired the good or service and whether its production emissions belong here rather than in another upstream category.
Materials and supplier processing sit inside the production boundary. The arrow then crosses into the direct supplier-to-buyer freight leg, assumed third-party transport here. Upstream transport already within material production must not be confused with that later leg.
Diagram text description
Cradle-to-gate describes the supplier-side chain from raw-material extraction through production of the purchased product, not every event in that product's lifetime. A purchased component may embody the emissions of upstream material production and manufacturing. The reporting company's later use of energy to process that component is its own Scope 1 or 2 activity, not more Category 1 emissions. Subsequent customer use or end-of-life treatment, where applicable, has separate downstream treatment.
The guidance allows companies to distinguish production-related purchases from non-production purchases. That split can align with procurement records and help teams find the largest evidence gaps. It does not create a licence to omit office supplies or bought services simply because they sit outside a bill of materials. Establish the reporting year, acquisition boundary and consistent purchase classes first; only then judge which class needs a more precise calculation.
Where does Category 1 end?
Category 1 is not a catch-all for every cost that procurement can see. The GHG Protocol assigns capital goods to Category 2. A production machine, long-lived facility equipment or other capital asset therefore needs a separate categorisation decision from the operating materials that flow through the business. Accounting treatment can help identify the candidate purchase, but the emissions-category decision still needs to follow the inventory methodology.
Transport is a second common trap. The Category 1 chapter expressly places transportation of purchased products from the tier-one supplier to the reporting company, in vehicles the company does not own or control, in Category 4, upstream transportation and distribution. The supplier's production chain may itself contain upstream transport before its gate; that is different from moving the finished purchase from the direct supplier to the reporting company. If a company owns or controls the delivery vehicle, its own Scope 1 boundary must also be considered.
Imagine a business buying aluminium housings and paying a carrier to deliver them from the direct supplier. The purchased housing's production footprint is a Category 1 candidate; the separately arranged inbound freight is a Category 4 candidate. Buying a new press used for years is a Category 2 candidate. A data model that records all three under “supplier spend” without activity and category flags risks double counting. Record the boundary decision beside each class, not only in a footnote after totals have been calculated.
A capital purchase is assessed in Category 2. A non-capital purchase still needs its activity classified: ordinary goods and services may belong in Category 1, while direct-supplier freight has a separate route. The branch is a first decision, not a complete category taxonomy.
Diagram text description
Map purchases before choosing a factor
Start with a procurement map rather than a factor database. Extract reporting-year purchases from the purchasing system and group them by meaningful goods and service classes. Keep the supplier, item description, quantity and unit, spend and currency, acquisition period, purchasing entity and any known destination. Reconcile the extract with finance records so a missing supplier account or duplicated credit note does not quietly become a carbon result.
Next, tag likely category exceptions: capital assets, energy purchases, freight, waste services, leased assets and employee travel may have their own accounting routes. A label such as “operations” is not enough. If a bundled contract includes equipment, maintenance and delivery, document whether the supplier can split the amounts or activities. If not, record an allocation rule and uncertainty rather than assigning the whole contract to the most convenient category.
Prioritise the classes where better evidence could change a sourcing or reduction decision. Purchase volume, estimated footprint, supplier concentration and factor uncertainty all matter. A frequently purchased material with a broad proxy may deserve attention before a small, well-documented office item. Keep both in the population; prioritisation determines the next data request, not which purchases disappear from the inventory.
Choose the calculation method that the evidence can support
The GHG Protocol offers four calculation families. Supplier-specific data uses product-level cradle-to-gate inventory information for the purchased good or service. It can capture the actual supplier's process, but only if the functional unit, allocation, time period, production boundary and data quality match the purchase. A glossy supplier corporate-total report does not meet that description on its own.
The hybrid method combines supplier-specific activity or allocated Scope 1 and 2 data with secondary information for missing upstream stages. It can improve a material input where a supplier has useful process records but not a complete product inventory. The team must show which stages are directly measured, which are modelled and whether any material or fuel flow is included twice. “Hybrid” is not a label for an undocumented spreadsheet mix.
Supplier-specific uses supplier product data; hybrid combines supplier and secondary information; physical-average and spend-based methods use different secondary-data denominators. Row lengths are schematic, not measured coverage or an accuracy ranking. The hybrid split does not prescribe a 50% data share.
