Indonesian Coffee Export to EU: REX & EUDR 2025 Buyer Guide

A practical, buyer-ready checklist for EUDR geolocation in Indonesian coffee. What data to request, exact coordinate formats, how to prove no deforestation after 2020, and a sample CSV/JSON field list you can use today.

If you’re buying Indonesian coffee for the EU in 2025, you’re juggling two realities. REX keeps your customs paperwork clean. EUDR geolocation keeps your coffee on the market. We won’t cover tariff preferences here, but we’ll show you the exact geolocation and proof stack EU buyers are asking for and the workflow we use internally. This is the EUDR geolocation Indonesian coffee playbook we wish everyone had last year.

The buyer-ready checklist we actually use for EUDR

In our experience, EUDR success rests on three pillars. If one wobbles, the whole file gets questioned.

Geolocation to the plot. Use WGS84 decimal degrees. Points for plots up to 4 ha. Polygons for plots over 4 ha. Include area in hectares and harvest year.
Clean chain-of-custody. A lot number should trace to a set of farm IDs with weights that add up. Any aggregation must be transparent.
Deforestation evidence. Show the plot existed before 31 Dec 2020 and no forest conversion after that date. Keep screenshots, layers, and a written note of your check.

What’s interesting is that once buyers see those three elements, approvals move quickly. The rest is housekeeping.

Quick note on REX vs EUDR

REX is about origin declarations for tariff treatment. EUDR is about deforestation-free and legal origin. You need them both in many cases, but they’re separate workflows. We make sure REX details appear on commercial docs, then keep the EUDR due diligence statement and geolocation datasets in a separate compliance pack.

Week 1–2: Map your source and validate formats

Here’s the thing. Most delays happen because coordinates arrive in the wrong format or without area. We standardize up front.

Coordinate system. WGS84 (EPSG:4326). Decimal degrees. Six decimals is a good practice. Example: -8.303211, 115.279441.
Point vs polygon. If a coffee plot is 4 hectares or less, a single point is acceptable. Over 4 ha requires a polygon (minimum three vertices plus closure). If in doubt, collect a polygon.
Accuracy. Typical smartphone GNSS is 3–10 m. That’s fine. Record accuracy readout if your app provides it.
File formats we accept. CSV for tabular data, plus GeoJSON/KML for polygons. We keep a simple CSV master and link out to geometry files.

Free offline tools we’ve used on Indonesian smallholder mapping:

SW Maps or QField on Android. Offline basemaps, forms, and polygon capture.
ODK Collect/KoboCollect for enumerator-friendly forms. Works well in patchy signal.
Organic Maps or MAPS.ME for backup waypoints.

Takeaway: Decide your toolset on day one. Force WGS84, decimal degrees, and a consistent field list across all suppliers.

Week 3–6: Build your MVP traceability and test submissions

We create a minimal but auditable chain-of-custody. No fancy blockchain needed.

Farm master table. One row per plot. Fields: farm_id, farmer_name, village, district, province, geotype (point/polygon), coords_or_file, area_ha, altitude_m, variety, processing, last_update.
Lot-farm link table. One row per farm contribution to a lot: lot_number, farm_id, harvest_year, weight_kg, collection_date, aggregator_id.
Shipment table. One row per physical shipment: shipment_id, lot_number(s), bags, net_weight, ICO mark, invoice_no, packing_list_no, container_no.

We then test an EUDR due diligence statement draft in the EU Information System sandbox. As of late 2024, the EU hasn’t published country risk benchmarking, so assume standard due diligence for Indonesia. Keep a PDF of the submission and the unique reference ID.

Need help pressure-testing your dataset or want our ready-to-use CSV template and SOP? Contact us on whatsapp.

Week 7–12: Scale and optimize

Once the MVP works on one supply chain, roll it out origin by origin. We find it easier to start with single-origin supply where farmer groups are already organized. For instance, buyers wanting a clean EUDR file often start with mapped single-origin lines like Arabica Bali Kintamani Grade 1 Green Coffee Beans. Multi-region blends are doable too. We publish clear farm lists and links in our lot specs, including for profiles like Blue Batak Green Coffee Beans, then tie them to aggregation weights.

Do I need polygon coordinates for smallholder coffee plots in Indonesia under EUDR?

Short answer. Only if the plot is larger than 4 ha. Under or equal to 4 ha, a point is accepted. In practice, we capture polygons for clustered smallholders when boundaries are known. It future-proofs your file if holdings get consolidated.

What format should I use for EUDR geolocation?

Use decimal degrees in WGS84. Example fields: latitude, longitude with six decimals. For polygons, provide a GeoJSON or KML with coordinates in WGS84 and include the area_ha in your attribute table. Avoid UTM, DMS, or national grids.

How can I map hundreds of small coffee farms without visiting each one?

