What Is Product Traceability?

When a supplier reports a contaminated ingredient, a retailer finds expired stock on a shelf, or a distributor disputes what was in a delivery, the same two questions come up: where did this come from, and where else did it go? Product traceability is the discipline of being able to answer them quickly and with evidence, rather than with guesswork and a warehouse search.

What product traceability means

Traceability is the ability to follow identified goods backwards and forwards through the supply chain: which inputs and production run they came from, how they were packed, who received them and where they are now.

It helps to think about it at three widths.

• One step back. Each business knows who it received a product or ingredient from, and when.
• One step forward. Each business knows who it sent the product to, and when.
• End to end. The links from every business join up, so a question about one jar or one batch can be followed from raw material to the store shelf.

One step back and one step forward is the minimum any business can control on its own. End-to-end traceability needs several companies to keep compatible records and be willing to share them. That second part is usually the harder one.

Tracking vs tracing

The two words are often used together, but they point in opposite directions.

Tracking follows goods forward: where did batch TL-0412 go, and which distributors and stores hold it now? Tracing follows goods backward: this jar was returned by a customer, so which batch, line, shift and ingredient lots did it come from?

A recall usually needs both. You trace back from a complaint to find the cause, then track forward from the cause to find everything else it touched.

How detailed does traceability need to be?

Traceability always works on identified things, but the "thing" can be more or less specific. The ladder at the top of this page shows the identifiers involved; the table below shows what each level of detail lets you answer.

The levels are not alternatives in a strict sense. A detergent maker might trace by batch, identify each pallet with an SSCC, and never serialize a single bottle. A perfume house fighting grey-market resale might do all four. More detail costs more on the packaging line and in data handling, so the right level is the coarsest one that still answers your real questions. The trade-offs are covered in batch-level vs unit-level identification, and the unit-level option in what is product serialization?

The building blocks

Every working traceability system, from a spreadsheet at a small food business to a multi-company event network, is built from the same five pieces.

1. Identifiers

You cannot trace what you cannot name consistently. GS1 provides the common identifiers most trading partners already use:

• GTIN, with a batch or serialization where needed, for the trade item itself.
• SSCC, an 18-digit number for a carton, pallet or parcel in transit. See what is an SSCC?
• GLN (Global Location Number), a 13-digit number for a company, site or specific location such as a dock door or packing line. See what is a GLN and where is it used?

These travel in barcodes or 2D codes using Application Identifiers, such as (01) for the GTIN, (10) for batch, (21) for serial and (00) for the SSCC. The barcode is only the carrier; the identifier is what your records use.

2. Data capture at every hand-off

Identifiers only help if someone records them when goods change state or change hands: produced, packed, shipped, received, sold, returned, destroyed. The moment matters. A record made at the dock door when the pallet leaves is evidence. A record reconstructed from a production report two days later is an estimate.

On the factory side, that capture starts on the packaging line, where codes are printed and read back. The journey of a serialized product through a packaging line shows where each piece of evidence comes from.

3. Aggregation

Warehouses do not scan every jar. They scan the pallet. For that to work, someone must have recorded which units went into which carton and which cartons went onto which pallet. That parent–child record is aggregation, and it is what lets one scan stand for hundreds of items.

The physical process is described in how individual products become cartons and pallets, the record itself in what is aggregation in product traceability?, and the warehouse side in why warehouses scan a carton or pallet instead of every unit.

4. Shared event records

Each step above produces an event: something happened to identified goods, at a time, at a place, for a business reason. EPCIS (Electronic Product Code Information Services) is GS1's standard for recording and sharing those events in a common shape, covering the what, when, where, why and, since EPCIS 2.0, how. Its companion vocabulary, CBV (Core Business Vocabulary), supplies shared words for business steps such as commissioning, packing, shipping and receiving, and for states such as in_transit or recalled.

Read in order, a short run of events like the ones above tells the story of a product across companies: commissioned at the factory, packed, shipped, received at a distributor. Each company records its own steps. The full picture is explained in what is EPCIS? and walked through for one bottle in EPCIS explained through the journey of one product.

5. Access and sharing between partners

Traceability data is not held in one central place. GS1 does not run a database of everyone's events; each business captures its own records and shares them with trading partners by agreement (does EPCIS mean sending all my data to GS1? covers this). So every programme has to decide who may see what, how requests are made, and how quickly a partner must respond during an incident. Without that agreement, end-to-end traceability exists only on paper.

Why businesses invest in traceability

The reasons are usually concrete, and most organisations have more than one.

• Recalls. The difference between withdrawing "everything with this product name" and "these three batches from these eleven stores" is traceability. It limits cost, waste and the reputational damage of a broad recall.
• Expiry and stock rotation. Knowing which batches sit where, with which expiry dates, helps distributors and retailers sell or pull stock in the right order.
• Compliance. Some sectors and markets set their own identification or traceability rules, for example for certain medicines. These vary by country and product category, so start from the rule that actually applies to your products rather than a general checklist.
• Diversion and grey markets. If goods intended for one market or channel appear in another, shipment and aggregation records show where they left the authorised route.
• Sustainability and origin claims. Statements about where materials came from, or how a product was made, are only as credible as the records behind them.
• Customer and partner questions. Retail buyers, distributors and consumers often ask where a product came from. A traceable answer is faster and more defensible than an email chain.

Where traceability usually breaks

Most traceability problems are not about missing software. They are about missing or untrustworthy records at a specific link in the chain.

No aggregation. Units are serialized and pallets are labelled, but nobody recorded which units went into which carton. The pallet can be tracked; its contents cannot. Rework that bypasses the system causes the same gap later.

No partner data. The brand's own records are excellent up to the factory gate. After that, the distributor's system uses its own product codes, or keeps only paper delivery notes. One step forward is known; the second step is not.

Paper and spreadsheets. Handwritten batch logs and emailed spreadsheets can be accurate, but they cannot be searched in minutes during an incident, and they rarely use the same identifiers as the next company's records.

Consumer scans mistaken for supply-chain events. A shopper opening a product page from a QR code tells you about engagement, and sometimes hints at where a product ended up. It is not evidence that the product was shipped, received or sold through a particular channel. Scan data can raise useful questions, as how consumer scans can reveal channel and diversion signals explains, but it does not replace shipping and receiving records.

How to start

The most common mistake is to start from technology: "we need serialization" or "we need an EPCIS system". Start instead from the question you need to answer, and let it set the scope.

1. Write down the question. For example: "If an ingredient lot is contaminated, which customers received finished goods made with it, within four hours?" Or: "Which distributor sold this unit outside its territory?"
2. Choose the granularity that answers it. The first question needs batch and shipment records. The second needs unit serials and aggregation down to the unit. If the question does not need unit-level data, you may not need serialization at all.
3. Map the hand-offs. List every point where goods change state or change hands, and who records what at each one today. Gaps usually become obvious here.
4. Fix identifiers before systems. Make sure products, batches, logistic units and locations are identified consistently, ideally with GS1 identifiers your partners already understand.
5. Agree sharing with the partners who matter. Start with the two or three distributors or suppliers who matter most, rather than the whole network.
6. Test it. Run a mock recall or a mock diversion investigation and measure how long the answer takes and where it fails.

Take Tamarind Lane, a fictional chilli-sauce maker. Its first question is recall scope by batch, so it starts with batch codes on every jar, SSCCs on every pallet and receiving confirmations from its four distributors. Unit-level serials can come later, if diversion ever becomes a real problem.

Traceability is less about a single system than about a habit: identify things consistently, record what happens to them when it happens, and agree in advance who can see those records. Get those three right at the level of detail your questions need, and the technology choices become much simpler.