EUDR’s extra year is really a data story

The European Union has pushed back the start of the EU Deforestation Regulation (EUDR) again. The new start dates are 30 December 2026 for large and medium operators and traders, and 30 June 2027 for micro and small enterprises.

On paper, that looks like breathing space. In practice, it shifts the pressure to a single question that supply chains struggle with even in calm years. Who pays to create, clean, store, and defend the data that proves a shipment is “deforestation-free” and legally produced?

What changed, and what did not

The postponement was set through an amending act published in December 2025.

A parallel change matters for how data moves through the chain. The EU’s own guidance now reflects a “first operator” logic. The first operator placing relevant products on the EU market submits the due diligence statement (DDS). Downstream operators and traders do not file fresh DDSs. They keep and pass on a reference number linked to the initial declaration.

That sounds like simplification, and it is in administrative terms. It does not reduce the amount of upstream data that must exist. It also does not remove the legal exposure if the underlying data is weak, inconsistent, or unverifiable.

The non-negotiable core: plot-level traceability

EUDR does not accept “country of origin” or “region of origin” as proof. It is built around traceability to the plot of land.

For covered commodities and relevant products (cattle, cocoa, coffee, palm oil, rubber, soya, and wood, plus listed derived products), operators must show the goods are deforestation-free and legal.

Deforestation-free has a fixed cut-off date. Products cannot come from land deforested after 31 December 2020.

The EU also expects geolocation coordinates for each plot involved in production, recorded and submitted as part of the DDS through the EU information system. If the geolocation is missing, the product cannot be placed on the EU market.

This is why the delay is a data story. The most expensive part is not a policy memo or a supplier letter. It is building a dataset that holds up when a customs authority asks, “Show me the plot.”

Why “more time” can still feel like a squeeze

The delay was not driven by a change in ambition. It was driven by readiness concerns, including the functioning of the digital information system and the practical ability of companies and administrations to handle the workload.

That matters for exporters in Asia-Pacific because the EU system is only one side of the pipe. The other side is the supply chain’s own data infrastructure. A typical APAC export supply chain has features that make EUDR data hard work.

Smallholder-heavy sourcing, which is common in cocoa, coffee, rubber, and palm oil, means thousands of plots and frequent aggregation. One weak link can contaminate a batch if the chain lacks segregation controls.

Land tenure and documentation vary sharply by country and by province. “Legal production” is not a single document. It can mean land-use rights, harvest permits, environmental approvals, and labour-related compliance under local law.

Supplier tiers are long. Data must travel from farms and intermediaries to processors, traders, and finally the EU-facing operator who files the DDS. Each handover creates loss, duplication, and mismatch.

Then there is the reality of data quality. Coordinates can be wrong by a few hundred metres. Names of farms can be inconsistent. Plot boundaries can overlap. A registry might not match the ground. EUDR does not reward “close enough”.

The real cost centres sit upstream

When people say “EUDR compliance cost”, they often picture an EU importer filling in fields on a portal. The expensive work happens earlier.

Geolocation collection costs money. Even if smartphones and GNSS devices are common, there is still training, rework, and supervision. There is also governance: who is authorised to collect, edit, and approve location data.

Document capture and verification cost money. A farm may have paper records. A mill may have partial records. Documents can be local-language and non-standard. Translating is not enough. The buyer needs confidence that the document set meets EUDR’s “legal production” expectation.

Supplier mapping costs money. Mapping is not only “who do we buy from”. It includes mass-balance logic, chain-of-custody design, and whether mixing can be ruled out. For many commodities, that requires process changes, not just data entry.

Audit and assurance cost money. Even with good internal controls, companies will want independent checks on geolocation sampling, documentation completeness, and traceability logic. This cost rises when supply chains rely on many intermediaries.

Dispute handling costs money. When a shipment is questioned, the business must respond fast with evidence. That is staff time, legal time, and often re-testing or re-verification.

