From Tech Pack to Luxury Bag Sample, Step by Step

Before step one: what has to be true

Three things must be settled before the pack goes to the bench. The specification must carry tolerances rather than dimensions alone. One person must be named as the approver for the round. And the hardware decision must be at least provisional, so tooling can run in parallel with pattern work rather than after it. Where any of these is missing, the first step is not cutting, it is a conversation, and that conversation is cheaper now than it will be in three weeks.

The fourth condition is subtler and worth stating: the brief should be frozen. A brief that moves during a round does not produce a corrected sample, it produces a different bag, and the difference will not be visible until the corrections are assessed against notes that no longer describe what was asked for.

Step one: pack review and the gap list

The bench reads the pack the way a builder reads a drawing: looking for what is missing rather than admiring what is present. The output is a gap list — every field that is blank, ambiguous or physically unrealisable, with a question against each. Common entries are substance, skive depth, seam allowance, interlining weight, edge build and thread ticket number. The list goes back to the brand and is answered in writing before anything is cut.

This step is where most programmes either save or waste their development time. Answering six questions costs a day; discovering the same six gaps in a finished sample costs a round.

Step two: pattern development

Pattern work converts the pack into pieces that can be cut. On a structured bag this means a full pattern set: body panels, gussets, facings, linings, pocket pieces, reinforcements and the interlining shapes that bond to each. Every piece carries a grain direction, a seam allowance and a note on where it is skived and to what depth. Digital pattern work allows the set to be adjusted and re-graded without cutting material, which is why iteration at this stage is inexpensive.

The decision inside this step is whether to cut a toile. On a novel construction, a toile in inexpensive material resolves proportion and assembly order before the final leather is committed. On a proven construction it is unnecessary, and the pattern goes straight to cutting.

Step three: material and component call-off

Every line of the bill of materials is either confirmed in stock, ordered, or flagged as client-supplied. Leather is called off against a single dye lot, with the full requirement for the sampling phase reserved rather than reordered mid-process. Hardware is checked for finish and shade against the reference clasp. Thread, interlining, adhesive and edge paint are matched to their specification lines.

• Leather: species, tannage, substance, temper, and dye lot reserved before cutting.
• Lining: fibre, weight and shade matched to the reference.
• Interlining and reinforcement: type, weight and the panels each bonds to.
• Thread: fibre, ticket number, needle size and colour for each stitch position.
• Hardware: alloy, plating thickness, finish, and the reference clasp for comparison.
• Edge paint: type, number of coats, target build and sheen against a chip.

Step four: cutting

Panels are cut in sequence, and on any bag where panels meet visibly, they are taken from adjacent positions on the same hide. This is a shade control rather than a cutting convention, and it is what keeps the front and the gusset of one bag from reading as two different colours. On exotic skins the scaled pattern is placed on the skin and photographed before the knife runs, because a mis-placed piece is not recoverable.

Each cut piece is labelled with its panel identity and its position in the sequence, so that a problem found in assembly can be traced back to a specific hide and a specific position.

Step five: closing and assembly

Closing turns flat pieces into a three-dimensional body. Skiving, bonding, folding and stitching run in a sequence determined by the construction notes, and that sequence matters: two routes to the same silhouette produce different hand and different durability. Stitch density is set to the specification, 8 stitches per inch on visible calfskin topstitching, and the tension is checked on a test piece before the first seam runs.

Where an assembly problem appears — a corner that will not turn, a gusset that fights the body — the decision is whether to adjust the pattern or the sequence. Adjusting the sequence is faster and usually correct; adjusting the pattern is slower and usually necessary only when the geometry itself is wrong.

Step six: edge finishing and hardware setting

Edge finishing is where a sample begins to look finished, and it is the operation most often under-specified. Coats are applied, sanded and re-applied to the target build, then compared for sheen against the reference chip. Hardware is set with a jig so that plates seat flatly and escutcheons centre within 0.5 mm, and every closure is cycled twenty times by hand before the piece leaves the bench.

Step seven: assessment against the pack

The finished sample is assessed line by line against the pack rather than admired as a whole. Dimensions are measured against their tolerances, stand and balance are checked loaded as well as empty, hardware is cycled and pull-tested, colour is measured against the signed panel under D65, and stitch density is counted. Every finding is recorded with a photograph and a measurement.

The assessment is also where the decision branches. If the findings are local — a stitch line, an edge build, a hardware shade — the piece goes to correction. If they are structural — proportion, stand, assembly order — the pattern goes back to step two, and that is a second round rather than a correction.

Step eight: correction and the second build

Every finding from step seven becomes a numbered correction, and the second build is assessed against that list before anything else is considered. A point is only closed when it is confirmed on the piece, not when the bench believes it has been addressed. Where a correction cannot be closed on a second build, the cause is usually upstream in the pattern rather than at the bench, and the honest move is to say so rather than to attempt a third time on the same geometry.

Re-cutting and re-finishing one panel costs around USD 4.80 in bench time, before material and re-assembly. That figure is why the correction list should be consolidated by one approver rather than assembled from several stakeholders: the count of corrections, not the count of opinions, is what determines the cost of the round.

Step nine: pre-production and sign-off

The final step is built on the production line with production tooling and production jigs, by the people who will run the order. It proves repeatability, which a sampling bench cannot: the sampling bench proves that the bag can be made, the line proves that it can be made five hundred times. The pre-production piece is inspected at the same level as any production lot and, if it passes, is signed, dated, photographed and stored away from light as the reference for bulk.

1. Confirm the pack carries tolerances, one approver is named and the hardware decision is at least provisional.
2. Review the pack for gaps and answer every one in writing before cutting.
3. Develop the full pattern set and decide whether a toile is needed.
4. Call off every component, reserving leather against a single dye lot.
5. Cut panels in hide sequence, photographing placement on exotic skins.
6. Close the body to the specified construction sequence and stitch density.
7. Build the edges and set the hardware with jigs, cycling every closure twenty times.
8. Assess line by line against the pack, recording photographs and measurements.
9. Correct, re-build and confirm each numbered point on the piece before signing off.

Where the process branches

The steps are sequential, but the process is not linear, because three findings send the work backwards. A structural finding returns the pattern to step two. A component finding sends the hardware back to tooling, and the tooling window then dominates the schedule rather than the bench. A material finding reopens the conversation about substance, temper or tannage, and may require a fresh dye lot to be reserved.

What should never happen is a branch that is not recorded. Every return to an earlier step is noted on the pack with the reason and the date, so that the finished sample carries a history rather than a mystery. That history is what allows a brand to understand why a style took three rounds rather than two, and it is the only reliable basis for planning the next one.

Steps from pack to signed sample

9

Working days per round, standard

6 to 10

Working days per round, complex

12 to 15

Stitch density

8 stitches per inch on visible calfskin topstitching

Tolerance, structured body

±1 mm; hardware position ±0.5 mm

Colour confirmation

Delta E 1.0 against the signed panel under D65

Hardware seating

Centred within 0.5 mm, plates seated flatly with a jig

Closure cycling

20 cycles by hand at the bench, 5,000 on a rigged frame per style

Correction cost per panel

Around USD 4.80 in bench time

Minimum order quantity

500 pieces per style per colourway

Bulk lead time

35 to 50 days from approved sample

Shipment terms

FOB Xiamen

Programme history

Export work since 2014 across 7 production lines, about 200,000 pieces per month

Frequently Asked Questions

People Also Ask