Traceability From Farm Gate to Buyer: What Contract Buyers Ask For
Contract buyers, exporters and processors increasingly ask questions that only a record can answer: which farmer, which day, which lot, and what happened to it after the gate. What traceability actually requires, and the honest distance between a record and a certification.
For decades the question a Kenyan aggregator had to answer was how much and at what price. Increasingly it is also: where did this come from, who grew it, when was it harvested, and can you show me? Contract buyers ask because their own customers ask them, and an aggregator who cannot answer is progressively confined to the spot market where price is the only variable.
The good news is that traceability is mostly a discipline about identifiers rather than a technology problem. The awkward news is that the discipline has to start at the buying point, which is the least controlled part of most operations.
A lot is the unit of traceability
Traceability rests on one idea: produce moves through your operation as identified lots rather than as an undifferentiated pile. A lot is created at intake, it can be split and combined, and every movement records which lot was involved.
The decision that matters most is how granular a lot should be, and it is a trade-off you make once. Too fine and your intake process becomes unworkable during a peak week; too coarse and a quality problem contaminates a lot so large that traceability gives you no advantage.
| Lot definition | Suits | The cost |
|---|---|---|
| Per farmer per delivery | High-value produce, certified programmes, small volumes | Heavy at intake; many lots to manage downstream |
| Per collection point per day | Most aggregation operations — the usual right answer | A quality issue implicates one site-day, which is a manageable scope |
| Per route per day | Bulk commodities with low differentiation | Wide implication if a problem is found |
| Per consignment out | Minimum viable — better than nothing | You can trace forward to the buyer but barely backward to the farm |
Per collection point per day is where most operations should land. It keeps intake fast, it maps onto how produce is physically handled, and it limits any quality investigation to one site on one day — a scope you can actually act on.
Traceability is not a technology problem. It is a decision about what a lot is, made once, and then applied by tired people at a buying point during a peak week.
Both directions, and the backward one is harder
Traceability means two distinct capabilities, and operations that have one usually assume they have both.
Forward — where did this go?
- A quality issue is found in a lot you still hold
- You need to know which buyers received from it
- Answered from despatch records against lots
- Relatively easy: you control the paperwork at both ends
Backward — where did this come from?
- A buyer rejects a consignment on a residue or quality finding
- You need to know which farmers and which days contributed
- Answered only if intake recorded the source against the lot
- Hard: it depends on discipline at the buying point months earlier
The backward direction is the one buyers actually test, and it is the one that fails, because it depends on data captured at the noisiest point of the operation by people under pressure to keep a queue moving. Every improvement in intake capture pays off here and nowhere else quite as visibly. The intake discipline itself is covered in member produce intake and payments.
What the buying point must capture
-
Farmer identity, consistently
A farmer number, used every time. Names are unreliable — spellings vary, families share them, and a name typed differently on two days is two farmers as far as any record is concerned.
-
Date, site and weight, at the moment of receipt
Not at the end of the day from a notebook. A weight recorded later is a weight somebody remembers, and it will not agree with the farmer's recollection.
-
Grade or quality observation
Even a simple one. Grade recorded at intake is what lets you separate a farm-level quality issue from something that happened in your own store.
-
The lot the delivery joins
This is the link that makes everything else traceable. Without it you have intake records and stock records that cannot be connected.
The identifier is the whole system
Every part of traceability reduces to whether a consistent identifier was recorded at the moment produce changed hands. Farmer number at intake, lot number through the store, consignment reference on despatch. Three identifiers, applied consistently, answer nearly every question a buyer will ask.
What a record is, and what it is not
This distinction needs to be blunt, because it is where aggregators get into difficulty with buyers. A traceability record shows what your own system recorded. It is not a certification, it is not verified by anyone external, and it does not attest to how the produce was grown.
