8 Things to Check on a Premium Handbag Production Floor: A Buyer's Walkthrough in Eight Questions

One: Is the Material Lot Named and Held?

The first check is documentary rather than physical, and it is the one that decides everything downstream. Ask which lot the leather came from and whether one lot covers the whole colourway, because a run drawn from two dye lots will show shade drift that no finishing technique can hide. A named lot should be traceable to a tannery certificate, and where the tannery is audited under the Leather Working Group protocol the certificate should name the tannery itself rather than an agent.

• Ask for the lot reference and the tannery name, then check that both appear on the pack.
• Confirm the lot covers the full colourway quantity, since topping up mid-run is how drift begins.
• Check the arrival record: a newly named European lot takes 15-20 working days to reach the bench.
• Ask whether compliance reports for the destination market were booked against that specific lot.

Two: Is the Cutting Marker Built for Yield or for Speed?

The second check is the marker, and it reveals more about a house than any showroom. A marker built for yield places panels around the natural shape of the hide and reaches roughly 78-85% utilisation; a marker built for speed runs panels in straight bands and quietly wastes material. Yield matters commercially because it sets the hide requirement, and it matters aesthetically because panels taken from the belly of a hide behave differently from panels taken from the shoulder.

The reason this check sits second is that it is the last point at which a material problem is still inexpensive to put right. Once panels are cut, every subsequent operation adds value to a piece that may already be wrong, which is why a floor that cuts first and checks later produces expensive corrections.

Three: Are Folds Skived, or Simply Folded?

The third check is tactile and takes ten seconds. Run a thumb along a fold: a premium fold tapers to 0.6-0.8 mm over a skive width of 8-10 mm, while a fold left at full substance steps and stands proud at the corner. Skiving is not cosmetic, because a double-thickness corner resists closing, wears at the edge finish from the inside, and eventually telegraphs through the surface as a ridge.

• Skive width of 8-10 mm gives a taper the hand reads as gradual rather than as a step.
• Fold thickness of 0.6-0.8 mm means two panels meeting at a corner make one thickness.
• Adhesive open time of 8-12 minutes before closing, so the bond has taken rather than slipped.
• Press dwell of 15-20 seconds at the fold, which is a gate rather than a habit.

Two conditions on the floor govern whether a skive holds: room humidity and adhesive discipline. A bench kept at roughly 50-60% relative humidity keeps leather at a stable moisture content, so a panel skived in the morning behaves the same way when it is closed in the afternoon. Adhesive applied outside its 8-12 minute open time either skins over or stays wet, and both produce a fold that lifts months later rather than immediately.

Four: What Is the Stitch Count on a Structural Seam?

The fourth check is arithmetic: count the stitches in one inch of a structural seam. Premium work runs 8-9 per inch with a bonded thread around 0.45 mm in diameter and a needle sized to match, commonly a 90/14 on calfskin. Standard work runs 6-7 per inch with a thicker thread that fills the needle hole instead of closing the seam, and the difference appears when the bag is overfilled rather than on the shelf.

Structural seam

8-9 stitches per inch, tolerance ±0.5 per inch

Handles and straps

9-10 stitches per inch, since they carry the load of the whole bag

Thread

Bonded filament around 0.45 mm in diameter, matched in shade to the lot

Needle

90/14 on calfskin at 1.2-1.4 mm, changed on a set schedule rather than when it breaks

Backstitch

Present at the start and end of every run, and inspected specifically

Needle discipline is worth watching while you are there. A needle that is changed on a schedule produces clean holes; one that is run until it breaks produces torn fibre at the seam line, which is invisible on day one and visible as a frayed run within a season of being carried.

Five: How Many Coats Build the Edge?

The fifth check separates a finish that lasts from one that photographs well. Ask how many coats build the edge and how long each cures: three coats of roughly 0.12 mm, sanded between passes, is the premium answer. A single heavy pass looks acceptable under showroom light and then hairlines at the corner, because a thick film cannot follow the leather flexing beneath it.

• Three coats of about 0.12 mm each, sanded and re-checked between passes.
• A curing interval between coats that is treated as a gate rather than as a preference.
• Corners radiused so the film has no sharp turn to bridge.
• Edge shade matched to the lot rather than to a standard swatch card.

Six: Where Does Hardware Wait Before It Is Fitted?

The sixth check is about staging, and it catches the substitution that causes most late disputes. Components should arrive with a record, wait in a identified place, and be compared against a signed reference from the approved sample before fitting. Plating of 0.15-0.3 µm, proved at 24 hours of neutral salt spray under ISO 9227, is a measurable target; a finish name is not.

Seven: Was the Inspection Plan Agreed Before Cutting?

The seventh check is a question about timing rather than about rigour. A plan drawn at AQL 2.5 under ISO 2859-1 before cutting changes behaviour upstream, because the bench knows what the end of the run will be measured against. A plan discussed after the goods are finished is a negotiation, and negotiation at handover is always more expensive than measurement at intake.

Also ask how defects are classified. Critical, major and minor should be defined before inspection rather than during it, because a classification agreed at the moment a defect is found is simply a negotiation conducted under time pressure.

Eight: Are Corrections Re-Checked or Merely Counted?

The eighth check is the one almost nobody asks, and it is the best predictor of whether the next run will be better. A floor that re-checks corrections and records them against the operation that caused them improves; a floor that counts corrections and packs the replacements repeats the same defect next season. Ask to see the record for the last rejected batch.

• Corrections are re-checked, not merely replaced, before the lot is released.
• Records name the operation rather than the individual, so the cause can be addressed.
• Rejected units are quarantined in a defined place, not returned to the line informally.
• The re-check result is recorded with the lot, which makes a later dispute resolvable on evidence.

What to Do With What You Saw

A walkthrough is only worth the record made during it. Write down the eight answers on the day, compare them with what the quotation promised, and raise any gap while the run is still in progress rather than at handover. Most disagreements that reach a lawyer began as a note someone did not write on a walkthrough.

1. Record the lot reference, the marker utilisation and the skive reading as numbers, not impressions.
2. Photograph the seam, the edge and the hardware reference under the same light each visit.
3. Note the inspection plan and the defect classification, and confirm both in writing.
4. Ask for the correction record of the last rejected batch before you leave the floor.

One further habit separates a useful walkthrough from a social one: ask the floor to state each value before you reveal what you expected it to be. A number given first is a number the house actually holds; a number confirmed after you have suggested it is worth very little. The same applies to the inspection plan, which should be quoted from the pack rather than recalled from memory.

On our own programme these eight checks sit inside a run of 7 production lines with about 200,000 pcs per month of capacity and export experience held since 2014, carried out through our validated partner factory Quanzhou Yingming Bags, an SGS-validated site. Sampling takes 6-10 working days, bulk 35-50 days, and third-party inspection can be added alongside the internal AQL 2.5 audit.

Frequently Asked Questions

People Also Ask