Batch traceability means answering two questions in minutes: where did the goods you shipped come from, and where did the goods you received end up. That is not the same as knowing how much stock you hold. An accurate inventory tells you the quantity; traceability tells you the identity of each unit and its history. In food, pharma and chemicals the difference is regulatory: with no batch on record, no product recall is possible. Here is what traceability requires, which data to capture at each step, and which part is solved by process discipline before buying any software.
What batch traceability means, and how it differs from counting stock
A batch is a set of units produced or received under the same conditions: same raw material, same line, same shift, same supplier. Tracing by batch means recording, at every movement, which batch the goods belong to. Counting stock answers how many units of a given SKU (stock keeping unit, each distinct item you handle) you have. Tracing answers which of those units is the one in front of you and where it came from. You can have a perfect inventory and broken traceability: if you know how many cases are on hand but not which ones belong to the affected batch, you have to handle all of them. The gap shows up the day something goes wrong, never before.
Which industries treat it as a hard requirement
In food, pharma and chemicals, batch traceability is not a best practice: it is a condition for operating. Food requires identifying the origin of every ingredient and the destination of every shipped lot. Pharma often pushes the requirement down to the individual pack, with expiry date and serial number. Chemicals add the safety data sheet and storage compatibility between products. Beyond those three sectors, others adopt it by contract: auto parts, cosmetics, medical supplies and anything destined for export, because the overseas buyer writes it into the tender. If you sell to a large supermarket chain or a pharmacy chain, the requirement will most likely reach you through them before it reaches you through a regulator.
Backward and forward traceability are two different questions
They rely on different records. Backward: given a finished product, which raw material batch it came from, which supplier delivered it, and on which inbound document it entered. You use it when the problem shows up in your product and you need the cause. Forward: given a suspect batch, which customers received it, on which shipping documents, and on which dates. You use it when the problem is already known and the goods must come off the market. Most operations have solved only one of the two, usually the backward one, because receiving is recorded carefully and dispatch is done in a hurry. On recall day, the half you need is the forward one, and it cannot be improvised.
Which data to capture at each point of the journey
At receiving: batch, expiry date, supplier, inbound document, quantity and date. It is the only point where the data arrives from outside; if you miss it here, it does not exist later. At putaway: the exact location each batch went to, plus a clear rule against mixing different batches in the same position. At picking: which batch was taken for each order line, not only which item. This is the link that breaks most often, because the operator picks whatever is closest. At dispatch: which batches left, on which shipping document, and to which customer. And at every point, who did it and at what time. Without operator and timestamp you cannot reconstruct anything, nor fix the step that failed.
What happens in a recall if you have no batch traceability?
The recall stops being surgical and becomes total. If you cannot prove which customers received the affected batch, you have to assume all of them did: you pull the whole SKU, from every customer, across whatever period you can justify. The cost is not only the goods. It is the team hours spent going through paper documents, the operation halted while the search runs, the customer relationship damaged by a generic notice, and the exposure before the regulator, who will ask for the record and will note how long you took to produce it. An operation that traces well narrows the recall to the affected documents and answers with a list. One that does not answers with an estimate, and an estimate does not close a case file.
What process solves, and what genuinely needs software
Inventory accuracy is measured by comparing records with physical stock. Results depend on receiving and dispatch records, location rules, training and controls; software or robots do not guarantee an accuracy percentage.
When you do not need to automate traceability
If you handle few SKUs, batches are large, they come in and go out whole, and you ship to a handful of customers, a well kept spreadsheet and tidy paperwork are enough. Automating there adds fixed cost and a dependency without solving a real problem. It is also worth pausing if the process is not defined yet: computerising a record nobody follows only produces bad data faster. First tidy the loop, run it by hand for two or three months, and measure whether the data comes out complete. If it does, and volume is growing, then the system pays back. The reverse order, buying first and organising later, is the most expensive way to discover the problem was in the process.
Where to start
Start with the forward question: if you had to pull a batch off the market tomorrow, could you say which customers received it and on which documents, today, without digging through paper? If the answer is no, the job is not buying a system. It is closing the four capture points, receiving, putaway, picking and dispatch, and sustaining them until the data comes out complete on its own. When volume justifies automating, the decision is made with the operation already in order, and the system is sized on real data rather than assumptions.
Frequently asked questions
A batch groups units produced or received under the same conditions and is tracked in blocks. A serial number identifies a single unit and lets you follow it on its own. Batch level is enough for most food and chemical operations; pharma and high value equipment usually require serials. The finer the level, the more capture points and the more recording discipline the operation needs.
Yes. A WMS, or warehouse management system, is the software that administers locations and movements, and it makes the job far easier, but traceability rests on complete records, not on a software brand. At low volume, spreadsheets and tidy paperwork do the job. The system becomes necessary when the volume of movements makes manual recording more expensive or less reliable than automatic capture.
You lose traceability on both. From the moment you cannot say which unit belongs to which batch, any recall has to cover both batches in full. That is why the one position, one batch rule is not a system quirk: it is what keeps the scope of a problem contained. If space forces you to mix, fix the space before you look at software.
Yes, and it becomes more critical there because expiry dates drive everything. In cold rooms every door opening costs energy and time, so the system should tell you exactly where the batch sits before anyone walks in. The automated equipment STOKA integrates works down to −30 °C for stacker cranes and −25 °C for shuttles, so the record can live inside the cold room instead of depending on a clipboard.
The manual loop can be organised in weeks, but the data only comes out complete once the team does it without reminders, and that takes a couple of months of practice. If you also automate, installing a STOKA system takes 3 to 4 months from the purchase order, and that lead time runs in parallel with tidying the process, which is best resolved before go live.