Fit Samples vs Production Samples for Luxury Bags: What Each One Decides, the Three Checks That Separate Them and When to Run Both

One Paris House, Two Rounds, Two Different Questions

A Paris house brought us a structured top-handle in box calf, four colourways and a trunk show eleven weeks out. The first sample came back and the founder rejected it because the handle drop felt wrong against the shoulder. The second came back and she rejected it because the gusset read too deep in the lookbook photographs. Neither rejection was about the same thing, and neither sample had been asked the question it was rejected for.

What had happened is common: the piece was being reviewed twice against two different standards, one about how it sat on the body and one about how it would reproduce 500 times, and nobody had written down which round was for which. Once the two rounds were named and separated, the programme settled in one further round and hit the show date.

What a Fit Sample Actually Decides

A fit sample decides how the piece behaves on the body: handle drop, strap fall, gusset depth, balance empty and loaded, and whether the opening can be worked with one hand. Nothing else about it is authoritative, and treating it as authoritative is the single most expensive misreading in sampling.

Handle drop

Measured from the crown of the handle to the top of the body, held at ±2 mm of the drawing

Strap length

1050 mm as a working cross-body reference, checked under a 4 kg load for 30 minutes

Gusset depth

90 mm on a day bag; the dimension most often misread in photographs and least often drawn

Balance

The piece must sit flat when set down and must not tip forward when empty

Opening

The closure must work one-handed, and it is cycled ten times before any judgement is made

Base

A 1.5 mm base board keeps the floor flat without adding visible bulk at the fold

A fit sample may be cut from a trial hide and may carry the closest stock fitting rather than the engineered one. That is acceptable, and it is the point: geometry is the question, material is not, and spending a round on shade before the shape is settled spends the round on the less important thing.

What a Production Sample Actually Decides

A production sample decides what the run will look like. It is the first piece of bulk rather than a better version of the prototype, and it is made on the bulk tooling, by the bulk method, from the programme hide.

• Leather cut from the programme hide rather than the nearest shade on the shelf.
• The specified hardware fitting, plated at 0.15-0.3 microns, cycled and photographed.
• Stitch pitch at 8-9 per inch in 0.45 mm thread, counted over 25 mm rather than estimated.
• Three thin edge coats, examined after conditioning cold and warm rather than at room temperature only.
• Interlining at 0.8 mm, the same weight the run will use, not a heavier one that flatters the panel.
• Interior layout, pocket depth and whether the lining shows at the opening.
• Packing identical to bulk packing, because a piece presented better than bulk proves nothing.

The discipline here is unglamorous. A production sample made by the strongest hand on the bench from the best hide in the store defines a standard the run cannot hold, and the rejection it eventually causes is really a rejection of the sampling method rather than of the work.

Three Checks That Separate Them

Three checks distinguish the two samples quickly, and they are worth writing at the head of the tech pack so that nobody on either side has to infer which round is running.

The last row is the one that carries weight. A fit sample invites change; a production sample ends it. Programmes that treat the production sample as another draft discover, somewhere in week three of bulk, that nothing has actually been approved.

Where Programmes Confuse the Two

Four confusions account for most of the extra rounds we see, and all four are procedural rather than technical.

None of these is a failure of skill. They are failures of labelling, and labelling is free.

The remedy we use is a single line at the head of every sample parcel note stating which round is running and what that round is entitled to decide. It takes twenty seconds to write and it removes the ambiguity that produces the extra round. Where a buyer wants a piece that can be photographed for press at the same time, we make a separate piece rather than loading one sample with two jobs, because the two jobs pull the review in opposite directions.

How the Two Sit Inside One Calendar

The sequence is short and fixed, and the two samples occupy specific positions in it rather than floating.

1. Tech pack complete, with the seven numbers written: leather 1.2-1.4 mm, stitch 8-9 per inch, thread 0.45 mm, three edge coats, plating 0.15-0.3 microns, interlining 0.8 mm, #5 zip.
2. Fit sample in 6-10 working days, or 12-15 where the piece carries a frame, an exotic skin or engineered hardware.
3. Fit review returned inside five working days, written as measurements rather than impressions.
4. Production sample made on the bulk tooling from the programme hide.
5. Sealing: three references, one held by each side and one held back, each signed across the seal.
6. Bulk in 35-50 days, inspection at AQL 2.5 under ISO 2859-1, handover on FOB Xiamen terms.

Two windows inside that sequence absorb almost all the delay in a programme: the review period after the fit sample, and the hide lead time before the production sample. Both can be run in parallel with other work if they are anticipated, and both stop the calendar dead if they are not.

What Each One Costs to Get Wrong

The cost of a mistake rises by an order of magnitude at each gate, which is the whole argument for separating the two samples rather than collapsing them into one.

Read as insurance, the arithmetic is not close. A round spent early is a few days; the same correction discovered in week four of bulk is a commercial problem rather than a technical one, and it is usually the same correction, because dimensional errors do not correct themselves across a larger run.

Tolerances and the Paper Trail

Written tolerances are what make the two samples comparable, and they belong on the tech pack rather than in correspondence.

A measurements sheet comparing the piece against the drawing, line by line, should travel with both samples. It is the item most often omitted from a parcel and the one that prevents the most disputes, because a discrepancy found on paper costs an email and the same discrepancy found at bulk costs a season.

Hardware, Edges and What Only the Second Sample Proves

Three attributes cannot be proven by a fit sample at all, and attempting to judge them there is what generates the third round.

• Plating durability: a stock fitting in the right finish says nothing about 0.15-0.3 microns carrying 24 to 48 hours of salt spray.
• Edge behaviour: three thin coats look identical to one thick one at handover, and only conditioning cold and warm separates them.
• Grain consistency: one hide tells you nothing about whether the tannery can supply 500 pieces that match it.

From Both Samples Into Bulk

Once the production sample is approved, the programme moves onto fixed commercial terms: MOQ 500 pcs per style and colourway, bulk in 35-50 days, inspection at AQL 2.5 under ISO 2859-1 with third-party inspection available, and handover on FOB Xiamen terms. Export experience on this programme runs since 2014 across 7 production lines with about 200,000 pcs per month of capacity, delivered through our validated partner factory Quanzhou Yingming Bags.

What carries across the boundary is not the sample itself but three things derived from it: the sealed reference, the written specification sheet, and the agreed tolerances. Those are the instruments that make bulk measurable, and they are all created on the day of approval rather than at any point afterwards.

Frequently Asked Questions

People Also Ask