Documenting farm-to-export traceability for coffee means assigning a unique lot code at harvest, then recording every custody handoff — farmer, collection point, wet mill, dry mill, warehouse and port — with dates, weights and geolocation. For Kintamani Arabica, that chain begins in Bangli Regency, north-east Bali, and ends at a Bali or Surabaya port.
Traceability is no longer a nice-to-have for specialty green coffee. Roasters in the EU and US now ask for it before they cup a sample, and the EU Deforestation Regulation (EUDR) has turned plot-level data into paperwork rather than a marketing story. This guide walks through how to build that record for a Kintamani Arabica lot, from the first picked cherry to the loaded container.
What does a coffee traceability record actually need to contain?
A useful record answers four questions for every kilogram: where it grew, who handled it, when it moved, and what it weighed at each step. Miss one and the chain breaks.
For Kintamani lots grown around Kintamani, Ulian Village (Desa Ulian) and Catur Village in Bangli Regency, the minimum data set looks like this:
| Stage | Who records it | Key data points |
|---|---|---|
| Harvest | Farmer / picker | Farmer ID, plot geolocation, harvest date, cherry weight, ripeness |
| Collection | Collection point / koperasi | Intake date, cherry weight in, village/subak group |
| Wet mill | Processor | Process method (washed/natural/honey), parchment weight, drying start |
| Dry mill | Miller | Hulling date, green (biji hijau) weight, screen size, defect count, moisture |
| Warehouse | Exporter | GrainPro/jute bag count, warehouse-in date, lot code confirmed |
| Port | Exporter / forwarder | Container number, port of loading, export documents |
Capture this at the moment it happens, not from memory a week later. A picker’s cherry weight recorded on delivery day is verifiable; a number reconstructed at the mill is a guess.
How do you build a lot code that survives from farm to port?
A lot code is the spine of the whole record. Every document, bag tag and spreadsheet row points back to it. Build it once, at intake, and never reuse it.
A readable structure encodes origin, village or subak, process, harvest year and a sequence number:
| Segment | Meaning | Example |
|---|---|---|
| KTM | Region — Kintamani, Bangli | KTM |
| ULN | Village / subak — Ulian | ULN or CTR (Catur) |
| W | Process — washed (proses basah) | W / N (natural) / H (honey) |
| 26 | Harvest year (2026) | 26 |
| 014 | Sequential lot number | 014 |
So a single washed Ulian lot reads KTM-ULN-W-26-014. Keep that same code printed on GrainPro liners, jute bags, the processing log and the export invoice. When a roaster’s question about a shipment lands, you should be able to pull every stage from one string. That single-code discipline is what lets a buyer move from a bag tag to the underlying full traceability records without a phone call.
How do you segment lots by village and subak?
Kintamani’s terroir is not uniform. Cultivation altitudes reported across sources range from roughly 1,000 up to 1,700 metres above sea level, and the coffee is farmed by smallholders organised into Subak Abian — the Balinese irrigation-and-farming cooperative structure that also governs coffee plots. Segmenting by that structure protects both quality and provenance.
Practical segmentation rules:
- Never blend two villages into one lot code. Keep Ulian and Catur separate even when volumes are small.
- Tie each lot to a named subak group where possible; buyers looking for GI-protected Kintamani Arabica specifically want the Bangli Regency and subak cue on the bag.
- Record altitude bands per plot, not per region, so a screen-17 microlot from a higher block is not averaged into a valley lot.
- Split by process at the same point — a natural and a washed from the same subak are two lots, two codes.
This segmentation is also what allows honest pricing conversations. As of 2026, indicative FOB bands run roughly USD 8–11/kg for washed Grade 1 specialty (screen 16+, cupping around 82–84) and USD 10–15+/kg for microlots and naturals scoring 84–87 and above, with commodity grade well below that and all figures subject to change. Those numbers only hold if the lot behind them is cleanly segmented and its score comes from an actual cupping report — never from a farm’s reputation.
What chain-of-custody documents move with the coffee?
Chain-of-custody is the paper trail that proves the coffee in the container is the coffee described in the record. For green, unroasted Kintamani shipping from Bali or Surabaya to EU and US ports, the core set is:
| Document | Issued / held by | Purpose |
|---|---|---|
| Farmer delivery slip | Collection point | Links picker to intake weight |
| Processing log | Wet / dry mill | Method, weights, moisture |
| Cupping report / COA | Exporter or lab | Green bean quality snapshot per lot |
| Certificate of origin | Chamber / authority | Proves Indonesian origin |
| Phytosanitary certificate | Agriculture quarantine | Plant-health clearance |
| Commercial invoice + packing list | Exporter | Commercial and bag/weight detail |
Grades and quality figures — Grade 1, screen 16+, moisture max 13%, defect value within specialty limits — belong on the COA and the spec sheet, tied to the lot code. Do not carry a cupping score forward on a company letterhead as if it were a fixed attribute of a farm; it describes one lot, one crop, one report.
How does this documentation support EUDR readiness?
Coffee is in-scope of the EU Deforestation Regulation, which means an importer must hold plot geolocation and due-diligence documentation showing the coffee is deforestation-free and legally produced. The farm-to-export record you have just built is most of what that due diligence needs: georeferenced plots at harvest, a village/subak layer, and an unbroken custody chain to the port.
A few honest caveats. Preparing EUDR-ready traceability improves a buyer’s position; it does not guarantee a customs or compliance outcome, and no exporter should promise one. Geolocation for smallholder plots is usually captured as a point or polygon per farm and aggregated to the Subak Abian group — workable, but it has to be collected deliberately during the May–October main harvest, not backfilled afterwards.
Build the record once, at the source, and the same data serves your quality story, your EUDR file, and the roaster who wants to know exactly which corner of Bangli their coffee came from.
Frequently Asked Questions
How long should I keep farm-to-export coffee traceability records?
Retain them for at least five years to align with EU Deforestation Regulation due-diligence expectations, and longer if a buyer’s contract requires it. Keep the full set per lot code — harvest data, processing logs, COA and export documents — in one linked file so any single shipment can be reconstructed years later without chasing separate archives.
Can smallholder lots be traced without individual farm GPS points?
Not reliably under current buyer and EUDR expectations. Each contributing plot should carry a geolocation — a point or polygon — even when many small farms feed one subak group. Points can be aggregated to the Subak Abian group for reporting, but the underlying per-plot coordinates must exist and be collected during harvest, not estimated afterwards.
What is the difference between a lot code and a traceability record?
The lot code is a short unique identifier, such as KTM-ULN-W-26-014, printed on bags and documents. The traceability record is the full dataset that code points to — farmer, plot, process, weights, dates, quality and custody documents. The code is the key; the record is everything the key unlocks.