The two questions a recall asks
Trace back starts from a problem and works towards its source: this batch of finished goods is defective, which raw material lot went into it, which supplier delivered that lot, on what date, and what else was made from the same delivery. It defines the scope of the problem.
Trace forward starts from a suspect input and works towards the market: this component lot is contaminated, which production runs consumed it, which finished units resulted, and which customers, shipments or locations received them. It defines who has to be contacted.
The two questions have to be answerable in either direction and usually in hours rather than days, because the window between discovering a problem and having to tell someone about it is short. A business that can answer one direction but not the other still ends up recalling far more than necessary.
The records that make the answers possible
At receipt, capture the supplier lot or batch number against the delivery, per line. This is the single most skipped step and the one that can never be recovered later, because a supplier lot printed on packaging that has since been discarded exists nowhere else.
At consumption, record which input lot went into which production run or output batch. In a manufacturing setting that is the link between a component lot and a work order; in a food or cosmetics setting it is the link between an ingredient lot and a batch code.
At despatch, record which output lot went to which customer, shipment or location. For direct customers this is straightforward. For anything sold through a distributor, one-up traceability means knowing the distributor and the quantity, which is what lets them run their own trace forward.
The through-line is that traceability is captured at three moments during normal work, not assembled during an emergency. Each one costs seconds when the goods are in front of someone and is impossible three weeks later.
- Receipt: supplier lot number, quantity, date, per purchase order line
- Consumption: input lot to work order or output batch
- Output: batch or serial assigned to finished units, with production date
- Despatch: output lot to customer, shipment or destination location
Batch numbers, serial numbers, and how much granularity to carry
Batch or lot tracking groups units made or received together under one code. It is the right level for anything produced in quantity from a shared input: food, chemicals, cosmetics, fasteners, coatings. The recall unit is the lot, so a smaller lot means a smaller recall.
Serial tracking identifies each unit individually. It is worth the effort for high-value or safety-critical equipment where you need to know exactly which unit went where, and it is overkill for anything you make ten thousand of.
Lot size is a decision with a recall cost attached. Running a whole day as one batch is less work than running each shift as its own; it also means a problem confined to one shift recalls three times as much product. That trade should be made deliberately rather than by default.
Expiry and best-before dates ride along with the lot. Where they apply, capturing them at receipt turns the same records into rotation control, which is a benefit you get on the ordinary days when nothing is being recalled.
Run a mock recall before you need one
Pick a lot at random, start a timer, and answer both questions from your records alone: what did this lot go into and where did it go. Do not let anyone use memory, and do not stop until you have a list of specific units and specific destinations with quantities.
The exercise fails in predictable places. A supplier lot that was never recorded at receipt, a production run that recorded which product was built but not which component lot went in, a despatch that names the customer but not the batch. Each failure is a specific record to start capturing on Monday.
Measure two things: how long it took, and what percentage of the quantity you could account for. Many regulated schemes expect the exercise inside a few hours with the great majority of quantity reconciled, and even outside regulation those are reasonable targets, because the gap between them and your result is the amount of extra product you would have to recall to be safe.
Repeat it periodically and after any process change. A traceability chain that worked last year does not automatically survive a new supplier, a new product or a change in how batches are coded.
Freezing stock and controlling the affected units
The first physical action is to stop affected stock moving. Quarantine on-hand units of the implicated lots so they cannot be picked, sold or consumed while the scope is still being established, and treat anything in production against those lots the same way.
Keep returned units separate from both good stock and general returns. Recalled goods that get mixed back into sellable stock recreate the original problem with better documentation, and a distinct quarantine location is the cheapest possible prevention.
Record every step with a timestamp and a name. Regulators and customers will ask what was done and when, and an activity log that shows who quarantined what at which time is worth considerably more than a reconstruction written afterwards.
In StockFlow, batch and lot tracking, serial number tracking, expiry dates and a full activity log run across receiving, production and despatch, so the trace-back and trace-forward lists come out of the records already being captured rather than out of an emergency spreadsheet.
Answer both recall questions from your own records
Capture batch and serial references at receipt, through production and out to despatch, with a full activity log behind every movement. Free to start.
Trusted by small businesses
What our customers say
“Super Kind! Quick replies from their support and very easy fixes, changed the dashboard a bit and customized it. Also gave me 450 items extra on the free plan just for me. Highly recommend and again great service!”
“Best customer service! Stockflow's customer support is fast and extremely helpful. They assisted me with customization of the software to improve my experience as a user.”