How to Barcode and Label Handbags for Amazon and Shopify

How should a handbag be barcoded and labelled for marketplace and Shopify channels? Follow the decisions in this order: channel first, because it decides the symbology; then identifier, EAN-13 or UPC-A for retail and Shopify and FNSKU for marketplace fulfilment; then placement, a flat panel at least 6 mm from any seam and never on leather; then print quality, 300 dpi thermal transfer with a 6.35 mm quiet zone; then carton labels at 4 x 6 inches. Most scan failures are placement and quiet-zone failures, not symbology failures: the code is printed correctly and then laid across a curve or crowded by a stitch line. Nine decisions below, each with the number to specify and the check to run before cartons close.

Labelling numbers to specify

The First Branch: Retail, Marketplace or Own Store

Channel decides everything downstream, so it is the first branch in the tree and the one to settle before artwork starts. Three channels have three different answers to the question what identifier goes on this bag, and a specification written for one will be rejected by another.

• Retail, through a buyer with a scanner. A GS1-issued GTIN, encoded as EAN-13 for most markets outside North America and UPC-A within it. The number must be licensed to your brand, not borrowed.
• Marketplace fulfilment. An FNSKU identifier generated by the channel for each SKU, applied as a label so that units received at the centre are attributed to your inventory.
• Own store, direct to consumer. A barcode is optional. What matters is that the SKU in the store platform matches the SKU on the carton and the packing slip.

The common error is assuming one code serves all three. A UPC-A printed on a hangtag will not be accepted as an FNSKU at a fulfilment centre, and an FNSKU will not scan at a retail till. Where a brand sells through more than one channel, the practical answer is a plain retail GTIN on the hangtag for retail and Shopify, plus a fulfilment label applied to the poly sleeve for the marketplace portion of the inventory.

Which Symbology You Need: EAN-13, UPC-A or FNSKU

Three symbologies cover almost every handbag programme. EAN-13 encodes 13 digits and is the default outside North America; UPC-A encodes 12 and is the North American retail default; FNSKU is a marketplace-specific identifier, usually beginning with an X, that ties a unit to a seller's inventory within one fulfilment network.

Two rules follow. First, a GTIN must be licensed through a GS1 member organisation rather than purchased cheaply from a reseller, because a retail buyer's system checks the prefix against the GS1 registry and a mismatch is discovered at goods-in, not at quotation. Second, one GTIN covers one product variant: a tote in tan and the same tote in black are two GTINs, and so are two sizes of the same style.

Where the Label Goes on a Handbag

Placement is the decision that causes most scan failures, and it is entirely within the brand's control. A barcode needs a flat, unstitched, uncurved surface with a clear margin, and a handbag is mostly curves, seams and hardware. That is why the label usually goes on the packaging rather than on the product.

1. First choice: the poly sleeve. A flat film surface, no curvature, no adhesive on leather, and the label is discarded with the film.
2. Second choice: the hangtag. Flat 350 gsm card, and the code can be printed rather than labelled, which removes a component.
3. Third choice: a care card or insert. Works where the bag ships unbagged, but the card must be in a fixed position.
4. Never: directly on leather or coated fabric. Adhesive leaves residue, can lift a finish, and a curved panel fails to scan.

Within whichever surface is chosen, three spacing rules hold: at least 6 mm from any seam or edge, no curvature across the symbol, and the full 6.35 mm quiet zone left and right kept clear of text and print. A code that meets all three scans first time; one that fails any of them scans intermittently, which is worse than never, because intermittent failures are discovered at goods-in rather than at sampling.

Print Quality: The Numbers a Scanner Cares About

Print quality is graded, not judged. ISO/IEC 15416 defines how linear barcodes are measured — symbol contrast, modulation, edge determination, defects and decodability — and produces a letter grade from A down to F. Specify grade C or better and verify it with a barcode verifier rather than by eye, because a code can look perfect at 203 dpi and still fail intermittently.

• Resolution. 300 dpi thermal transfer on a coated label stock. Direct thermal prints fade under heat and warehouse lighting.
• Quiet zone. 6.35 mm clear on both sides of the symbol; this is the specification most often lost in artwork.
• Contrast. Dark bars on a light background. Metallic foil behind a barcode destroys contrast and should be avoided.
• Size. Do not scale the symbol below 80 percent of nominal without re-verifying; shrinking narrows bars past tolerance.