Diagram text description
Average-data calculation multiplies physical units, such as mass or item count, by a relevant industry-average factor. Spend-based calculation multiplies monetary purchase value by an appropriate economic factor. Both rely on secondary data. A spend proxy can rapidly screen a large purchase book but may react to price changes unrelated to physical emissions; an average mass factor may miss supplier-specific process differences. The GHG Protocol says the listed methods progress in supplier specificity, not that the first is automatically most accurate. Choose the method that the evidence can actually defend, and record why.
What should a supplier data request contain?
A useful supplier request names the purchased product or service, its quantity and functional unit, the reporting period and the cradle-to-gate boundary the buyer needs. Ask whether the figure represents a product-level footprint, an allocated share of the supplier's corporate emissions or only one production stage. Request the method, factor sources, geography, energy mix, allocation basis, included gases, any third-party assurance and the data's age. For services, the functional unit and attribution may need more explanation than for a material sold by kilogram.
Do not ask for “your Scope 1, 2 and 3 number” and paste the reply into Category 1. A corporate total covers the supplier's whole organisation, possibly many products and customers. Without allocation and boundary work, adding that total to a purchase may substantially overstate or misattribute emissions. Conversely, a product figure that includes only final assembly may omit major upstream material stages. Record what is known and what remains a proxy.
Give the supplier a reason to respond with usable evidence. Identify the high-volume product family and the decisions the estimate informs. A small, consistent template can request activity and factor provenance while allowing an explicit “not available” answer. Preserve the original response, date and version. If a supplier later improves its method, the team should be able to explain the change between inventory years rather than silently replacing a historical number.
Remove overlap before adding production and freight
A delivered-product factor may include both production and the direct supplier's shipment to the buyer. If the buyer also calculates that shipment from carrier records, adding the full delivered factor creates an overlap. The solution is to identify the components, not to subtract an unrelated freight estimate from an unexplained total.
For a fictional example, suppose the supplier documents a delivered factor of 3.2 kg CO2e per kg of material: 3.0 for cradle-to-gate production and 0.2 for the specified inbound delivery. A 1,000 kg purchase therefore contains 3,000 kg CO2e of production and 200 kg of delivery, a combined 3,200 kg. Counting the full 3,200 and then adding the same 200 kg transport again would produce 3,400 kg.
The fictional delivered footprint contains 3,000 kg CO2e production and 200 kg delivery, totalling 3,200. Bars share one linear scale. Count the delivery once; adding the same 200 kg again would produce 3,400. The component breakdown must be documented before splitting a real factor.
Diagram text description
The corrected split is 3,000 kg in Category 1 and 200 kg in the applicable Category 4 line for the assumed third-party movement. This allocation relies on the supplier's explicit component breakdown and the same shipment boundary. It is not permission to derive a production factor by subtracting any carrier number that happens to be available. Where the components are unknown, obtain clarification or use a consistent alternative estimate.
Apply the same principle to capital goods. A new production press may already be classified separately in Category 2 while remaining in the raw procurement extract. Flag that purchase before aggregating Category 1, rather than trying to identify duplicates after a supplier-spend total has been calculated.
For each material class, trace the production endpoint, the direct supplier-to-buyer transport leg and any separately recorded activity. The factor documentation determines which stages are present. Short labels such as “cradle-to-gate” or “delivered” help locate the question, but the underlying coverage needs to resolve it.
Distinguish better data from better physical performance
An estimate can change because quantities changed, because the supplier changed its process, or because the estimation method changed. Those explanations lead to different decisions. Preserve the activity and factor components separately so that an apparent improvement can be attributed correctly.
A spend-based teaching case makes the distinction visible. Assume purchases of 10,000 currency units and an illustrative factor of 0.30 kg CO2e per matching currency unit. The estimate is 3,000 kg CO2e. If the same physical order costs 12,000 next year and the factor is held fixed, the proxy rises to 3,600 kg. That 20% increase follows the monetary input; it does not demonstrate a 20% increase in the supplier's physical emissions.
At an unchanged illustrative factor of 0.30 kg CO2e per currency unit, spend rising from 10,000 to 12,000 raises the estimate from 3,000 to 3,600 kg. This is a price sensitivity, not proof of higher physical emissions. Real currency and price-year rules still apply.
Diagram text description
This is a sensitivity example, not a forecast or an instruction to ignore price-basis adjustments. A real spend factor has a currency, price year and economic coverage. Use its documented treatment of inflation and currency conversion, and avoid applying a nominal invoice amount to a factor with an incompatible monetary basis.
