How to Design a Lot Number Format That Survives an Audit

A lot number is the one identifier that has to work everywhere: on a label at receiving, in a pick list at 6 a.m., in a recall spreadsheet at midnight. Most sellers invent their batch coding system in a hurry and only discover its flaws during an audit, when a code that looked fine on a carton turns out to be ambiguous, unsortable, or duplicated. This guide walks through what a lot code must do, the common schemes with real examples, and the pitfalls that quietly break traceability.

What a lot code actually has to do

Strip away the formatting debates and a lot number has four jobs. Judge every scheme against these before you print a single label.

• Unique — one code maps to exactly one batch of one product, forever. Reused codes turn a recall into an archaeology project.
• Sortable — codes that sort chronologically as plain text make FEFO checks, stock counts, and audits dramatically faster.
• Human-readable — a picker glancing at a shelf label should be able to compare two lots without a scanner. Eight characters is comfortable; twenty is not.
• Scannable and typeable — the code will be keyed into systems by hand at some point. Avoid characters that fail that test (more below).

Notice what is missing: the code does not need to encode everything. Supplier, product, expiry date, and receiving quantity belong in your batch record, linked to the lot — not crammed into the code itself. A lot number is a key, not a database row. If you need extra structured data per batch, attach it as custom batch fields rather than lengthening the code.

The four common schemes, compared

Almost every working lot number format is a variation on one of four schemes. Here they are with lot code examples and honest trade-offs.

For most e-commerce sellers receiving finished goods, date-based with a suffix (260915-A, 260915-B) is the right default: it is sortable, readable, and self-explanatory to a temp picker on day one. Julian dates earn their keep only when label space is genuinely scarce. If you manufacture or repack, a sequential scheme scoped per product avoids the awkward question of which date a multi-day production run should carry.

The ambiguity pitfalls that break codes

Codes fail in the field, not on the whiteboard. The classic failure is character ambiguity: O versus 0, I versus 1, S versus 5, B versus 8. A handwritten receiving log or a squinted-at thermal label turns LOT-1O0 and LOT-100 into two different keys pointing at the same physical batch, and your ledger silently forks. The fix is simple: ban the letters O, I, and ideally S and B from any alphanumeric portion of your codes, or go fully numeric with letter suffixes drawn from a safe set (A, C, D, E, F).

• Ban ambiguous characters: no O, I, and preferably no S or B in codes.
• Fix the length. Variable-length codes invite truncation and padding errors; 200915 and 0200915 should never both be plausible.
• Never reuse a code, even years later and even for a different product. Uniqueness is forever.
• Skip homemade checksums unless you print barcodes. A check digit helps machines but adds a character humans must copy correctly; a barcode with the code beneath it solves both problems at once.

Supplier lot versus internal lot: keep both

A recurring audit finding is a warehouse that overwrote the supplier’s lot code with its own at receiving. When the supplier later issues a recall for their lot 7734-K, you cannot answer the only question that matters: which of your internal lots came from it. The rule is keep both and link them. Record the supplier’s code exactly as printed on the carton, assign your own internal code in your format, and store the pair in the batch record. Trace-back (your lot to their lot) and trace-forward (their lot to your customers) both become a single lookup. If terms like trace-back and trace-forward are new, the batch and lot tracking glossary covers the vocabulary auditors use.

Where lot codes go to work: FEFO and recalls

A lot code earns its keep downstream. In daily operations, each lot carries its expiry date, and FEFO allocation uses that pairing to route the earliest-expiring lot onto each order automatically as orders sync from Veeqo or Shopify. In a crisis, the flow reverses: a supplier notice names a lot, and your records must produce every shipped order, customer, quantity, and date for it. That is precisely the export a one-click recall CSV generates — but only if the codes going in were unique and consistently captured. Once your format is settled, pressure-test it with a one-hour mock recall; a code scheme that survives that exercise will survive an auditor.

Frequently asked questions