Three approaches we’ve actually used:

Enumerator hub-and-spoke. Train one local enumerator per village to capture 30–50 farms each using SW Maps or ODK. Typically two weeks for 500 farms if transport is organized.
Remote digitizing with farmer validation. Draw polygons from recent imagery, then validate with farmers via WhatsApp video or village meetings. Capture consent and corrections.
Cooperative records plus spot checks. Start with existing membership lists, map a representative sample in person, then roll out the rest with remote validation.

What proof shows my Indonesian coffee wasn’t grown on land deforested after 2020?

We combine geospatial and documentary evidence:

Geospatial. Overlay your plot geometries on Global Forest Watch tree cover loss layers and Sentinel-2 time series. Save screenshots showing presence before 2020 and no new clearing after 31 Dec 2020. For Indonesia, we also consult KLHK deforestation maps when available.
Documentary. Farmer statements on land history, local letters, cooperative membership records. Not a substitute for geospatial proof but helpful context.
Written risk assessment. One paragraph per supply chain stating method, data sources, dates, and outcome.

How do I link farm GPS points to a specific shipment, lot, and invoice for EUDR?

Create a simple mapping chain:

Farm master. Each plot has a farm_id.
Lot-farm link. Each lot references all contributing farm_ids with exact weights.
Shipment. Each shipment references lot_number(s), invoice number, and container.
On bags. Use lot_number plus internal bag IDs. We often print QR codes in the warehouse.

When an auditor asks “Which farms are in this container?” you can filter on shipment_id and follow the joins.

What if my supplier won’t share precise farm locations—can I still place the coffee on the EU market?

No. Under EUDR you must submit geolocation of plots where the coffee was produced. Workarounds that hide GPS data or aggregate at district level won’t pass. What can work is a data escrow or NDA, or limiting visibility to the EU Information System submission only. We’ve had success with NDAs that specify who can access raw coordinates and for what purpose.

Which free offline apps work best to capture EUDR-compliant coordinates from farmers?

SW Maps. Easy polygons, photos, and custom forms. Good offline.
QField. Powerful, ideal if you already manage QGIS projects.
ODK Collect/KoboCollect. Form-first, great for enumerator teams.
Organic Maps/MAPS.ME. Lightweight waypoint backup when batteries run low.

Our tip. Preload village basemaps and a simple form with required fields before fieldwork.

Sample EUDR field list you can copy

Use one CSV for the master and link to geometry files.

Farm master CSV fields

farm_id
farmer_name
group_or_coop
village, district, province, country
geotype (point or polygon)
latitude, longitude (if point)
geometry_file (GeoJSON/KML filename if polygon)
area_ha
altitude_m
variety/process (e.g., Gayo, fully washed)
harvest_year
consent_record (Y/N + file ref)
last_update

Lot-farm link CSV fields

lot_number
farm_id
harvest_year
weight_kg
collection_date
aggregator_id

Shipment CSV fields

shipment_id
lot_number(s)
bags
net_weight_kg
invoice_no
packing_list_no
container_no
export_date

JSON geometry example (stored per plot)

type: Feature
properties: {farm_id, area_ha, last_update}
geometry: Point [lon, lat] or Polygon [[[lon, lat], ... , [lon, lat]]]

Due diligence statement notes to include

Operator/trader identity and role
Product: coffee, CN 0901, origin Indonesia
Geolocation reference to your farm master (attach CSV and geometries)
Risk assessment method and sources (GFW, Sentinel-2, KLHK) with dates
Risk mitigation steps taken (field audit, enumerator checks, NDAs)
Statement of negligible risk and compliance with the deforestation cut-off date (31 Dec 2020)

5 mistakes we see that sink EUDR submissions

Missing area_ha on polygons. Auditors often ask for it. Calculate once in QGIS.
Wrong coordinate system. UTM or DMS sneaks in. Force WGS84 decimal degrees.
Lot weights don’t add up. If 18,900 kg shipped, your farm weights should reconcile within a small tolerance.
No dated screenshots for deforestation checks. Save images with timestamps and plot IDs.
Privacy missteps. Share coordinates only with consent. Use NDAs, and store PII separately from geometry wherever possible.

Trend watch for 2025

Country benchmarking is delayed. Treat Indonesia as standard due diligence for now.
The EU Information System is stabilizing. Expect stricter validation on geometry schemas and file attachments in the coming months.
Buyers are shifting to origins with known farmer groups first. If you’re building a 2025 program, start with organized origins, then add complex blends.

Final takeaways

Decide your formats up front. WGS84, decimal degrees, CSV master plus GeoJSON/KML.
Build a minimal chain-of-custody that reconciles. Farm → Lot → Shipment → Invoice.
Prove no deforestation after 2020 with geospatial evidence and a short written risk assessment.

Questions about your current dataset or want our field templates pre-filled for Bali, Java, Gayo, and Sumatra origins? Contact us on whatsapp. If you’re evaluating mapped lots with unique processing profiles, you can also View our products to see which single-origin or blended lines align with your compliance timeline.