If you want to understand traceability in a food production business, do not begin with a software demonstration. Begin with one batch.
A batch is wonderfully unforgiving. It does not care what the procedure says. It does not care what the system was supposed to do. It simply moves through the business, leaving evidence — or failing to leave evidence — as it goes. Follow it closely enough and it will show you where the real traceability system works, where it creaks, and where everyone has quietly learnt to step around it.
Map the journey, not the org chart
This is the essence of the “Map” stage. It is not a theoretical process map drawn in a meeting room. It is a factory-floor walk-through. Goods-in, storage, production, packing, labelling, finished goods, dispatch, quarantine, rework and returns all need to be looked at as part of one connected journey. Traceability fails most often in the joins between these areas, not inside the neat boxes on an organisational chart.
Choose a normal product. Not the simplest. Not the most troublesome. Pick something typical enough that the exercise reflects everyday work. Ask the team to show how one ingredient batch becomes part of one finished product batch, and how that finished batch reaches one or more customers. The aim is not to catch people out. It is to find the quiet gaps before an auditor, customer, enforcement officer or complaint does it for you.
Start at goods-in
Start at goods-in. A delivery arrives. Who checks it? What do they check against? Is the supplier approved? Is the product description the same as the purchase order? Is the supplier batch clear? Are dates readable? Is the temperature taken where relevant? What happens if the delivery note says one thing and the label says another? Does the person accepting the delivery have the authority to place stock on hold, or does stock drift into production while someone looks for a decision?
A small moment here can create a large problem later. A smudged batch code may be accepted because “we know the supplier”. A handwritten correction may never reach the production sheet. A delivery note may be filed by date, while stock is recorded by product code. None of this feels dramatic in the moment. Yet, six weeks later, when a supplier reports a problem with a lot number, these small compromises become the difference between a tight investigation and a broad, expensive guess.
Into storage, and “just for now”
Next, follow the material into storage. Where does it physically go? Is the location recorded? Are chillers, freezers, ambient stores and allergen-controlled areas clearly identified? Are part-used ingredients labelled in a way that preserves their original batch identity? Are decanted materials given a code that links back to the supplier batch? Does the team know what “on hold”, “released”, “rejected” and “awaiting inspection” actually mean?
Temporary storage is often where traceability becomes untidy. A pallet is put “just there” until someone has time to move it. A part-used bag is placed on a lower shelf because the usual space is full. A tray is held back for a quality check. A tub is relabelled quickly because production is waiting. “Just for now” is one of the most dangerous phrases in a busy factory. It sounds harmless because everyone can see the item at that moment. Traceability, however, is not about what everyone can see at that moment. It is about what can be proved later.
Weighing, units and top-ups
Then move into weighing or preparation. This stage is often a bridge between warehouse discipline and production urgency. Ingredients are selected, weighed, issued, sometimes substituted, sometimes split, sometimes returned. Ask how the batch identity follows the ingredient. Does the weighing sheet capture supplier batch numbers? Does it record quantities used and quantities remaining? What happens when more ingredient is needed halfway through a run? Can two supplier batches be used in one production batch? If so, is that clearly recorded?
This is also where unit confusion can begin. One record may use kilograms, another cases, another bags, another trays. A supplier may deliver in cases of six, production may use kilograms, and dispatch may think in pallets. That may be perfectly manageable if the conversion is controlled. It is rather less manageable if the conversion lives in someone’s head or on a battered note taped to a cupboard.
Production and the trouble with rework
Production itself brings another test: can the batch record tell a coherent story? It should show what was made, when, by whom, using which ingredients, under which controls, and with what checks. But the reality can be less tidy. Lines stop and restart. Product is reworked. A label roll is changed. A recipe version is updated. A quality check fails and is repeated. A small amount is held back. A packer spots a mistake and corrects it. These are not unusual events. They are food production. The question is whether the traceability record can cope with them.
Rework deserves particular attention. It is a perfectly legitimate part of many production processes, but it must not become a black hole. If rework is added into a new batch, can the original batch be traced? Are the quantities recorded? Are allergen, date-code and quality implications understood? Can the business show not only that rework was used, but exactly where it went? A system that traces fresh production but loses sight of rework is only partly working.
Labelling and finished goods
Labelling is another critical handover. The finished product may be safe and correctly made, but if the label is wrong, the traceability story becomes unreliable. Label checks should confirm product name, date code, batch code, allergen information where relevant, and pack format. The label is often the customer-facing part of traceability. It is also the link between the physical pack and the internal record. If label control is weak, digital records will not save the business from confusion.
Now follow the finished goods. Are pallets or cases identified with the correct product and batch? Are split batches recorded? If finished product is placed on hold, is that status visible to dispatch? If stock is released later, who records the change? If finished goods are moved to another location, does the record move with them? A physical move without a data move is one of the classic causes of stock uncertainty.
Dispatch is where traceability becomes forward-facing. Can the business show which customer received which batch, in what quantity and on what date? If a batch was split across several customers, can that split be reconstructed quickly? If a customer asks for evidence, does the answer come from one reliable source or a small office treasure hunt?
Finally, consider what comes back. Returns and complaints are often outside the neat forward flow of goods. Returned product may sit in a holding area. Samples may be retained. A customer may send a photograph of a code. A complaint may identify a date but not a batch. These events still form part of traceability. If the return or complaint process is disconnected from production and dispatch records, the business may miss useful evidence.
The value of following one batch is that it makes these issues visible without requiring a grand project. You do not need to digitise everything to learn from the exercise. You need a clipboard, a few honest conversations and the willingness to write down what actually happens.
Three kinds of gap
When the walk-through is complete, look for three types of gap.
The first is a record gap: information is needed, but no one captures it reliably. The second is a handover gap: information exists, but responsibility changes and the record does not follow. The third is a timing gap: the physical product has moved, but the record is updated later, sometimes much later.
Timing gaps are particularly important. In many factories, people trust the paperwork because it is complete by the end of the day. But traceability under pressure often depends on what was known at the time. If records are always written up after production, the business may be relying on memory more than it realises. That does not mean every small business must scan every movement instantly. It does mean the owner-manager should understand where the delay creates risk.
At the end of the walk-through, resist the urge to turn every finding into a software requirement. Some problems are simpler than that. A location needs a name. A form needs one extra field. A handover needs a sign-off. A status label needs to be clearer. A spreadsheet needs to be retired. A production sheet needs to capture the second ingredient batch when a top-up is used.
Choose the software last
The best digital solutions are chosen after this work, not before it. If you cannot describe the batch journey, you cannot sensibly digitise it. Software suppliers may show dashboards, alerts, barcode scanning and reports. These may be useful later. But the real question is whether the system fits the journey your product actually takes.
For an SME food producer, following one batch may be the most valuable traceability exercise you can do this month. It is simple, practical and revealing. It respects the reality of the factory floor. It also creates the kind of operational insight that makes future digital choices far more sensible.
Traceability is not proven by saying “the process is documented”. It is proven by following the batch and finding the evidence.
Manager’s pain point
“The process looks controlled in the procedure, but the real journey through the factory is messier.”
AI prompt to try
Create a batch traceability walk-through checklist for a small UK food production business. Include goods-in, storage, weighing, production, packing, labelling, finished goods, dispatch, quarantine, rework and returns. For each stage, ask what data is captured, who captures it, where it is stored, what could go wrong, and what evidence would be needed during an audit, complaint, withdrawal or recall.Why this helps
This prompt gives the manager a practical factory-floor checklist. It moves traceability from theory to observation, making it easier to find weak handovers, timing delays and places where stock becomes invisible.