Thermal transfer on coated stock costs about USD 0.02 per label and is the cheapest reliable option. Printing the code directly onto the hangtag costs nothing extra at print and removes a component from the pack, which is the better answer where the channel allows it — but it means the code must be verified on the printed card rather than on a separate label.

Carton Labels and Shipment Identification

Carton labelling is a separate specification from unit labelling and is where inbound shipments most often get delayed. Each carton in a shipment to a fulfilment centre carries a shipment identifier label, typically 4 x 6 inches or 102 x 152 mm, printed at 300 dpi and applied to a flat panel of the carton rather than across a corner or over the tape seam.

• One label per carton, on a flat side panel, never wrapping an edge.
• Two printed side panels with SKU, colourway, quantity, carton number and country of origin.
• Carton numbering as n of m, so a shortage is identifiable without opening everything.
• Dimensions and weight on the label, matching the packing list within a small tolerance.

Marketplace prep also imposes a rule on film: where a poly bag is used to prep a unit, a suffocation warning in the destination language is required. That rule is worth confirming at brief stage rather than at packing, because a non-compliant prep is chargeable at the destination and the fix — re-bagging 500 units — costs more than the entire label budget. The marketplace packaging rules cover the current requirements.

Country of Origin and Content Labelling

Origin and content marks are legal rather than logistical, and they belong on the product and the carton rather than only on an insert. Getting them wrong is the version of a labelling error that stops a shipment at customs rather than at a scanner.

• Country of origin. On the bag, on the hangtag or on the carton depending on the destination's rules, and always matching the customs documentation.
• Fibre and material content. Where the destination requires a composition statement, it should name the material of the body, the lining and the trim.
• Care information. A care card with cleaning language approved for the destination market, not translated from a template.
• Children's items. CPSIA documentation and tracking labels where a bag is marketed to children.

Two documents support the rest of the file: an ISO 9001 quality system reference, and BSCI or Sedex / SMETA audit references for the validated partner factory. Both can be supplied for our validated partner factory Quanzhou Yingming Bags. Where hides are sourced through Leather Working Group tanneries, that traceability is documented the same way, and it belongs on the material story card rather than on the label.

Shopify Versus Marketplace: Two Different Data Models

The two channels want different things from the same number, and understanding the difference prevents a whole category of inventory errors. Shopify treats the barcode as an attribute of a variant — a lookup key for your own stock system — while a fulfilment network treats its identifier as a claim on a specific unit of inventory belonging to a specific seller.

The practical consequence is that the variant boundary must be defined once and used everywhere. A tote in two colours and two sizes is four variants, four sets of labels and four carton line entries. Defining that boundary in the brief — before label artwork exists — is the cheapest way to avoid a relabelling exercise at the destination.

Verifying Before the Cartons Close

Verification is cheap and it belongs in the pre-shipment inspection rather than in a separate process. Under AQL 2.5 and ISO 2859-1, the drawn units are opened and checked, and the label check is four items: presence, position, print quality and scan.

1. Presence. Every drawn unit has the correct label for its variant — the most common failure is a mixed-label carton.
2. Position. At least 6 mm from any seam, flat, with the 6.35 mm quiet zone unobstructed.
3. Print quality. A verifier reading to ISO/IEC 15416, grade C or better, on a sample rather than by eye.
4. Scan. A handheld reader on a sample of labels, recorded in the inspection report with the grade.

Record the results in the report, because a scan failure discovered at the destination is charged back and a scan failure recorded at origin is negotiable before the goods leave. Where third-party inspection by SGS is booked, the same four checks are included and the report is issued to you.

What to Send With the Brief

Labelling is specified once and then copied for every reorder, which makes the brief the highest-leverage document in the whole exercise. Seven items, on one sheet, alongside the bag specification.

• Channel and destination, which decides symbology and prep rules.
• Variant list with the boundary defined: style, colourway, size, and the identifier for each.
• GTIN or FNSKU numbers per variant, with the GS1 registration confirmation where retail is involved.
• Label artwork at 300 dpi with the 6.35 mm quiet zone drawn in, not implied.
• Placement stated in millimetres from a datum edge, with the surface named.
• Carton label size and content, 4 x 6 inches, with carton numbering as n of m.
• Origin and content marks approved for the destination market.

