When a quality problem is found, does your team need to find every pack made on Tuesday afternoon, or this one pack? That question is at the heart of the choice between batch-level and unit-level identification. This comparison explains both, side by side, so you can match the identity to the job.
The identifier ladder
- GTIN“What product?”The product type — every bottle of the same 100 ml fragrance shares it.
- Batch / lot“Which production run?”Units made together, usually sharing a production date and expiry.
- Serial“Which exact unit?”Combined with the GTIN, one individual bottle.
- SSCC“Which logistic unit?”A specific carton or pallet as it moves through the supply chain.
- GLN“Which party or place?”A company, factory, warehouse, store or dock door.
For background on unit identity itself, start with the cornerstone what is product serialization?
Two levels of identity
Both levels start from the GTIN, which identifies the product type.
Batch-level identification adds a batch or lot number: a code shared by every unit made in the same production run. A thousand bottles filled from the same mix on the same shift share one batch. GS1 carries it in Application Identifier (10), up to 20 characters.
Unit-level identification adds a serial number, unique to one unit within its GTIN, carried in Application Identifier (21). This is serialization. A serialized code normally also carries the batch, so unit level includes everything batch level offers.
An analogy: a batch is like a school class — everyone in Year 9B shares it. A serial is like each student's roll number. You can announce something to the whole class, or call one student by name.
Side-by-side comparison
| Batch level (GTIN + batch) | Unit level (GTIN + serial, usually with batch) | |
|---|---|---|
| What it identifies | A production run of one product | One physical unit |
| Data on the pack | Same for every unit in the run | Different for every unit |
| Printing | Can be set once per run | Unique data sent to the coder for every unit |
| Line inspection | Check the code is present and readable | Also check each unit carries the expected, unique code |
| Records to keep | One record per batch | One record per unit, plus status changes |
| Recall precision | Whole batch | Specific units or ranges, if packing and shipping were recorded |
| Expiry and stock rotation | Yes, with expiry in the code | Yes |
| Batch-specific content (test results, origin) | Yes | Yes |
| Once-per-pack rewards or warranty | No — one code is shared by many packs | Yes |
| Duplicate-scan detection | Limited — many genuine scans per code are normal | Yes — one pack should behave like one pack |
| Relative cost | Lower | Higher: data, printing, inspection, storage |
The pattern is clear: anything that depends on counting or recognising one pack needs unit level. Anything that depends on what happened in production works at batch level.
Where batch level is enough
Batch-level identity, especially in a 2D code that also carries expiry, does more than many people expect:
- Recalls. If a contaminated ingredient is traced to two batches, you recall those batches rather than the product.
- Expiry management. Retailers that can read expiry from a 2D code can rotate stock and stop expired items being sold. See how retailers can use GTIN, batch, serial and expiry.
- Batch-specific consumer content. A resolver can route a batch-level GS1 Digital Link to a page showing that batch's test certificate or origin story, while falling back to the product-level page for anything not set at batch level.
For a carton of detergent or a jar of spices, this is often all the identity the product needs. We set out how to decide in do all products need serialization?
Where unit level is needed
Unit level becomes necessary when the business must treat one pack differently from its neighbours:
- Per-unit rewards and warranty. A loyalty programme for electricians must credit each box once. With a batch code, one photographed label could be claimed hundreds of times.
- Anti-counterfeit monitoring. A server can flag a unit-level code scanned in two distant cities within an hour. A batch code is legitimately scanned in many places at once, so that signal is lost.
- Precise recalls and investigations. If you recorded which units went into which cartons and shipments, a problem can be traced to specific units or customers.
- Regulated products, where a rule requires a unique identifier on each pack.
A worked example
A fictional skincare brand makes a 50 ml face cream. One morning, a supplier warns that a lot of preservative may be out of specification.
At batch level, the brand checks its production records: that preservative lot went into batches 2611A and 2611B. It recalls both batches — about 18,000 jars — and its product page, reached through the batch-level code on each jar, tells consumers which batches are affected.
At unit level, the brand also knows the preservative was only used for part of batch 2611B, and which serials were packed into which cartons and shipped to which distributors. It can narrow the recall to the affected units and contact the specific distributors. But to get that precision, it had to print, inspect and record a unique code on every jar, every day — including the many days when nothing went wrong.
Neither answer is wrong. The question is whether the extra precision is worth the daily cost for this product.
Choosing the level
Ask two questions:
- Will anyone act on the difference between two packs of the same batch? If not, stay at batch level.
- Can our lines print and check unique data at speed? If not, either upgrade the line or start with a pilot.
Remember that the choice is reversible upward. Moving to a 2D code with GTIN, batch and expiry today does not prevent you adding serials later — the same code structure simply gains one more field. See GTIN + serial: how an individual product gets its identity for how that extra field fits.
Create a batch-level or serialized GS1 Digital Link QR code