Custom handbag sampling is a single engineering build completed in 6–10 working days once hides and fittings are physically held, or 12–15 days where the brief calls for a frame, a moulded handle or a tooled closure. The unit is not made to be admired. It is made to prove what a drawing cannot: that the card pattern closes into the intended volume, that the handle drop still reads correctly under load, and that the operation sequence repeats.
Sampling therefore sits between the signed brief and the run. On our Quanzhou production floor the window opens only when components are in the building, and it closes with a measurement sheet rather than a photograph, because the sheet is what bulk is built and inspected against. The 35–50 day production window, inspection to AQL 2.5 under ISO 2859-1, and the 500-piece threshold per style and colourway are all costed from what this one unit establishes.
What One Sample Is Meant to Prove
A drawing settles appearance; a physical unit settles behaviour, and behaviour is where programmes fail. Leather stretches more across the fibre structure than along it, thickens at a turned seam and relaxes under sustained load. A hide that behaved beautifully as a swatch can refuse to fold at a 12 mm corner radius, or creep 15 mm at the handle after a week of carrying weight. None of that appears in a flat sketch, and all of it appears within ten working days.
The second purpose is repeatability, and it is the one buyers forget to ask about. An experienced hand can make almost any construction work once. Whether it works when cut in quantity, on a marker planned for yield, by operators meeting it for the first time, is a different question. Where an operation demanded unusual skill at the bench, that operation drifts across a run and should be flagged in writing while the pattern is still open.
Card, Pattern and the Substitute Shell
The first two days rarely involve the material you specified. They involve card, a pattern cutter and a set of numbers pulled off the tech pack — the least glamorous phase, and the one that decides whether the rest of the week is spent building or rebuilding.
From drawing to a working block
A card pattern is a set of pieces, each carrying three things beyond its outline: a seam allowance sized to the substance, a grain arrow fixing cutting direction, and balance marks that let an operator align pieces without measuring. A structured top-handle bag typically breaks into twelve to twenty-four pieces once gussets, facings, base boards, handle plies and pocket parts are counted; missing one from the schedule is the most common cause of a stalled day three.
Why the first build is not made in leather
Where the silhouette is new, the first shell is assembled from a substitute — a low-cost split, a coated paper or a non-woven that holds a fold without stretching. It answers questions that are expensive to ask in the specified hide: whether the gusset depth produces the intended volume, whether the mouth opens wide enough to use, whether the corner turns cleanly at that substance. Corrections then transfer back onto the card.
Seam allowance is the number briefs most often get wrong. A single-ply lining join closes at 6 mm; a turned seam through two plies of 1.6 mm bovine plus reinforcement needs 8–10 mm before it lies flat. State finished dimensions and substance together, or the allowance is guessed — and the guess surfaces at assembly.
Handle Drop and Strap Load
Handle drop is the measurement most argued about after a sample arrives, almost always because the brief never said where it was measured from. Drop runs between two points: the top edge of the mouth or the highest point of the shell, and the underside of the handle at its apex. Name both on the drawing and the argument disappears; leave it open and two people disagree by 20 mm on the same bag.
Load matters as much as datum. A drop of 110 mm cut empty opens towards 125 mm after carrying weight, because leather and thread both relax; a full-leather strap creeps along its length where a webbing-cored strap does not. Measure the unit empty and again after a stated load has hung for a stated time, and put both figures on the sheet.
- Datum. Name the two points and draw them on the sketch.
- Empty figure. What the pattern is cut to, and what the pattern room will defend.
- Loaded figure. What a customer's shoulder actually experiences.
- Anchor panel. Where each end lands, named by panel, so reinforcement sits behind it rather than near it.
The anchor is where engineering and aesthetics separate. A handle end landing on a mouth seam will eventually pull that seam however good the stitching is; one landed on a reinforced panel, with a stiffener behind the top edge and a boxed or bar-tacked stitch at each end, puts load into the shell instead of into a join. Where a strap carries load across a seam, a tensile check against a method such as ASTM D2209 settles it cheaply.
