Request agricultural supplier data that matches the crop you buy

Fictional I036 crop-purchasing example. The buyer has 250 matched tonnes out of 1,000 purchased tonnes (25%), under this case's defined evidence request. Separately, activity packs cover 300 sampled tonnes out of the aggregator's 3,000-tonne intake population (10%). The records do not establish whether or which sampled intake supplied the buyer's deliveries. These percentages have distinct denominators; neither a shared scale of absolute tonnes nor a sample-to-delivery match is shown. No emissions, certification or environmental-benefit result is implied.

Before counting a supplier’s farm records towards your purchased-crop evidence coverage, ask which delivered lots those records support. In the case below, the processor can establish a matched activity pack for 25% of its purchased maize. A separate 10% sample of an aggregator’s intake cannot be added to that result or applied proportionally to the processor’s deliveries.

Set the purchase population and the evidence test

This fictional example follows a processor buying 1,000 tonnes of maize. It is a purchasing-evidence exercise for procurement and reporting teams; no real supplier records have been verified.

The processor’s purchase population is exactly 1,000 metric tonnes of maize received from 1 April 2025 inclusive to 1 April 2026 exclusive. All listed deliveries arrive in January–March 2026. The three supplier groups are disjoint and exhaustive: A supplies 600 tonnes, B supplies 250 tonnes and C supplies 150 tonnes.

All purchase, pool and sampled quantities use the same stipulated supported net-mass and moisture basis. Duplicate receipts, returns, rejected deliveries and quantity adjustments have already been resolved. These assumptions isolate the evidence-link question; this case performs no drying, stock or physical-loss reconciliation.

For this exercise, “matched” means that a purchased quantity has the stated crop, farm/lot and harvest/crop-cycle links, plus the agreed requested activity pack. Pack completeness is stipulated where stated below. This is Nitium’s fictional evidence-completeness convention, not an official traceability definition, certification or assurance conclusion. It does not establish suitability for every inventory method.

Keep three date fields separate

H25 is an invented crop-cycle label with a documented harvest window of 1 September–30 November 2025. It describes no regional crop calendar. The farm-activity pack must identify the actual crop-cycle start and end separately. Those activity dates, the harvest window and the January–March 2026 receipt dates answer different questions. A 2026 receipt does not by itself make a calendar-year 2026 farm report the right evidence.

Read the three supplier replies against the purchase lots

A: 600 purchased tonnes. The deliveries are documented as coming from maize intake pool A-H25, which totals 3,000 tonnes. Valid activity packs cover 300 unique intake tonnes from that pool and crop cycle. The supplied records do not identify whether, or which, sampled intake supplied this processor’s deliveries. Keep the pool reference and packs, but leave matched purchased tonnage unestablished. The true sample-to-delivery overlap remains unknown.

B: 250 purchased tonnes. Two distinct lots, B-01 at 150 tonnes and B-02 at 100 tonnes, have records linking them to named farm, crop and harvest references and complete requested activity packs for their corresponding crop-cycle intervals. These links and the review outcome are assumed in the scenario. Both lots meet the exercise’s evidence test: 150 + 100 = 250 tonnes.

C: 150 purchased tonnes. Supported country-of-origin information is available; the crop-harvest link and matched farm-activity pack are unavailable. Retain the origin information under that label. It does not supply the missing crop-cycle and lot links.

Calculate two percentages with two named populations

Purchased-mass evidence coverage = 250 ÷ 1,000 × 100 = 25%. Only B’s 250 tonnes have an established matched pack. The other 750 tonnes do not meet this defined evidence test on the supplied records. They still include A’s pool/sample information and C’s origin evidence; they are not evidence-free, wasted or zero-emission tonnes.

A-H25 intake-volume sample coverage = 300 ÷ 3,000 × 100 = 10%. This describes the upstream pool’s intake mass with activity packs. It is neither the share of farms sampled nor the share of the processor’s purchased lots with matched evidence.

Reject the two shortcuts

Adding unlike evidence populations gives (300 + 250) ÷ 1,000 × 100 = 55%. The arithmetic is correct, but 300 is sampled intake mass, not identified purchased mass. Calling 55% “matched purchase coverage” would assert a delivery link that the records do not establish.

Applying the sample share to A’s deliveries gives 600 × 300 ÷ 3,000 = 60 tonnes. This is a proportional inference, not an observed match. Adding it to B would give (250 + 60) ÷ 1,000 × 100 = 31%, which is also unsupported as directly matched purchased-mass coverage.

The case assumes no random sampling, homogeneous mixing, proportional distribution or representativeness. A separately justified modelling approach would need its own method and labels. It cannot turn the inferred 60 tonnes into identified deliveries. Recording no established matched quantity for A is also not a finding that its true overlap is zero.

Ask for the missing connection in a usable record

Use one row per buyer receipt or purchase lot, with linked upstream records where available. For this exercise, the request needs these fields:

• Purchase: supplier ID; buyer receipt ID; receipt date; buyer lot ID; maize quantity in tonnes; and the documented net-mass/moisture basis.

• Crop: crop name; crop-cycle ID; actual activity-period start and end; harvest start and end; and farm/crop reference. Keep unknown dates explicit.

• Link: upstream pool ID; intake-lot references; activity-pack ID; and the record reference connecting evidenced intake to the buyer lot, including the quantity that connection supports.

• Review: evidence state; missing-link note; and the record version used. Keep a supplier reply’s administrative status separate from tonnes meeting the evidence test.

For A, ask: Can you link our delivered lots to the evidenced intake, and identify the supported quantity and record references? If that link is unavailable, retain A-H25, the crop-cycle period, the 3,000-tonne pool and the 300-tonne sample as separate fields. Request the selection and weighting information for a separately scoped methodological review. Do not ask A to manufacture a 60-tonne match.

For B, retain B-01 and B-02 with their farm/crop/harvest references, activity-pack IDs and actual activity periods. If B sends a newer document, ask which of those lots and crop cycles it covers before replacing the linked version.

For C, request the supported crop/harvest references and upstream link to the purchased lots, or an explicit unavailable response. Preserve the country-only record while the missing link remains open.

Hand over the evidence states, then the next request

Send the reporting reviewer the 1,000-tonne purchase schedule, B’s 250 matched tonnes, A’s separately labelled 10% pool-intake sample and C’s origin-only evidence. Attach the linked record references and missing-link notes. State the 25% result with this exercise’s evidence test and purchase period. No emissions, environmental benefit, supplier performance or legal-compliance conclusion follows from this coverage calculation.

Source context

GHG Protocol’s Agricultural Guidance identifies existing producer records as possible inputs and says exact data needs depend on the calculation tool.

For Scope 3 inventories, GHG Protocol considers technological, temporal and geographical fit, completeness and reliability when assessing data quality.

For organising replies and outstanding evidence, use the Supplier ESG response tracker.

Discuss a purchased-crop evidence request with Nitium Advisory: +919322161562.