Send that sheet with the first brief and labelling never becomes a launch-week problem. Sampling runs at 6-10 working days, bulk at 35-50 days, and inspection at AQL 2.5 under ISO 2859-1 covers labels alongside the bag and the rest of the pack. The twelve-point packaging checklist places labels in the full pack, and the private label service overview shows where they sit in a brief.

One more branch worth settling early, because it changes the pack rather than the label: whether labels are applied at origin or at the destination. Applied during packing, a thermal transfer label costs about USD 0.02 and is inspected with everything else at AQL 2.5. Applied by a prep service at the destination, the same label can cost several times that per unit and sits outside any inspection you control. Where the channel permits origin labelling, take it — and put the label artwork on the same sheet as the bag specification so nothing is left to a packing-week decision.

Will It Still Scan After Ten Weeks in a Container

Labels are specified for the shop floor and judged at the destination, forty days later, after a humid crossing and a week in a warehouse. Most failures at goods-in are not printing failures; they are failures of stock and adhesive under conditions nobody specified. Four decisions at drawing stage remove nearly all of them, at a cost measured in cents.

• Face stock. Seventy to eighty gram paper is fine on a hangtag but weak on a poly sleeve: fifty to seventy micron polypropylene film resists scuffing and keeps the symbol flat.
• Adhesive. Permanent acrylic holds across roughly minus 20 to 80 degrees Celsius; hot-melt softens above about 40-50 degrees, and the inside of a sun-loaded container easily reaches sixty.
• Application conditions. Labels applied below 10 degrees or above 40 degrees lose adhesion immediately, which is why applied-at-origin labels need a stated conditions line in the work instruction.
• Rub resistance. The printed symbol should survive ten cycles of a cotton cloth rub without losing its grade; a barely adequate direct thermal print fails this within weeks.
• Why not on the leather. Plasticiser migration lifts an adhesive from a coated surface over eight to sixteen weeks, which is the mechanical reason the code belongs on the sleeve or the hangtag.

How Verification Is Actually Run

A hand scanner gives a yes or a no; a verifier gives a grade per parameter and says which one failed and by how much. The distinction matters because a symbol that scans today can be one grade away from failing after a week under warehouse lighting. Verification is also a routine that has to be scheduled, not an event, and it sits naturally inside the pre-shipment checks described in the AQL 2.5 inspection guide.

• Eight to ten scan lines are averaged across the symbol height; a single scanner beep tells you nothing about consistency.
• Parameters graded A to F for symbol contrast, modulation, edge determination, defects and decodability, with the symbol taking the lowest grade of the five.
• Grade C or better is the usual commercial requirement, measured on a calibrated verifier rather than on a handheld reader.
• Verify after finishing. A code printed on a hangtag must be verified after varnishing or laminating, because a gloss layer can drop decodability by a full grade.
• Sampling through the run. Verify the first-off label and then one label in every 500, keeping the reel identity and print date with the verification record.

What a Scan Failure Costs at Goods-In

Labelling errors are discovered at the receiving door, which is the worst possible place in the whole chain: the goods are already paid for, the clock is running, and the fix is done at local labour rates. The numbers below are typical rather than universal, and they explain why the packaging rules for marketplace fulfilment get quoted at brief stage rather than at packing.

• Per-unit preparation charges. Relabelling is commonly billed at USD 0.20-0.60 per unit, and a full prep where the unit must be bagged or tagged again at USD 1.50-3.00.
• Receiving hold. A lot that cannot be received adds three to ten days to inbound appointment time, and considerably longer during an autumn peak.
• Retail chargebacks. Buyers levy these per purchase order line, often USD 25-75 plus handling, regardless of how many units were affected.
• Quarantine rather than sorting. Where a symbol fails intermittently, the safest warehouses quarantine the whole receipt rather than triage it, so ten bad labels can hold five hundred good units.
• The cheap alternative. Verification plus the four-item presence, position, quality and scan check at origin costs about USD 0.01-0.03 per unit.

Frequently Asked Questions

People Also Ask