Closures: Trial-Fitting Hardware Before It Meets Leather
Closures fail distinctively: they look right, photograph well and misbehave in the hand. The cause is nearly always the fixing stack — the plies the fitting has to pass through. A turn-lock specified for two plies will not close over two plies plus a stiffener and a facing, and by the time that is visible the leather has been drilled.
| Closure family | What the trial fit must establish | Failure symptom on a first build |
|---|---|---|
| Magnetic snap | Pull strength against contents pushing outward; clearance for the halves to meet | Holds empty, pops open when filled |
| Turn-lock or twist clasp | Post length against the full ply stack; rotation clearance over the facing | Will not turn, or loosens over cycles |
| Zip set into a gusset | Tape width against seam allowance; pull travel against the handle | Wave in the tape, or the pull jamming |
| Slide or bar lock | Alignment of both halves across the mouth under asymmetric load | One side seated, one side proud |
| Flap with press stud | Stud position on the flap's relaxed curve, not its flat outline | Misses its socket once the flap sets |
The trial fit is unglamorous and worth insisting on: hardware is set first into a scrap build cut from the same stack the unit will use, cycled thirty to fifty times and checked for wear at the contact points. It also settles plating questions, because a finish that marks at the setting tool marks on the unit — and a marked finish cannot be repaired, only replaced.
Lining, Interlining and Padding
Interiors are returned more often than exteriors, usually for something invisible from outside. Lining is cut to the shell's internal dimensions, not to its flat pattern dimensions, and the difference is the thickness of every ply it sits against. Cut to the flat figure and it pools at the base; cut to the internal figure with a small ease allowance and it sits drum-tight without dragging the shell inward.
Interlining decides whether a soft shell stands or slumps: non-woven stiffeners give body without bulk, 0.6–1.0 mm poly board gives a base its flatness, 2–3 mm foam gives a panel its shape. Each occupies volume, and volume added inside is volume removed from the contents. Padding added after the shell closes is the classic second-round trigger — the bag still measures correctly outside and no longer fits the device inside.
Day 1 to Day 10: Task, Output, Confirmation
The allocation assumes a moderate structured construction with a new silhouette. Flat shapes compress towards six days; framed or size-matched builds run long or into the 12–15 day band. Treat the last column as your own job list.
| Working day | Engineering task | Physical output | What you confirm |
|---|---|---|---|
| Day 1 | Tech pack parsed into a panel schedule; card pattern drafted and balanced | Numbered card pieces with allowance, grain arrows and balance marks | Panel count, cut versus finished dimensions, allowance for the substance |
| Day 2 | Substitute shell assembled; gusset depth and mouth geometry corrected on the card | A mock-up shell that holds volume but carries no finish | Volume, proportion and clear mouth opening |
| Day 3 | Corrected marker laid on the hide; panels cut, sorted and matched | Cut panel set matched for grain, substance and shade | Hide placement, grain direction, consumption figure |
| Day 4 | Skiving, reinforcement laid, stiffener boards cut, hole templates marked | Prepared panels with thinned joins and marked hardware positions | Substance at each seam, reinforcement behind each load point |
| Day 5 | Shell closed: gusset set, base board inserted, corners squared | An unlined shell that stands on its own | Silhouette, corner radius, base flatness, symmetry |
| Day 6 | Handle built to the stated drop; strap length set and ends boxed | Handle and strap anchored into reinforcement, not into a seam | Drop empty and loaded; strap adjustment range |
| Day 7 | Closure trial-fitted on scrap, then seated on the shell and cycled | Closure operating, with the fixing stack proven on a scrap panel | Alignment when shut, resistance, behaviour when filled |
| Day 8 | Lining cut to internal dimensions; interlining, padding and pockets set | Lined interior with pockets checked against real objects | Pocket fit, no drag or pooling, silence in handling |
| Day 9 | Edge build completed, thread ends secured, branding applied | A finished unit ready for measurement and photography | Edge build, colour against the swatch, branding datum |
| Day 10 | Measurement sheet completed, photography taken, unit packed | The sample, its measurement record and its image set | Approval, or written revision notes within the turnaround |
Decisions grow steadily more expensive as the week advances, so close every structural question before day three and leave finish variables to the last two days. Day ten is not an afterthought either: a unit dispatched without a measurement sheet invites comments based on an impression, and impressions cannot be actioned.
The Tolerance Grid Your Sample Is Measured Against
A sample cannot be approved against taste alone. The grid below is what a measurement sheet should close against, stated before the unit is built so that approval is a comparison rather than a negotiation.
Weight does more work than it appears to. If unit five hundred weighs 40 g less than the approved sample, something in the build has changed, and it is worth asking what before the cartons are sealed.
