Most food footprints are still built from spend. Spend-based numbers are fine for screening a portfolio and useless in a procurement negotiation, because the supplier cannot defend them.
A spend-based model multiplies what you paid by an industry average. It is cheap, fast and frequently wrong by a factor of two — because the same euro of ingredient can come from a hydro-powered processor or a coal-fired one. When a buyer puts that number into a scope 3 target, the supplier is being asked to defend somebody else's average.
Primary data replaces the average with a measurement. That requires the processor to know their own line items, which is exactly what a verification engagement produces.
| Line item | Typical share | Evidence |
|---|---|---|
| Ingredient production | 45–65% | Supplier footprint statements, traceable to a batch |
| Processing energy | 15–25% | Sub-metered lines, hourly where available |
| Cold chain | 5–15% | Temperature logs plus electrical meter on the cold store |
| Packaging | 6–14% | Material mass × published factor |
| Inbound transport | 3–10% | Distance × mode, not a global default |
One line makes three products and one clean-ingredient declaration. Splitting the electricity is a choice: by mass, by economic value, or by machine hours. Each produces a different number and all three are defensible if declared.
We require the allocation rule to be fixed before verification starts and published in the statement. A verifier who lets a client pick the rule after seeing which one flatters the result has verified nothing.
| Grade | What you must supply | What it buys you |
|---|---|---|
| A | Sub-metered lines, calibrated, 12 months | Buyer-facing claim, defensible in audit |
| B | Site-level meters plus line-level estimates | Buyer-facing claim with a stated uncertainty range |
| C | Utility invoices only | Internal screening, not a supplier claim |
| Stage | Days | From you | You get |
|---|---|---|---|
| Scope | 3 | Product families, sites, allocation question | Boundary decision |
| Document review | 4 | Meter data, mass balance, factor sources | Findings log |
| Verification | 3 | Line access, sampling windows | Test results |
| Technical review | 2 | Nothing | Statements per family |
A roaster imports green coffee from three origins and blends into four SKUs. Traceability stops at the blend, so the honest move is to verify the green coffee at origin and carry the blend as a mass-weighted combination — with the weights published.
| Step | What happens | Why |
|---|---|---|
| 1 | Verify each origin lot | The origin is where metering exists |
| 2 | Weight by mass in the blend | Economic weighting would let marketing move the number |
| 3 | Verify roasting energy separately | It is the one line the roaster controls |
| 4 | Publish the blend recipe share | So a buyer can recompute the blended footprint |
No. Cover the ones that carry the mass. In most food businesses 20% of suppliers carry 80% of the footprint — start there and label the rest as estimated.
Yes, once it is issued — the buyer cites the statement reference and the label id rather than copying numbers into a spreadsheet.
About twelve weeks from kick-off, including the time it takes the supplier to produce meter records they did not previously keep.