Lot and Batch Tracking

A contaminated ingredient enters your warehouse on a Tuesday. It gets received, put away, consumed in three production runs over the following week, and the finished goods ship to forty customers across two countries. On the Friday of the week after, a supplier notice arrives: that ingredient lot is subject to recall. The question you now have to answer is brutally simple and, without lot tracking, almost impossible. Which finished products contain that ingredient, and where did every one of them go? This scenario is exactly where lot and batch tracking earns its entire cost. It is one component of the wider automation picture I set out in the warehouse automation complete guide, and it is the component that keeps compliance and quality leaders awake at night.

The message up front: lot and batch tracking is not a feature you switch on, it is a discipline you enforce at every touch point where material changes hands or changes form. The value is not in capturing the lot, it is in never breaking the chain. A single unscanned receipt, a single blind issue to production, and your traceability has a hole exactly where the recall will need it.

1. What lot and batch tracking is

Lot and batch tracking is the practice of assigning a shared identifier to a group of items that were produced, received or processed together under the same conditions, then recording the movement of that identifier through every step of the supply chain. A lot (or batch, and the terms are used almost interchangeably in practice) is a quantity of material that shares a common origin: the same production run, the same supplier delivery, the same raw-material mix, the same day of manufacture. Everything in the lot is assumed to share the same quality characteristics, which is precisely why it is treated as a single traceable unit.

The distinction from serial tracking matters and I will return to it in detail, but the short version is this. A serial number identifies one individual physical item. A lot number identifies a group of identical items that cannot be meaningfully told apart. You do not care which specific bottle of syrup a customer received; you care which batch it came from, because if that batch is bad, every bottle in it is suspect. That group-level identity is the whole point. It gives you enough resolution to isolate a problem without the crushing overhead of tracking every individual unit.

Lot tracking answers three questions that no amount of general inventory accuracy can answer on its own. What went into this batch? Where did this batch go? And which other batches share a common cause with it? Standard inventory tells you how many units of a product you hold. Lot tracking tells you the pedigree of each of those units, which is a completely different and far more valuable kind of knowledge when something goes wrong. For the foundation this sits on, accurate stock positions in real time, see the real-time inventory tracking guide.

2. The traceability chain

The heart of lot tracking is an unbroken chain of custody. Every time material moves or transforms, the lot identity has to be carried forward or linked, so that a continuous thread runs from the supplier who delivered the raw material all the way to the customer who received the finished product. Break the thread at any link and traceability fails at exactly that point. The diagram below shows the chain as it should run, and the two directions in which you need to be able to follow it.

Read left to right, this is the physical flow of material. Read as a data structure, it is a genealogy. At receipt, the supplier lot is captured and bound to the received quantity. At put-away, that lot travels with the material to its storage location. At production, something important happens: the input lots are consumed and a brand new finished-goods lot is born, and the system records the link between the child lot and every parent lot that fed it. That parent-child link, the genealogy record, is the single most valuable piece of data in the entire chain, because it is what lets a recall jump across the transformation from raw material to finished product. From finished-goods storage, the new lot ships to customers, and the shipment record binds the lot to the destination.

3. Lot versus batch versus serial

People conflate these three tracking models constantly, and the confusion leads to systems that either under-track (and cannot support a recall) or over-track (and drown the operation in scanning overhead). The right model depends on what you are tracking and why. The table below lays out the practical differences.

The practical rule I give clients is to match the tracking model to the failure you are protecting against. If a defect would affect an entire production run identically, lot or batch tracking gives you all the resolution you need at a fraction of the cost of serialising. If you need to trace, warranty or recall an individual unit, for example a pacemaker or a car component, you need serial numbers, covered in depth in the serial number tracking guide. Many mature operations run both: serial numbers on the finished device for warranty and service, lot numbers on the components inside it for supplier recall. The two models are complementary, not competing.

4. Forward and backward traceability

A traceability system has to run in two directions, and a system that only runs one way is only half a system. The two directions answer different questions and get triggered by different events, but they rely on exactly the same underlying genealogy data.