Failure Points That Push a Build Into Round Two
Revision rounds are normal but not random. The same handful of decisions accounts for most second builds, and each is visible in the brief before a panel is cut.
- Drop stated without a datum. Built to one interpretation, reviewed against another.
- Allowance sized for the wrong substance. Puckers on a thick hide, splits on a thin one.
- Gusset depth taken from the flat drawing. Flat geometry is not assembled volume.
- Closures specified for two plies on a three-ply build. Discovered at day seven.
- Lining cut to external dimensions. Pooling at the base, or a lining that drags the shell.
- Strap anchors landing on a seam. Load carried by thread instead of by a panel.
- Padding added after the shell closes. Outside holds; the device does not fit.
- Edge coating over adhesive residue. The ridge appears a week later.
All of those are documentation failures rather than craft failures, which is the encouraging part. They are prevented by stating the numbers a brief usually leaves implied.
Revision Rounds: Calendar Cost First, Money Second
Buyers treat rounds as a cost line. The larger exposure is calendar time: each round occupies another window and defers bulk by the same days, so three rounds add roughly a month before the 35–50 day run even opens.
| Round | What usually triggers it | Working days added | Cumulative from day 1 | What else changes |
|---|---|---|---|---|
| First build, as quoted | Complete brief, materials held | 6–10 | 6–10 | Baselines: consumption, measurement sheet, photography |
| Second, dimensional | Drop, depth or proportion corrected on the card | +6–10 | 12–20 | Marker re-planned; consumption moves slightly |
| Second, material | Hide, lining or stiffener changed | +10–15 | 16–25 | New lot reservation, possibly re-testing |
| Third, closure or geometry | Mechanism changed or mouth re-cut | +8–12 | 24–37 | Hardware re-ordered; stack re-proven |
| Third, construction | Structure level or interior rebuilt | +10–15 | 26–40 | New bill of materials; price re-quoted |
| Beyond the agreed cap | Scope reopened late | Priced per round | Open-ended | Bulk start slips; deposit timing may shift |
Cap the rounds in writing and agree the price of extras before the first unit is built. The cap is not a trust issue; it keeps scope honest and concentrates attention on the first review, because a buyer who knows there are two rounds examines the first unit far more closely.
Ten Checks Before You Sign Anything
Approve against evidence. The sequence below takes twenty minutes with a tape, a scale and the objects the bag was drawn to carry.
- Measure the unit unloaded in the three dimensions stated in the brief and write the figures down.
- Take the handle drop from the datum on the sketch, not from wherever looks natural.
- Hang the stated load for an hour, re-measure the drop and note the difference.
- Operate the closure thirty times and feel whether the resistance changes across the cycles.
- Run a hand inside every corner for trapped stiffener, residue or a folded lining seam.
- Put your own phone and notebook into each pocket rather than measuring with a ruler.
- Set the bag on a flat surface and press each corner in turn to check for rock.
- Sight along the mouth line from directly above and compare the left and right halves.
- Weigh the unit to the gram and record it as the reference for bulk variance.
- Photograph every fault against a ruler before writing a single line of comment.
The last of these saves the most time. A marked-up image turns a preference into a location, and a location is something a pattern cutter can act on.
What the Revision Packet Should Contain
A revision packet is a change order, not a list of complaints. Group corrections the way the pattern room works — dimensional together, material together, hardware together — so one pass through the card captures everything. State each as a number against a named location and say explicitly what is approved: silence reads as dissatisfaction, and an unrequested adjustment is the fastest route to a third round. Return it inside the turnaround agreed at the outset.
From Approved Sample into Bulk
Approval of a sample is approval of a design, not of a run. A pre-production unit built on production tooling from bulk materials is the last point at which a gap between a careful sample and an efficient run can be caught at no cost. Where the schedule cannot accommodate it, place in-process checks on the operations flagged as needing unusual skill.
The arithmetic from there is fixed: 35–50 days of production, inspection to AQL 2.5 under ISO 2859-1 single sampling at general level II, shipment FOB Xiamen with CIF or DDP on request, and T/T with a 30% deposit and the 70% balance before shipment. A golden unit held by both parties settles what no specification resolves — hand, finish, colour and how a closure should feel.
If you are testing demand rather than committing to a season, ask about a pilot run below the standard threshold before committing to the full quantity.
