Pick one lot you already sold, follow its papers back to the field it grew in, and write on one sheet how long that took and where one paper stopped leading to the next. A lot’s origin is spread across the sale document (the settlement or invoice), the freight and scale ticket, and the bin or storage record if you keep one. The last paper is the field log, where you write what was done in each field and when. Without that test, traceability is a guess, and you find out whether the papers connect only when a buyer or an auditor asks where a lot came from.
The fix is to run that test once, on purpose, before anyone asks. Food-safety programs call it a mock recall, a practice run of pulling a product back, and here it is aimed at a single lot. Take a lot you already sold, follow its records back to the field and forward to the buyer, write down how long the search took, and mark where the records stopped connecting. You end with one sheet: the lot, its path from the sale document to the field, the time it took, and the break point.
Why is traceability unproven until you trace one lot?
Traceability is the ability to follow a lot, not a pile of papers about it. The Codex Alimentarius Commission, the international food standards body of the FAO and the WHO, defines it as the ability to “follow the movement of a food through specified stage(s) of production, processing and distribution”. Having receipts is not the same as being able to run that movement backward on request, and the gap shows only when someone asks.
Perennia, a food and agriculture agency in Nova Scotia, puts it plainly in its recall guidance: “testing your recall plan is just as important as having one”. If you have a written recall plan, tracing one real lot is how you test it; if you do not, the test lets you see what such a plan would have to cover.
How do you follow the records back to the field?
Each record covers only one step: from the sale document you get to the load, from the load to the bin, from the bin to the harvest, and from the harvest to the field. Reaching the field means linking those steps one after another, and where one record is missing, the trace stops.
The links hold while every hand-off carries the same lot number, and they give way where product is pooled. Researchers led by Pearson, writing in the journal Global Food Security, note that this step-by-step tracing “can be easily lost in commodity products that are blended”, like grain from several fields poured into one bin or milk from several farms into one tank. Past that point, one lot no longer leads back to one field.
Which lot would you least like to be asked about?
Test a real lot you already sold, and make it the hardest one, not the tidiest. Perennia’s procedure is to “select a product, ingredient/input that has been sent to a customer” and trace that particular lot. A lot picked because its papers are clean tells you nothing about the day the question is real, so pick the load you would least want a buyer to ask about. Start from the sale document, the settlement or invoice, because that is where the buyer and the auditor start.
Keep the test on the farm. Perennia advises you to “not let customers know you are performing a mock recall”, since a call asking a customer to confirm a shipment can be taken for a real recall and set one off. You are checking your own records against each other, not raising an alarm.
Where do the records stop connecting?
Finding the break point is what the exercise is for, not a sign it failed. A trace often stops where the records pass from one business to the next, from the farm to the trucker or the elevator, and no shared lot number carries across. Your own records can line up from field to truck and the trail still stops at the elevator, because the link between the two businesses is missing.
An oversized lot breaks the trace a second way. The Economic Research Service, the research agency of the US Department of Agriculture, notes that systems with “large tracking units, such as an entire feedlot or grain silo, will have poor precision” when a problem has to be isolated: a lot that big leads back to many fields instead of one.
Many breaks are low-tech. Pearson’s team notes that the food industry is “still largely reliant on paper records” that are rarely made searchable, so a trace often stalls not for a missing paper but for the hour it takes to find one. Mark the spot where you stopped: the hand-off with no lot number, the blended bin, or the record you could not find. With that mark you know which link to fix before the next sale.
How long does the trace take?
The time the trace takes matters as much as whether it reaches the field, because when a lot has to be pulled back, a slow answer can come too late. In the United States, the FDA’s Food Traceability Rule will set that clock for the foods it covers, but it is not yet in force: compliance, scheduled for January 20, 2026, will not be required before July 20, 2028, by order of Congress. From then on it will require that traceability records reach the FDA within 24 hours of a request, as an electronic sortable spreadsheet.
That is one country’s rule, cited to show such deadlines exist; the one that applies to you depends on your country and your buyer’s contract, and it goes at the top of the sheet.
Write down the minute you open the sale document and the minute you reach the field. The elapsed time is a truer measure of readiness than the thickness of any binder: a trace that finishes in an hour and one that takes a week are different states of readiness, and only timing it lets you tell them apart.
What does one traced lot tell you, and what does it not?
One lot traced is a real test, not proof that every lot would trace as well. The Codex Commission warns that traceability “does not in itself improve food safety outcomes unless it is combined with appropriate measures”: the trace shows the path can be rebuilt, not that the product was good or the field well run.
Complete tracing is not the target either. The USDA economists write that “complete traceability is impossible”, since no system records every attribute of a food; you trace this lot to its field, not every input to its source. A trace that breaks is a gap to close, not a verdict on the farm.
The trace also helps in two other situations. If a buyer disputes a lot’s quality, with the path rebuilt you can show which grade it was sold under, provided you keep the buyer’s grade standard in writing: the list of what the buyer pays for and what they discount. Certification audits, the ones behind a seal or label on the product, often ask for this same trace. Before you pay for one, compare what the certifier requires with what your farm already records; what the seal asks you to document is a step-by-step guide to that comparison.
On the way back you will pass the pre-shipment quality check, the signed record of the lot’s condition when it left, and the trace is where you learn whether it was filed. Comparing what happened with what should have happened, and fixing the gap before the next sale, is a large part of running a farm well. There is more on selling what the farm produces in the marketing library.
Where to start
An afternoon, with the last settlements, the freight and scale tickets, the bin records and the field log on the table.
A binder of records is not the same as a trace you have already run. With the lot you pick today, followed with a clock running and the break circled, you have the answer before a buyer or an auditor asks for it. The papers were on the farm all along; what was missing was one afternoon following them to the field.