The postponement to 30 December 2026 does not remove these costs. It spreads them. That can be good if the chain uses the extra year to build a cleaner dataset. It is bad if the chain treats the extra year as a pause.

So who pays?

In many APAC supply chains, the first EU-facing operator will try to push costs upstream, because they carry the formal obligation to file the DDS and face penalties if the due diligence fails. Reuters has reported that penalties under the regime can include fines linked to EU turnover, which gives boards a strong incentive to treat EUDR as a hard compliance risk rather than a sustainability add-on.

But a pure “supplier pays” approach often fails for three reasons.

First, the suppliers who can least absorb cost are often the ones controlling the critical data, especially smallholders and local aggregators.

Second, unilateral cost-shifting weakens data honesty. If suppliers fear chargebacks, they may avoid disclosure of gaps, which makes the dataset brittle.

Third, the importer still needs confidence. Paying less does not reduce liability if the dataset collapses under scrutiny.

In practice, the chains that cope best tend to treat EUDR data as shared infrastructure. Cost follows control and benefit, not hierarchy.

A workable rule of thumb is that the party that needs a dataset for multiple markets and multiple buyers should invest more in the core data layer, because the asset can be reused. The party that needs data mainly to satisfy a single buyer’s template should not be left to fund the entire system.

Four funding models that are showing up

These models are not mutually exclusive. Most large supply chains will mix them.

Contracted cost-sharing is the simplest. Buyers and suppliers agree a per-tonne or per-container fee that funds mapping, document verification, and system upkeep. The key is that the fee must be tied to service levels, such as re-mapping frequency, dispute support, and dataset refresh cycles.

Premium for verified supply is commercially cleaner. Suppliers who provide plot-level data and robust documentation receive a price premium or longer contracts. This works best where products can be segregated, so the premium is attached to a definable stream rather than a mixed pool.

Pre-competitive pooling works where many buyers share the same origin landscapes. Firms co-fund baseline mapping, smallholder support, and documentation standardisation. They then compete on procurement and performance, but they do not duplicate the same geolocation work village by village.

Public and development co-finance can fill gaps, especially where smallholders risk exclusion. Several governments and agencies have already treated EUDR delay as a window to upgrade national traceability tools and support exporters.

What supply chains should do with the extra year

This is the practical part. The delay only helps if it is used to reduce future uncertainty.

Start with “data triage”, not full coverage. Identify the highest-risk origins and the most fragmented tiers. Fix those first. This lowers the chance that the first enforcement waves hit the weakest parts of the chain.

Design the chain-of-custody decision early. Decide which streams will be segregated and which will rely on controlled mixing. The data architecture depends on that choice.

Build a “defensible dossier” template. It should include plot geolocation files, land-use/tenure documents, evidence of legality checks, and the internal control narrative that shows how data is governed and updated. This makes responses faster when goods are questioned.

Treat reference numbers as an internal control, not an admin field. If downstream operators rely on reference numbers instead of filing fresh DDSs, then reference numbers become a compliance-critical identifier. Put them under the same control discipline as certificates and licences.

Run one stress test per commodity stream. Pick a shipment and do a “customs drill”: trace it back to plots, produce the legality evidence, and check whether the dataset survives a sceptical reader. Most gaps show up quickly in drills.

The bottom line

EUDR’s postponement to end-2026 is not a retreat from plot-level accountability. It is a signal that the EU’s system and the private sector’s data plumbing were not ready at scale.

For Asia-Pacific exporters, the risk is not only compliance cost. The bigger risk is market exclusion through data failure, especially where smallholders and intermediaries cannot produce plot-level evidence fast enough.

If there is one strategic shift worth making now, it is this. Stop treating EUDR data as a one-off reporting exercise. Treat it as shared supply-chain infrastructure, with a clear funding model and clear rules on data ownership, upkeep, and dispute support. The extra year is only valuable if it buys credibility.