Organic status, GlobalG.A.P. and equivalent schemes, residue testing, fair-trade claims — all of these require an accredited body, an audit and, usually, laboratory testing. A buyer asking for certification is asking for something a records system cannot supply, and presenting a traceability report as though it were certification damages a commercial relationship badly and permanently.
What the records genuinely do is make certification achievable. Auditors for these schemes ask precisely the questions this article describes — which farmer, which day, which lot, and show me — so an operation with good records is auditable and one without cannot even begin.
What we do and do not do
What AWRA OpsHub does today
- Batch and lot tracking, with allocations recorded as stock moves and depletes.
- Trace events on inventory, so a lot's movement history is retrievable.
- Intake recorded against a supplier or farmer record, with weights, dates and locations.
- Grades captured where you set them up as attributes on receipt.
- Mobile offline capture, so the buying point works without a connection and syncs later.
- Despatch records linked to lots, which is what makes the forward direction answerable.
What it does not do
- No certification management. Nothing tracks scheme membership, audit dates or certificate validity, and no record here constitutes a certification.
- No residue or laboratory test management. Test results can be stored as documents; they are not modelled against lots.
- No GPS farm mapping or plot registry. Farmer records are commercial records, not geographic ones.
- No export documentation. Phytosanitary certificates, certificates of origin and export paperwork are outside this entirely.
Read the first and last lines before any conversation with an export buyer. What you can offer is a complete internal record; what a scheme requires is an external audit of a system, and the record is the input to that audit rather than a substitute for it.
Start where the buyer will test you
Aggregators attempting traceability usually begin by trying to trace everything, and the intake process collapses in the first peak week. The version that survives starts with the produce line and the buyer relationship that actually demands it — typically one contract, one crop — and applies the discipline completely there.
One line traced properly is a commercial asset you can demonstrate. Five lines traced intermittently is a set of records that will fail the first time a buyer tests them, which is worse than not claiming it at all.
Our take
Define a lot as a collection point on a day, record a consistent farmer number at every intake, and make sure the lot travels with the produce through store and despatch. Start with the one crop your best buyer cares about. And never let a traceability report be described as a certification — that is a different thing, requiring an auditor, and confusing the two costs relationships that took years to build.
See batch traceability
Lot and batch tracking with allocations on every movement, intake recorded against farmer records, and offline capture at the buying point.
Explore traceabilityFrequently asked questions
How granular should a lot be?
Per collection point per day suits most aggregation operations. Per farmer per delivery is right for high-value or certified produce but makes intake heavy during peak weeks, and per route per day means a quality finding implicates far more produce than you would like. The trade-off is between intake speed and how wide a problem spreads when one is found — decide once and apply it consistently rather than varying by season.
Does traceability mean we are certified?
No, and this distinction must never be blurred with a buyer. A traceability record shows what your own system recorded; certification — organic, GlobalG.A.P., fair trade — requires an accredited body, an audit and usually laboratory testing, none of which a records system provides. What good records do is make certification achievable, because auditors ask exactly the questions traceability answers. Presenting one as the other damages commercial relationships permanently.
Why is backward traceability harder than forward?
Because it depends on data captured at the buying point, months earlier, by people under pressure to keep a queue moving — while forward traceability depends on despatch paperwork you control at both ends. Backward is also the direction buyers actually test, since it is the one that matters when a consignment is rejected. Every improvement in intake capture pays off almost entirely here.
What is the single most important field to capture?
A consistent farmer identifier. Names fail — spellings vary, families share them, and the same person typed two ways is two suppliers as far as any record is concerned. A farmer number used at every intake is what connects a rejected consignment back to a specific set of deliveries, and it is worth the effort of issuing and enforcing even when it slows the first few weeks.
Can you store our laboratory test results?
As documents, yes; modelled against lots with pass or fail status, no. There is no test-result management and no certification tracking, so if residue testing is central to your buyer relationships, plan for that record to live alongside rather than inside. The lot history, intake records and despatch links are what the tests attach to, and that connection is currently made by your own referencing.
