Supply chain fraud detection across buyers: duplicates, volumes and side-selling
Duplicate farmers, impossible volumes and side-selling, detected in your supply base and, if buyers choose, across buyers.
Farmer1st's fraud detection finds duplicate farmers, volumes that exceed what a plot can produce, and side-selling patterns. It always works on your own supply base, and buyers who choose to pool it can also catch patterns across buyers, without exposing one buyer's data to another. Each buyer receives alerts on its own supply base and keeps its data private.
Fraud detection
Duplicate farmers
The same person registered several times, by one buyer or by several, is detected and resolved to one verified record.
Impossible volumes
Deliveries are compared with what the farmer's plots can plausibly yield, so laundered volumes stand out.
Side-selling across buyers
Patterns that only appear when several buyers' data is combined are detected without revealing who bought what.
Private by design
Each buyer sees alerts about its own suppliers. Prices, volumes and relationships stay private.
Explainable alerts
Every alert comes with the reason it was raised, so teams can check it in the field before acting.
How it works
- 01
Connect your data
Farmer, plot and transaction data from your systems join your private space through data integration.
- 02
Check, alone or across buyers
Records are checked within your supply base and, among buyers who choose to pool detection, across their records, with privacy controls that keep each buyer's data separate.
- 03
Raise alerts
Duplicates, implausible volumes and suspicious patterns are flagged to the buyer concerned, with the reason.
- 04
Verify and resolve
Field teams verify alerts on the ground; outcomes are recorded and improve future detection.
What you get
- Volumes you can defend in due diligence
- A cleaner farmer registry
- Fraud caught that no single buyer could see
- Commercial data that stays yours
Why is fraud a first-mile problem?
In smallholder supply chains, product changes hands several times before it reaches an exporter: farmer, collector, cooperative, trader. At each step, it is possible to add volume from unknown sources, register a farmer twice, or sell the same declared harvest to more than one buyer.
This matters more than ever. Under the EUDR, each due diligence statement must be backed by the plots that produced the product. If volume from unverified land is mixed in, the declaration no longer matches reality.
What does Farmer1st’s fraud detection look for?
| Pattern | What it means | How it shows up |
|---|---|---|
| Duplicate farmers | One person registered several times | Matching identities, plots or contacts across records |
| Impossible volumes | More product than the plots can produce | Deliveries compared with plot area and plausible yield |
| Side-selling across buyers | The same declared harvest sold to several buyers | Combined volumes that exceed a farmer’s capacity |
| Ghost plots | Plots that do not exist or belong to someone else | Overlaps and inconsistencies in geolocation |
Every alert explains the reason it was raised, so teams can verify it before acting.
Why does it have to be pre-competitive?
Some fraud cannot be seen by a single buyer. If a farmer declares a full harvest to two buyers, each one sees a plausible volume. Only when both sets of data are compared does the problem appear.
On Farmer1st, fraud detection can be pre-competitive when buyers choose it: the buyers who agree pool the checks, on terms they set together, so patterns that span several of them become visible. They benefit from a cleaner supply base, and the cost of detection is shared instead of paid separately by each buyer. Buyers who do not pool still get every check on their own supply base. See the pre-competitive network for the principle.
How is each buyer’s data kept private?
Detection across buyers does not mean sharing data between buyers. The system compares records with privacy controls that keep each participant’s information separate:
- each buyer sees alerts about its own suppliers only;
- other buyers are never named in an alert;
- prices, volumes and supplier relationships stay private to the buyer that owns them;
- farmers remain in control of the data they have agreed to share.
What is shared is the result of the check, not the underlying commercial data.
How do alerts turn into action?
An alert is the start of a verification, not a verdict. Typical next steps:
- Review the alert and its reason.
- Verify on the ground through a mission planned in field agent management.
- Resolve: merge duplicate records, correct a plot, adjust volumes or change the supplier’s status.
- Record the outcome, which improves future detection.
Resolved cases feed directly into traceability and certification and EUDR compliance, so your declarations stay consistent with what you have verified.
Where does the data come from?
From the tools you already use. Farmer, plot and transaction data joins the network through data integration, whatever system or spreadsheet it lives in today. Data collected in the field with Farmer1st is checked the same way.
Farmer1st is the first-mile infrastructure built around farmers: a farmer-controlled network where companies, governments and organisations collect, verify and share origin data, and build the services they need on top of it.
Can detection follow our own risk rules?
Yes. Yield assumptions, thresholds and the patterns you care about can follow your risk policy. Modules switch on as you need them, and anything specific to your supply chain, we prototype with you in the Nursery.
Frequently asked questions
What kinds of fraud does it detect?
Mainly three: duplicate farmers registered several times, volumes that exceed what a farmer's plots can plausibly produce, and side-selling or laundering patterns that appear when several buyers' data is combined. Other patterns can be added to match your risks.
Can other buyers see our suppliers, prices or volumes?
No. Even when buyers pool detection, each buyer only sees alerts about its own supply base. Commercial data such as prices, volumes and supplier relationships is never visible to other buyers.
Why does fraud detection need to be pre-competitive?
Some fraud is invisible to a single buyer. If a farmer sells the same declared volume to two buyers, or appears under two identities, each buyer sees only half of the picture. Only a check pooled across buyers can find it, which is why buyers can choose to pool it, and those who do benefit from a cleaner supply base.
How does this help with EUDR?
Laundered volumes can bring product from unverified or non-compliant plots into a compliant shipment. Checking volumes against plot yield and detecting side-selling strengthens the link between each DDS and the plots it declares.
What happens when an alert is raised?
The alert explains why it was raised. Your team can send a field mission to verify, request documents or adjust the supplier's status. The outcome is recorded with the alert.
Related services
Traceability & certification
From plot to lot: first-mile chain of custody, mass balance and audit-ready certification records, with origin attested by farmers.
Explore the module →EUDR compliance
From plot polygons to a submitted DDS: EUDR due diligence built on verified first-mile data, sent directly to EU TRACES.
Explore the module →Data integration
Any tool, any spreadsheet, one verified record: AI-assisted mapping checked by a person, with every version kept.
Explore the module →Field agent management
Missions, routes, offline collection and supervision for the teams who gather first-mile data, with quality checked as it comes in.
Explore the module →Tell us about your supply chain.
Ten minutes with our scoping assistant. We come back with a first sketch of your service.