Backward traceability (also called upstream or trace-back) starts from a finished product and works backward to discover everything that went into it. A customer reports illness after eating one of your products. You take the finished-goods lot number from the packaging and trace back: which production run made this, which raw-material lots fed that run, which suppliers delivered those lots, and on what dates. Backward tracing is how you find the root cause. It answers the question, what went into this?

Forward traceability (downstream or trace-forward) runs the other way. It starts from a raw material or component and discovers everywhere it ended up. A supplier tells you a specific ingredient lot is contaminated. You take that lot number and trace forward: which production runs consumed it, which finished-goods lots those runs created, and which customers received those finished lots. Forward tracing is how you scope and execute a recall. It answers the question, where did this go?

The honest limitation: traceability stops at the boundary of your records. If a supplier ships you material with no lot number, or a customer distributes onward without capturing yours, the chain is broken at that seam no matter how good your internal system is. Real traceability is an end-to-end supply-chain agreement, not just an internal warehouse feature. Push lot identifiers onto inbound purchase requirements and onto outbound shipping documents, or your carefully maintained internal chain dead-ends the moment goods cross your fence line.

5. Capturing lots at receipt, production and shipment

Every gap in traceability traces back to a missed capture point. There are three capture points that carry almost all of the weight, and getting the discipline right at these three is most of the battle.

At receipt. When inbound material arrives, the supplier lot number has to be captured and bound to the received quantity and its storage location. In a well-run operation the receiving clerk scans a GS1 barcode that already encodes the lot, so the capture is a single scan with no keying. Where suppliers do not barcode the lot, someone has to key it from the certificate of analysis or the packaging, and this manual step is where most receipt errors are born. Capturing the lot at receipt, together with any expiry date, is also what enables correct rotation later; the mechanics of that are in the expiry date management guide.

At production (the transformation point). This is the hardest and most important capture. When raw materials are issued to a production order, the system must record which input lots were consumed, in what quantities. When the order completes and finished goods are declared, the system assigns a new finished-goods lot and links it to every input lot that fed the run. This link is the genealogy. Skip it, or record it loosely, and the chain snaps precisely at the transformation, which is the one place a recall absolutely must be able to cross. Backflushing consumption without capturing actual issued lots is the single most common way operations quietly destroy their own traceability.

At shipment. When finished goods ship, the outbound lot number has to be bound to the customer and the shipment. This is what makes forward tracing to the customer possible. In practice the pick confirmation captures the lot as the picker scans stock, and the despatch note records it against the order. If shipment capture is weak, you can trace a problem all the way to your own loading dock and then lose it, which is worthless for a recall because the whole point is knowing which customers to notify.

6. Recall readiness and compliance

The real test of a lot-tracking system is not how neatly it captures data on a normal day. It is how fast and how completely it can answer a recall on a bad day. Regulators in most jurisdictions expect regulated food and pharmaceutical operators to trace one step forward and one step back, and increasingly to do so within hours rather than days. The one-up, one-down principle means you must be able to identify your immediate suppliers of any input and your immediate customers of any output, for any given lot.

Run the mock recall before the real one: the only reliable way to know your traceability works is to test it under time pressure. Pick a finished lot at random, and time how long it takes to produce the full backward genealogy and the full forward distribution list. If it takes days, or if there are gaps, you have found the weakness before a regulator or a sick customer does. Mature operators run mock recalls quarterly and treat the recovery time as a headline KPI. This discipline sits inside the broader automation and control picture covered in the warehouse automation complete guide.

A targeted recall and a catastrophic one differ almost entirely on the quality of lot data. With precise genealogy, you recall the exact finished lots that contain the suspect input and notify only the customers who received them. That might be three pallets and eight customers. Without it, you cannot prove which products are clean, so you have to recall everything that might contain the ingredient across the whole date range, which could be your entire output for a month and every customer you have. The contaminant is identical in both cases. The difference in cost, brand damage and regulatory exposure is enormous, and it is created entirely by whether the chain held.

