EUDR for cocoa and chocolate businesses

Cocoa has the widest consumer-facing footprint of any EUDR commodity: the regulation covers not just beans but chocolate itself. A ten-person bean-to-bar maker importing two pallets of beans from Ecuador has obligations under Regulation (EU) 2023/1115, with a deadline of 30 December 2026 - or 30 June 2027 for micro and small enterprises. This guide covers the cocoa-specific scope, the West African data reality, and what a chocolate business actually files.

In-scope products: beans to bars

Annex I covers the full cocoa chain:

1806 is the one that surprises people: filled bars, couverture, drinking chocolate and cocoa-containing confectionery are all in scope. If you import finished chocolate from Switzerland or the UK into the EU, you are an operator placing an Annex I product on the market - the due diligence obligation lands on you, traced back to the plots where the cocoa was grown.

Who files what in a typical chocolate chain

• Importing beans, paste, butter or powder from origin: you are the operator; full due diligence and a DDS.
• Making chocolate from cocoa products bought inside the EU: the upstream importer filed; as an SME you reference their DDS numbers instead of filing new statements.
• Importing finished chocolate from outside the EU: you file, and you need plot geolocation from your manufacturer's bean supply - negotiate for it now, not in December.
• Exporting chocolate out of the EU: export requires a DDS too.

The West African data problem

Côte d'Ivoire and Ghana grow around 60% of the world's cocoa, overwhelmingly on smallholder plots under 4 hectares. Two consequences follow. First, most plots can be declared as single points at six decimal places rather than polygons - see the geolocation guide for exactly when a point suffices. Second, both governments have built national traceability systems that assign farm and plot identifiers, and large exporters increasingly ship with plot files attached. Ask your supplier for the plot data they already hold before assuming it does not exist; for indirect ("pisteur") supply chains, expect gaps and lead time. The two origins now diverge procedurally, so we cover each separately: Côte d'Ivoire (standard risk, national mapping push) and Ghana (low-risk tier, COCOBOD mapping, and the cross-border mixing trap). Collection tactics for the farm-gate level are in the smallholder geolocation playbook.

Cocoa-driven deforestation is the reason this regulation exists in the form it does: documented forest loss inside protected areas in West Africa made "trust the paperwork" untenable. Screening is therefore empirical. Each plot is compared against tree-cover, land-use and deforestation-alert datasets for evidence of clearing after 31 December 2020 - the convergence-of-evidence approach of the open WHISP methodology, which was piloted heavily on West African cocoa. Shade-grown agroforestry cocoa can resemble forest in a single dataset; agreement across several is what separates "agroforestry since 2015" from "cleared in 2022".

Country benchmarking and your checks

Major cocoa origins sit in the standard-risk tier of the EU country benchmark, meaning authorities aim to check 3% of operators annually. The practical implication for an SME chocolate maker is not the inspection odds - it is that standard risk requires the full due diligence cycle: information collection, risk assessment, and documented mitigation where risk is more than negligible. Your evidence file, kept five years, is what makes that defensible.

Mass balance and mixed lots

Cocoa is blended at every stage: fermentation, export lots, grinding. The regulation does not accept book-and-claim certificates as a substitute for geolocation - the DDS must carry the plots for the actual physical supply shed of the product. Where a lot aggregates many farms, you declare all of them. If any single declared plot shows post-2020 deforestation, the consignment has a problem, so screening the full plot list before contracting is cheaper than discovering a red plot at filing time. The DDS field guide shows how multi-plot geolocation files attach to one statement.

Certification data: reusable, not sufficient

A large share of EU-bound cocoa is Rainforest Alliance or Fairtrade certified, and both schemes have collected farm polygons for years. That data is genuinely reusable: ask certified suppliers to release the plot geometry attached to your certified volumes, and you may find most of your geolocation problem already solved. What certification does not do is discharge the legal obligation. The regulation is explicit that certification schemes can inform your risk assessment but cannot replace it - the DDS is filed in your name, on your conclusion, with your plots. Treat scheme data as a head start on collection and one input to the risk file, and keep the scheme's own audit reports alongside it as supporting evidence.

For small direct-trade importers the pattern is different but easier: two or three known farms, a personal relationship, and plot boundaries that can be walked with a phone in an afternoon. If that is you, your gap is rarely data collection - it is TRACES registration and knowing the filing mechanics.

Order of operations for a chocolate business

1. Classify your products by HS code and split purchases into "we import" versus "bought in the EU".
2. For imports, request plot geolocation per lot from suppliers - Ghanaian and Ivorian exporters increasingly provide GeoJSON on request.
3. Validate and screen plot files on receipt; push flagged plots back to the supplier immediately.
4. Set up EU Login and TRACES access, run a test DDS, then file per consignment or annually for recurring identical sourcing.

The same aggregation logic applies to coffee, while palm oil shows what happens when derivatives get involved and soy where scope boundaries cut differently.