Collecting plot data from smallholders and cooperatives

The shape of the problem

A European importer buying from a smallholder origin typically sits four or five steps from the land: exporter, consolidator, cooperative, section, farmer. Each step aggregates, and aggregation destroys the plot identity you now need. The commercial chain was built to move volume efficiently; it was never built to preserve provenance at parcel level.

The scale is the second problem. A single cooperative may have two thousand members with three parcels each. That is six thousand geometries to collect, validate, screen and maintain, for one supplier, in one country, for one commodity.

Four routes to the data

1. 1.Direct capture by the producer. A phone app, walked around the plot for a polygon or stood in the middle for a point. Highest quality and highest ownership. The farmer knows their own boundaries better than any registry, but it needs devices, training, connectivity and a season.
2. 2.Cooperative-held registries. Many cooperatives already hold member plot data for internal allocation or certification. Fastest starting point by far, and usually the right first call. Expect gaps, stale boundaries and duplicate members.
3. 3.Certification scheme data. Some schemes hold polygons for certified members. Useful as a base layer, never sufficient alone, and it only covers the certified fraction of a cooperative's membership, which is often a minority.
4. 4.Cadastral records. Authoritative where they exist and are digitised. In many origin countries they are neither, and where informal tenure is common the registry may not reflect who is actually farming.

Start with the cooperative's own data, then verify. The instinct is to launch a field mapping programme. The faster path is usually to take whatever the cooperative already holds, screen it immediately, and use the results to target field effort at the plots that actually need it. Screening is cheap; walking boundaries is not.

Sequencing that works

1. 1.Map your volume, not your suppliers. Rank origins by EU-bound volume and start where the exposure is. A supplier representing 2% of your intake does not deserve the same effort as one representing 40%.
2. 2.Get whatever exists, immediately. Ask every cooperative for what they hold, in whatever format. A messy spreadsheet this month beats a clean survey next year.
3. 3.Screen early and screen everything. Running the deforestation screen on imperfect data still tells you where the problems are. Most plots pass cleanly; you only need to work on the ones that do not.
4. 4.Target field effort. Send people to plots that failed geometry validation, that flagged on screening, or that could not be located, not to the 90% that were fine.
5. 5.Close the loop into procurement. Unmapped plots have to be visible at the point of buying, or mapped and unmapped material bulks together and contaminates the lot.

What goes wrong

• Coordinates that are not plots. The most common failure: a hundred farmers sharing the coordinates of the cooperative office or the village centre. It is detectable instantly on a map and it invalidates the filing.
• Boundaries drawn generously. Polygons extended to "make sure" the plot is covered pull in roadsides, river margins and neighbouring parcels, and with them, tree-cover loss that has nothing to do with the farmer. Over-drawing creates alerts you then have to spend money resolving.
• Reversed coordinates. GeoJSON is longitude first; almost every GPS app shows latitude first. See geolocation formatting.
• Collection outside the season. Some plots are inaccessible, or simply unidentifiable as the right parcel, outside the harvest window. Missing it costs a year.
• One-off collection. Membership changes, plots are sold, new land comes into production. A dataset collected once and never refreshed decays quickly, and the annual review will expose it.
• No consent conversation. You are collecting location data about identifiable people. Explain what it is for and who will see it, both because it is required and because farmers who understand the purpose supply better data.

Who pays

The regulation is silent, which in practice means the cost lands on whoever has least leverage, usually the producer. That is worth resisting on straightforwardly commercial grounds: a cooperative that cannot fund mapping becomes a cooperative you cannot buy from, and replacing a long-standing origin costs far more than the mapping would have.

Article 11 explicitly contemplates supporting supplier compliance as a mitigation measure. Funding the collection is not charity; it is the documented risk mitigation the regulation asks for, and it belongs in your file as such.

Exclusion is the failure mode to design against. The predictable outcome of a rushed programme is that smallholders, the hardest to map, the least able to evidence legality, get dropped from EU chains in favour of large estates that are easier to document. That is the opposite of what the regulation intends, and it is a reputational exposure of its own. Building mapping capacity is the alternative, and it takes the time you have between now and the deadline.

Related

• Geolocation: coordinates, polygons and the four-hectare rule
• Resolving a flagged plot
• Article 11 mitigation