Beyond recalls, lot data underpins quality holds and releases. When a batch is under investigation, you place a lot-level hold that prevents that specific material from being picked or shipped while everything else keeps moving. When the certificate of analysis clears, you release it. Without lot granularity you cannot hold selectively; you either freeze far too much or you let suspect stock ship. Lot-level status control is quietly one of the most valuable everyday uses of the whole system, long before any recall ever happens.

7. Lot and batch tracking in the WMS and ERP

Lot tracking is not a bolt-on. To work end to end it has to be native to both the warehouse management system that governs physical movement and the ERP that governs the financial and production record, and the two have to agree on the same lot identity. When they disagree, you get the worst of both worlds: a WMS that knows where material physically sits by lot, and an ERP that has lost track of which lot was consumed where.

In the WMS, the lot is an attribute of every stock record. Put-away, picking, cycle counting and stock moves all carry the lot forward, and the system can enforce rules such as first-expiry-first-out picking, which requires the lot and its expiry to be known at pick time. Directed picking that always pulls the oldest usable lot is only possible because the WMS holds lot-level detail on every location. In the ERP, the lot flows through purchasing, inventory valuation, production orders and sales, and it is the ERP that typically owns the genealogy record linking input lots to output lots across a production order.

In practice, mid-market operations often run lot tracking directly inside their ERP where it is built in. Microsoft Dynamics 365 Business Central, for example, handles lot and serial numbers natively through item tracking, carrying the lot from purchase receipt through warehouse activities to sales shipment, with the genealogy captured on production output. How that works in Business Central specifically, including item tracking codes and the trace tool, is covered in the Business Central inventory management guide. Larger or more automated operations layer a dedicated WMS on top for the physical execution and integrate lot data back to the ERP, and the integration boundary between those two systems is where I spend a great deal of my time, because a lot number that is captured cleanly in the WMS but arrives garbled or delayed in the ERP is a traceability gap waiting to happen.

The integration discipline is the same one that runs through every serious warehouse project: define the lot as a first-class data element, agree exactly which system is the source of truth for genealogy, and make sure every interface preserves the lot identity without transformation or loss. Get that right and the two systems present one coherent traceability story. Get it wrong and you have two half-truths that no recall can reconcile.

8. References

The most useful reference framework for anyone building lot and batch traceability is the GS1 system of standards for supply-chain traceability. GS1 defines the identifiers and barcode formats that let a lot or batch number travel between trading partners without re-keying, including the batch and lot number as a standard application identifier alongside the global trade item number and expiry date. Building your capture around GS1-encoded barcodes at receipt and shipment is what turns internal traceability into cross-company traceability, and it is the practical backbone of the one-up, one-down principle regulators expect.

Beyond GS1, the relevant reference points depend on your sector: food operators work to hazard-analysis and traceability requirements set by national food-safety authorities, and pharmaceutical operators work to good manufacturing and distribution practice with serialisation obligations layered on top of batch records. The common thread across all of them is the same principle this guide is built on: capture the lot at every touch point, preserve the genealogy across transformation, and be able to trace both directions on demand. The standards differ by industry; the discipline does not.

Final thoughts

Lot and batch tracking looks, on a normal operating day, like pure overhead. Extra scans, extra fields, extra discipline, for a payoff that never seems to arrive because most days nothing goes wrong. That is exactly why it is so often allowed to erode, one skipped capture at a time, until the day it is needed and the chain has a hole in it. The organisations that get this right treat the traceability chain as non-negotiable infrastructure, enforce lot capture at receipt, production and shipment without exception, and prove the whole thing with regular mock recalls that they time and improve.

When the supplier notice arrives on that Friday, the difference between the operators who built the chain and the ones who let it slip is not measured in convenience. It is measured in whether the recall is three pallets or three months of output, whether the regulator is satisfied in an afternoon or camped in your facility for a fortnight, and in some industries whether people are protected or harmed. Lot and batch tracking is the quiet discipline that decides which of those outcomes you get, and it is worth building properly long before you ever need it.

Related reading: Warehouse automation complete guide, Serial number tracking, Expiry date management, Business Central inventory management, Real-time inventory tracking.