Colombian coffee is the opposite data problem from Brazil. Around half a million families grow coffee on holdings that average well under two hectares, which means a single container can trace to several hundred farms. The good news: almost every one of those plots can be declared as a single point, and Colombia's coffee institutions hold more plot data than almost any origin. This page covers how to use both facts.
Risk tier and what it requires
Colombia is not in the benchmarking annex, so it is standard risk: full Article 9 information collection, a documented risk assessment, mitigation where needed, and a DDS per consignment. The 3% annual inspection target applies to operators sourcing there.
Smallholder points, at scale
With typical farms of 1 to 2 hectares, nearly all Colombian coffee plots fall under the 4-hectare threshold and can be declared as a single six-decimal point rather than a polygon. The volume is the challenge, not the geometry: your plot file per lot will have hundreds of rows, each needing a coordinate, an honest area value and a production date range. The smallholder collection playbook applies directly, with one Colombian addition: steep Andean terrain degrades GPS fixes, so points captured under canopy on a 40-degree slope deserve a plausibility check against imagery before they go in a filing.
The federation's data infrastructure
Colombia's coffee sector runs on institutions. The national federation maintains SICA, a coffee information system that registers growers, farms and plots, built up over decades of extension work. Cooperatives and exporters draw on it, and many can already produce georeferenced grower lists per lot. Certified volumes (a large share of Colombian exports) add scheme-collected farm records on top. None of this discharges your obligation: the data reaches you through your exporter, its quality varies by municipality, and the DDS is still filed on your conclusion. But in Colombia the right first question is rarely "can you map the farms" and usually "can you export what is already mapped, per lot, in GeoJSON". Convert whatever arrives into the TRACES GeoJSON profile and validate it before filing.
Screening Colombian coffee landscapes
The traditional coffee axis (Caldas, Risaralda, Quindío, plus Antioquia, Huila, Nariño and Tolima) is long-settled farmland where post-2020 forest loss inside coffee plots is uncommon. Screening attention concentrates on two patterns. First, shade coffee reads as tree cover in naive datasets, which is why convergence-of-evidence screening rather than a single layer keeps false flags manageable. Second, the deforestation frontier in Caquetá, Putumayo and the Andes-Amazon transition has seen post-conflict land clearing, and coffee has expanded in some of those municipalities. Plots there warrant the full evidence treatment: prior imagery, planting records, and a hard look at production dates against the 2020 cutoff.
Working sequence for an importer of Colombian coffee
- Ask the exporter which growers and plots feed your lot, and whether SICA-derived or scheme-derived plot records exist for them.
- Contract the plot file per lot: point coordinates, area in hectares as a number, grower identifier, production dates.
- Validate the file (coordinate order and range, areas, duplicates), then screen every point.
- Resolve flags with the exporter while the lot is still at the mill, then file in TRACES and archive for five years.
Comparable smallholder origins behave similarly: see Ethiopia for the traceability-poor version of this problem and Vietnam for the low-risk variant where the procedure lightens but the data does not.