How to Approve a Handbag Sample Without Hidden Surprises Later

How do you approve a handbag sample so nothing surprises you later? Run nine checks in a fixed order — dimensions under a 1 kg load, stitch density and seam allowance, hardware cycles and plating, colour under D65, interior and pockets, branding position and permanence, weight against the golden unit, packaging and barcode, then a written approval that names the revision. A sample approval is a standard, not an opinion: whatever you do not measure at approval becomes the tolerance the bulk run silently adopts. Priced out, a defect caught at approval costs nothing, the same defect caught at inspection costs a re-sort on 500 units, and caught after delivery it costs a return at retail value plus the listing.

Where the same defect is cheapest to catch

What a Bad Approval Costs, Priced Out

Approval is the only stage in the programme where a defect is free to fix. Before it, the bag does not exist in quantity; after it, the bag exists 500 times. That asymmetry is the entire argument for a disciplined approval, and it is easier to act on when it is priced rather than described.

Read the third row carefully, because it is the one brands walk into. A defect found at inspection on a lot of 500 is not a 500-unit problem in the way it feels; it is a calendar problem, because the remedy — re-sort, rework or re-make — runs against a launch date that cannot move. And the fourth row is worse than its price suggests, since a return rate that shows up in a marketplace account affects the listing rather than the order.

This is why the checks below are ordered by money rather than by convenience. The first two catch the defects that generate field returns; the middle ones catch the defects that generate bad reviews; the last two catch the defects that generate operational pain at the warehouse. Running them takes under an hour on a bench and about the same again on paperwork.

The Nine Checks, in the Order They Catch Money

Nine checks cover the failure modes that actually cost money, and the order matters because the early ones are the ones that cannot be fixed later. Work down the list with the sample in front of you and the pack beside it, and write the result of each check rather than an impression of it.

1. Dimensions measured under a 1 kg load.
2. Stitch density and seam allowance at every structural seam.
3. Hardware: cycles, closing force, plating and salt-spray evidence.
4. Colour under D65 against the retained swatch.
5. Interior: lining, pockets, and anything sharp or unfinished.
6. Branding position and permanence under abrasion.
7. Weight against the golden unit.
8. Packaging, barcode and carton marks.
9. A written approval naming the revision and everything accepted.

Check One: Dimensions Measured Under Load

Measure the bag loaded, not empty. Leather and webbing relax: handle drop measured dry differs from drop measured with 1 kg inside by 5-8 mm, and a base that reads flat on a bench can sag by 4-6 mm once there is weight in it. The number that matters at approval is the number a customer will experience, which is the loaded one.

• Body width, height and gusset. Within ±3 mm of the pack, measured on a relaxed unit.
• Handle drop. Within ±5 mm under a 1 kg load, from the top edge to the apex.
• Strap length on an adjustable strap. Measured at the mid setting, ±8 mm, with both sliders holding position under load.
• Base. Sits flat with no rock; a base that rocks by more than 2 mm reads as a fault in a photograph.

Carry the loaded bag for a minute before writing anything down. Comfort is not sentimental: a 25 mm strap carrying 3 kg distributes load differently from an 18 mm one, and a handle that digs is a handle that generates returns regardless of how the bag photographs. This is the check that separates a sample that looks approved from one that is approved.

Check Two: Stitch Density and Seam Allowance

Count stitches over 25 mm at three places: a structural seam, the strap attachment and the lining. Target 7-9 stitches per inch with Tex 40 bonded polyester. Below about 6 per inch the seam opens under load; above about 10 the needle perforates the material and the seam becomes the failure line rather than the join.

The strap attachment row is the one to insist on. A box stitch of 25 × 25 mm with a rivet through a 2 mm reinforcement patch is the difference between a bag that survives a season and one that fails at month three, and it is the joint that fails first in the field. It also costs very little at approval and a great deal at inspection, which is exactly the profile of a thing worth checking.

Check Three: Hardware Cycles and Plating

Operate every moving part more times than feels necessary. Cycle the zip over its full travel repeatedly; open and close the magnetic snap or turn-lock; extend and retract an adjustable strap. Binding, gritty travel or a slider that catches is a bulk failure waiting to happen, because it will appear on a proportion of 500 units rather than on the one in your hand.

• Zip. Smooth over full travel both ways, no binding at the corners; a branded puller should not snag the tape.
• Magnetic closure. Closing force in the region of 1.2-1.8 kgf: enough to hold a loaded flap, light enough to open one-handed.
• Plating. A neutral salt-spray result to ASTM B117 of 24 hours to first corrosion, with the report attached rather than promised.
• Nickel. Release tested to EN 1811 for anything in prolonged skin contact where the destination is the EU.
• Sharp edges. Run a finger around every component; a toggle or a plate with a sharp edge is a return and a complaint.

Ask for the finish reference at this stage rather than at inspection. A physical approved component, tagged with its finish code and retained, is what keeps the second season matching the first — plating baths vary by lot and light gold specified as gold comes back three shades different next time.

Check Four: Colour Under the Light It Sells In

Judge colour under a stated light source, and where possible the light it will be sold under. D65 is the standard reference; store lighting is frequently warmer and shifts a dyed hide noticeably. The acceptance criterion is a Delta E of 2.0 or better against the retained physical swatch, measured rather than eyeballed, and the swatch is the reference rather than the screen you first saw the design on.

• Body colour against the swatch card, under D65, from three angles.
• Contrast elements — thread, edge paint, lining, zip tape — each against its own reference.
• Lot consistency where more than one unit is available: place two side by side and look for a difference that a customer standing at a shelf would see.
• Fastness documented: rubbing to ISO 105-X12, light to ISO 105-B02, and perspiration on handles.