Open a sampling window with the numbers settled
Send your tech pack, dimensions and the datum you measure handle drop from. The programme team in Quanzhou confirms material availability first, then runs a 6–10 working day build and returns the unit with a completed measurement sheet.
Frequently Asked Questions
What is a card pattern and why does it take two days?
A card pattern is the set of pieces cut in card, each carrying a seam allowance sized to the substance, a grain arrow fixing cutting direction, and balance marks for alignment. Drafting it means converting a drawing into twelve to twenty-four parts for a structured style, then checking each against the stated dimensions. Rushing it produces panels that fit once and not repeatedly.
Why is the first handbag sample built in a substitute material?
Because geometry questions are cheaper to answer in a low-cost split, coated paper or non-woven than in the hide you specified. The substitute shell proves whether the gusset depth gives the intended volume, whether the mouth opens far enough to use, and whether the corner turns cleanly at that substance. Corrections then go back onto the card.
How is handle drop measured on a handbag sample?
Between two stated points: a datum such as the top edge of the mouth, and a second point such as the underside of the handle at its apex. The brief must name both, because measuring from a different datum on the same bag can shift the figure by around 20 mm. Record it empty, then again after a stated load has hung for a stated time.
Why does the handle drop change once the bag is loaded?
Leather and the thread holding it both relax under sustained load, so a drop of 110 mm cut empty commonly opens towards 125 mm after carrying weight. Full-leather straps creep along their length more than webbing-cored straps with a leather facing. Measuring only the empty figure is why a bag that matched the brief at approval feels different in use.
What load should a handbag strap be tested at before approval?
Use the load the bag was drawn to carry rather than a generic figure: a laptop and a wash bag for an everyday tote, less for an evening piece. Hang it for an hour, re-measure the drop, and check the anchor points for movement. Where a strap carries load across a seam, a tensile check against a method such as ASTM D2209 settles it quickly.
What is a closure trial fit and why does it come before assembly?
Hardware is first set into a scrap build cut from the same ply stack the production unit will use, cycled thirty to fifty times and checked for wear at the contact points. This establishes whether the post length suits the full stack including stiffener and facing, and whether the plating marks at the setting tool. Without it, the discrepancy is found after the leather is drilled.
Why does a magnetic closure hold empty but open when the bag is full?
Because the contents push outward with more force than the magnet was rated for, and any gap between the halves reduces the effective pull sharply. The trial fit must test the closure with the bag filled rather than closed on itself. Where more pull is needed, either the magnet grade changes or the land area behind it does.
How should lining dimensions relate to the shell dimensions?
Lining is cut to the shell's internal dimensions, not to its flat pattern dimensions, and the difference is the thickness of every ply it sits against. Cutting to the external figure leaves fabric pooling at the base; cutting too tight drags the shell inward. A small ease allowance lets the lining sit drum-tight without distorting the exterior.
How does padding change the internal volume of a bag?
Every millimetre of foam added inside a panel is a millimetre removed from what the bag holds. Padding introduced after the shell closes is a classic second-round trigger, because the exterior still measures to the brief while the device the pocket was drawn around no longer fits. Confirm pocket dimensions against the physical object on the substitute shell.
What is a measurement sheet and who signs it?
It is the record of every figure taken from the approved unit: finished dimensions, drop empty and loaded, strap range, mouth opening, closure alignment, base behaviour, edge build and weight. The programme team completes it and the buyer confirms it. It then becomes the reference the bulk run is inspected against, which is why it matters more than the photographs.
How much calendar time does each revision round add?
A dimensional correction usually adds 6–10 working days, a material change 10–15, and a closure or geometry change 8–12. Cumulatively, three rounds can put a programme 26–40 days from its start date, and the 35–50 day bulk window only opens afterwards. Cap the number in writing and agree the price of extra rounds first.
What happens if I change the closure after the first sample?
The fixing stack has to be re-proven, which usually means a scrap build and a fresh trial fit, and the new fitting has to be ordered. Expect 8–12 working days and a change to the bill of materials. Where the closure shapes the mouth rather than sitting on top of it, the panels around it are re-drafted as well.
What should a revision packet contain?
Corrections grouped the way the pattern room works: dimensional changes together, material changes together, hardware changes together. Each stated as a number against a named location, each supported by a photograph taken against a ruler, plus an explicit statement of what is approved. Silence on a detail is easily read as dissatisfaction and produces unrequested changes.