Physical averages and supplier-specific data have their own change questions. A mass factor can miss a supplier's process improvement; a newly supplied product result can omit upstream stages that an older average included. Compare boundaries before presenting the difference as a reduction. More specific evidence is useful when it fits the purchase and improves the estimate, not merely because its label sounds more advanced.
Keep a short change note for material revisions: the affected purchase population, old and new methods, the reason for the change and its treatment in comparative reporting. That record lets procurement distinguish a genuine sourcing opportunity from a recalculation driven by better information.
A procurement-team evidence checklist
A compact evidence file can travel with the calculation and survive staff changes. It should state the reporting year and organisational boundary; list purchase classes, source quantities and reconciliation checks; show each class's Scope 3 category assignment; and identify the calculation method, factor, supplier document and reviewer. It should also capture exclusions, unresolved allocations and planned improvements. The file is a decision record, not merely a data dump.
Table 1: scroll horizontally if needed.
| Checkpoint | Question for the reviewer | Evidence to retain |
|---|---|---|
| Category | Is this an operating purchase, capital good or separate transport service? | Purchase description, capitalisation or freight rule, category map |
| Activity | What quantity or spend was acquired in this reporting year? | Purchase extract, units, period and reconciliation |
| Factor | Does the source match the product, geography, unit and stage boundary? | Factor record, method, version and assumptions |
| Supplier | Is the figure product-specific and allocated, or only a corporate total? | Original supplier response, allocation and assurance notes |
| Review | What remains estimated, excluded or at risk of overlap? | Exception log, owner, decision date and next action |
The reviewer should be able to trace a reported figure from class total back to original procurement and factor evidence. If an inbound carrier charge and a purchased-product factor both include the same route, the overlap is visible before approval. If a supplier document arrives later, the owner can decide whether it improves the relevant period and preserve the previous version.
Work through one physical purchase
Suppose a manufacturer acquires 1,000 kg of a material during the reporting year. For this fictional case, assume a relevant average-data factor of 3 kg CO2e per kg material covering its cradle-to-gate production. Multiplying the purchased quantity by that factor gives 3,000 kg CO2e, or 3 tonnes, for the Category 1 production estimate.
Table 2: scroll horizontally if needed.
| Calculation element | Illustrative input or output | Meaning |
|---|---|---|
| Purchased material | 1,000 kg | Physical quantity in the reporting population |
| Production factor | 3 kg CO2e/kg material | Assumed matching cradle-to-gate factor |
| Category 1 estimate | 3,000 kg CO2e | Quantity multiplied by production factor |
| Separate inbound transport | 200 kg CO2e | Only the separately defined delivery in the overlap example |
| Combined production and delivery | 3,200 kg CO2e | Two non-overlapping components, not a Category 1-only total |
The invented 1,000 kg material purchase and 3 kg CO2e/kg production factor yield 3,000 kg CO2e. This is the average-data method structure under an assumed boundary match. The separately discussed inbound freight is outside this subtotal.
Diagram text description
The physical-factor multiplication is a recognised average-data method structure. The factor value and purchase are invented, and their assumed match does not establish a real dataset's suitability. In an actual record, check material grade, production geography, period and included processes before using the factor. A superficially similar material name is not enough.
The separate transport line also has a limited job. It demonstrates the non-overlapping decomposition described earlier; it is not derived from a new freight calculation here. The table must therefore not be read as a complete method for estimating a carrier's emissions. Obtain that line from its own compatible transport method and activity evidence.
As a simple sensitivity check, changing only the production factor from 3.0 to 3.3 raises the production estimate from 3,000 to 3,300 kg CO2e. The 300 kg difference says how responsive this purchase is to that assumed factor change. It does not establish the factor's actual uncertainty. Use source-quality evidence to decide which scenarios deserve investigation.
Choose the next supplier request from the calculation
Start improvement work with a purchase class large enough that its uncertainty could change a decision. In the fictional case, the useful questions are whether the production factor represents the material and whether the delivered factor separates inbound transport. A broad request for the supplier's company footprint would answer neither question directly.
Send the supplier the product identifier, quantity unit, reporting period and required production boundary. Ask for the corresponding product-level result or the information needed to explain the relevant gaps. Where that evidence is unavailable, retain the stated proxy and make the next improvement concrete: a better material match, a resolved allocation or an explicit transport breakdown.
Category 1 becomes easier to manage when every subtotal has a purchase population and a method that fit one another. Preserve that relationship as suppliers and datasets change. It gives procurement a reason for each data request and keeps the inventory useful for decisions beyond producing a single annual total.
Company carbon accounting