Chemistry belongs with colour rather than after it. Restricted substances in a dyed hide or a coated trim are tested, not asserted: chromium VI under REACH Annex XVII is tested to ISO 17075, California Proposition 65 covers listed substances in hardware and coated trim, and CPSIA applies where an item is marketed to children. Ask for the reports while the order is running, because a test commissioned after packing delays shipment by the length of the test.

Check Five: Interior, Lining and Pockets

The interior is where a sample gets approved on looks and fails on use. Turn the lining out and look at the seams; run a hand around every pocket for a sharp edge or an unfinished binding; check that a phone and a small notebook actually fit the pockets as drawn. A pocket specified at 150 × 110 mm that finishes at 140 × 100 mm is a pocket that does not hold the thing it was drawn for.

• Pocket dimensions within ±3 mm, and tested with the object they are meant to hold.
• Lining seams at 8 mm allowance, clean, with no loose thread ends inside.
• No sharp edges on bindings, stabilisers or hardware backs inside the bag.
• Base board sitting flat with no ridge visible through the lining.
• Odour. A strong solvent smell at approval is a complaint waiting to happen; it should be noted rather than tolerated.

Check Six: Branding Position and Permanence

Measure the mark from the datum edge on both axes and compare it with the pack: ±1 mm is a reasonable tolerance. Then test permanence, because a mark that rubs off in a month is a branding failure rather than a cosmetic one. Rub a foil mark firmly with a thumb; flex an embossed patch; tug a woven label at its stitch line.

• Position within ±1 mm of the datum, measured rather than judged, on every unit available.
• Emboss or deboss depth at 0.4-0.6 mm, clean-edged and consistent across units.
• Foil adhering after a firm thumb rub; a foil applied at the wrong temperature lifts after three to four weeks.
• Woven label secure at the stitch line, ends tucked as specified.
• Die ownership recorded in writing before the deposit moves.

The ownership line is the commercial half of this check. The tooling paid for should be identified, stored, and never run for another brand, and that clause is what makes the repeat cheaper: tooling amortised once amortises again. Where the branding method is changed after the die is cut, the charge is paid twice — which is the single clearest example of money wasted through a late decision in this trade.

Check Seven: Weight Against the Golden Unit

Weigh the approved unit and record the figure. Weight within 3% of the golden unit is a reasonable band, and the check is valuable because it catches what a visual inspection misses: a dropped interlining, a thinner reinforcement patch, a lighter gauge of base board. All three look fine on a bench and all three shorten the life of the bag.

Record the figure on the approval note as well. It means the pre-shipment inspector has a second, objective check alongside the dimensional ones, and it gives the repeat order something to compare against before the cartons are opened. This is the same logic that makes a pre-production unit worth building — three units compared side by side reveal station variation that one unit hides.

Check Eight: Packaging, Barcode and Carton

Scan the barcode rather than looking at it. A code that reads on a phone is not the same as a code that reads at a warehouse scanner, and a failed scan is a rework at the destination, which is the most expensive place to rework anything. Check the polybag gauge and venting, the dust bag, the hangtag stock and string, the carton ply and dimensions, the pieces per carton, the gross weight and the carton marks.

• Barcode scans first time with a scanner rather than a phone app, in the symbology the market requires.
• Polybag at the specified gauge and vented, with warning text in the destination language.
• Carton to the specified ply and dimensions, with stated pieces per carton and gross weight.
• Marks legible and matching the packing list, including any marketplace prep code.
• Country-of-origin and care wording present and correct, because both are printed rather than applied.

The Approval Note That Protects You Later

The approval note is one page and it is the document that resolves every later disagreement. It names the style code and revision letter, the date, the golden unit retained by each side, the measurements recorded at approval, and — critically — an explicit statement that everything not listed as a residual item is approved. Without that last sentence, every subsequent conversation can reopen a closed decision.

1. Style code and revision letter, matching the drawing the bulk run will use.
2. Date of approval and the name of the person who approved it.
3. The nine check results, with numbers rather than ticks.
4. Residual items listed with measurements, each marked must-fix or defer to the next order.
5. The golden unit reference, dated and held by both sides.
6. A statement that everything not listed is approved.

Send the deposit and book the slot in the same action as the approval. Slots tighten before Chinese New Year and during the autumn shipping rush, and a slot booked at approval rather than a week later is worth more than any concession negotiated on price. Bulk then runs 35-50 days, inspection draws a sample under ISO 2859-1 and applies AQL 2.5, and goods are handed over FOB Xiamen with CIF or DDP available. The full sequence, with the gates where money moves, is set out in the private label process from enquiry to shipment.

What to Do When Something Is Almost Right

Almost right is the dangerous state, because it invites a conditional approval and conditional approvals are how defects reach 500 units. There are only three honest options: accept it as is and record the deviation, request the change and accept the days, or defer it to the next order in writing. What is not an option is approving while hoping bulk will be better, because bulk will be the same.

• Accept and record. A deviation of 2 mm within a ±3 mm tolerance is not a defect; note it and move.
• Request and pay the days. A dimension change costs 2-4 days for a new pattern; a hardware change costs 5-10; a branding method change 7-10 plus tooling.
• Defer in writing. A preference that is nice-to-have rather than must-fix belongs on the next order, listed and dated.
• Never approve conditionally. Bulk reproduces the approved unit, including the thing you hoped would improve.

Where a programme runs several silhouettes, apply the same approval note to each and keep them in one file. A brand that holds a specification, a golden unit and a dye-lot card per style can often skip sampling entirely on the repeat, removing 6-10 working days — which is the practical reward for an hour spent on a bench now, and the reason the sampling timeline is worth holding rather than compressing.

Frequently Asked Questions

People Also Ask