Most traceability programmes start the same way: a field officer stands in a smallholder’s plot, opens a phone, and takes a photo with GPS coordinates attached. It feels like proof. It ends up in a folder. Eighteen months later a buyer’s auditor asks a simple question — which delivery did this plot produce? — and nobody can answer it.
The photo was never the problem. The problem is that a coordinate is not a claim anyone can check.
What a buyer is actually asking
Due-diligence questions from European and Southeast Asian buyers are remarkably consistent. They are not asking “do you know where the farmer is.” They are asking:
- Identity — is this the same farmer across every delivery, or five spellings of one name?
- Plot — which specific area of land, with defined boundaries, produced this volume?
- Continuity — what happened between the plot and the mill? Who took custody, when, and how much?
- Reconciliation — does the recorded volume make physical sense for that plot area?
A geotag answers none of these on its own. It is one attribute of one record, and usually the only one that survives.
The three failures that show up in audits
Duplicate farmers. The same person is registered twice because the first record had a typo in the village name. Volume doubles and the audit fails on plausibility, not on deforestation.
Undefined boundaries. A pin marks the centre of a plot. It says nothing about the edge. If two neighbouring farmers both claim land around the same point, there is no way to see the overlap from coordinates alone.
Broken handover chain. The plot record exists. The mill intake record exists. Nothing links them. In a chain-of-custody audit, missing linkage is treated as missing evidence — the product is unverified regardless of how good the farm-level data is.
What a verifiable record looks like
A traceability record becomes evidence when it is a linked structure, not a folder of media:
| Layer | What must be captured | Why it matters in an audit |
|---|---|---|
| Farmer | Stable ID, name, village, contact | Prevents duplicate identities |
| Plot | Polygon boundary + area, not a single point | Detects overlap and checks yield per hectare |
| Handover | Date, parties, weight, lot reference | Establishes custody transfer |
| Reconciliation | Plot area vs. delivered volume | Flags implausible claims before the auditor does |
The polygon is the important shift. A boundary can be measured, compared, and checked against delivered volume. A pin cannot.
The practical implication for exporters
If you are preparing for a buyer audit, the sequence that survives scrutiny is boring but reliable:
- Register the farmer once, with a stable identifier that does not change when names are spelled differently.
- Map the plot as a boundary, and store the computed area.
- Record every handover with weight and lot reference at the moment it happens, not reconstructed later.
- Keep the chain queryable in one direction — from a delivered lot back to the plots that produced it.
That last point is what separates a system from an archive. An archive answers “do we have a photo of this farmer.” A system answers “which plots produced lot X, and does the volume add up.”
QThink-X builds this record structure for Indonesian commodity supply chains. If you are preparing for an EUDR or GDST-related buyer audit, the bottleneck is usually the handover chain, not the farm data.
Primary sources
Preparing for a buyer audit?
The platform records farmers, plots and every handover in one auditable trail — exportable for a buyer, an auditor or a lender.
Request a traceability review