SKUs, Barcodes and Product Data for a Growing Range
By The Velocity Wear Team
Apparel multiplies. One style in four colours and six sizes is twenty-four separate things to count, order and sell. The data structure that felt unnecessary at three products becomes essential at thirty, and retrofitting one is considerably harder than starting with it.
What a SKU should encode
A stock keeping unit identifies one specific sellable item — this style, this colour, this size. A code covering the style but not the size is not a SKU and will not work for stock control.
- 1**Style,** as a short code you will still understand in two years.
- 2**Colour,** abbreviated consistently. BLK, NVY, WHT.
- 3**Size,** in a consistent scheme.
- 4**Optionally a season or version,** if you reissue designs.
Something like HOOD-BLK-M. Readable by a person, sortable by a computer, and unambiguous when spoken over a phone — which matters more than it sounds when someone is checking stock.
Rules worth following
- **Fixed-length segments** where possible, so codes sort predictably.
- **No spaces and no special characters,** which break in spreadsheets and platforms.
- **Avoid characters that look alike** — O and 0, I and 1 — since these are read aloud and typed.
- **Never reuse a code.** A retired product’s SKU stays retired, or your historical data becomes untrustworthy.
- **Decide the scheme before you need it,** because renumbering a live catalogue is genuinely painful.
Barcodes
Your internal SKU is yours to design. Barcodes for retail are different — selling through shops or many marketplaces generally requires globally unique identifiers, which are issued rather than invented.
You need one per sellable variant, so every size and colour combination needs its own. Plan for that quantity, and get them before you need labels printed rather than after — barcode labels are frequently applied during production, which means the numbers must exist before the run.
The product data worth holding
Beyond identifiers, hold the data you will be asked for repeatedly: composition, fabric weight, garment measurements per size, care instructions, country of origin, and decoration specification.
That list is not arbitrary. It is what a product page needs, what a retailer asks for, what labelling requires, and what product-passport regulation is heading towards. Capturing it once per production run serves all four, and reconstructing it later is close to impossible once the fabric lot has gone.
Where to keep it
A spreadsheet is enough for a long time, provided it is one spreadsheet rather than several partial ones. The failure is not lacking a system; it is having data spread across a platform, an inbox and someone’s memory.
