How Kering and LVMH Influence Luxury Bag Sourcing

The Audit That Arrived Without Warning

A client rang on a Tuesday to say that an auditor would be at the site on Thursday, and could the batch records be ready by then. The client did not own the site and had not commissioned the audit; the requirement had come down from a group one step further up the chain, which had decided that every house in its supply network would be re-verified that quarter. Nobody had told us because nobody had thought to.

What saved the day was not a scramble. It was a habit: every material delivery is logged against a batch, every batch is linked to the pieces cut from it, and the link is written down rather than remembered. The auditor asked for a traceability path on a finished piece and got it in under twenty minutes. Two other houses in the same network were asked the same question and took three weeks.

That is how group practice reaches independent houses: not through a directive, but through a requirement arriving at the bottom of a chain with a short deadline attached. The houses that hold the record answer in twenty minutes. The houses that do not spend three weeks reconstructing one, and the difference is the whole subject of this page.

How Group Practice Reaches Independent Houses

The largest groups employ, commission and audit at a scale that sets the working assumptions of the entire category. When they adopt a traceability standard, a testing cadence or an audit protocol, it becomes the default expectation for everyone selling into the same market, including houses that will never supply them. Five cascades do most of the work.

None of the five arrive as an announcement. They arrive as a line in a client's questionnaire, a clause in a purchase order, or a request from an auditor who assumes the answer exists. Independent houses experience them as administrative noise until one of them carries a deadline, at which point the house either has the record or does not.

The useful response is to treat all five as standing requirements rather than as client-specific asks. Built once, they cost little to maintain; built under pressure, each one costs a scramble and at least one uncomfortable conversation with a client who is themselves under pressure from above.

Cascade One: Traceability Becomes the Default

Traceability used to be a claim. It is now a test with a time limit: given a finished piece, produce the batch it came from, and produce it quickly. The working target worth setting is traceability within 24 hours of a request, and the tougher internal bar is under one hour, because an auditor standing on the floor is not going to wait a day.

The mechanism is three links: the material delivery is logged against a batch, the cutting record links the batch to the panels, and the assembly record links the panels to the finished piece. None of the three is difficult. All three are usually missing, because each is maintained by a different person and nobody owns the join between them.

Assign the join. One record, one owner, tested once a quarter with a real request rather than a theoretical one. Pick a finished piece at random, ask for its batch, and time the answer. If it takes longer than an hour, the record is not working, and it will not work on the day it is needed either.

Cascade Two: Testing on Every Delivery

Restricted-substance testing has moved from an annual exercise to a per-delivery expectation. A material arrives, it is tested or it carries a valid test report covering the batch, and the result is retained for the life of the product plus a period afterwards. Five years of retention is a defensible rule, and it is the figure most questionnaires are moving toward.

Two details decide whether this is manageable. First, the report has to cover the batch actually delivered, not a generic certificate issued for the tannery years earlier; auditors check the date and the batch reference. Second, the retention period has to be stated somewhere, because a result that is kept until somebody deletes it is not a retention policy.

Then test the assembly and not just the inputs. Leather, lining, adhesive, thread and coating each have their own exposure, and a piece assembled from five compliant inputs can still fail if a process introduces something. Testing the finished piece once per style per season is the least costly way to find that out before an auditor does.

Cascade Three: Capacity Gets Booked Earlier

When the largest groups secure capacity, they secure it months ahead, and the effect is that everyone else has to as well. A house that used to book eight weeks out now books 90-120 days out, not because its own programmes changed but because the line it shares is fuller earlier in the season than it used to be.

The consequence for an independent label is calendar discipline. Sampling at 6-10 working days and bulk at 35-50 days have not changed, but the slot those windows occupy now has to be reserved before the sample is approved rather than after. Houses that wait for approval before booking frequently find the next available slot is a season away.

Say so to clients early. A house that quotes a delivery window contingent on a slot being held is being honest; one that quotes the same window without the slot is setting up the conversation nobody wants to have in month two. Capacity is a fact, and it is better stated as one.

Cascade Four: The Calendar Speeds Up

Collections have moved from two drops a year to four to six, and the cadence has followed the groups rather than the independents. That compresses development: what used to be a season of sampling becomes a window of weeks, and the number of styles a house can develop properly in that window falls unless something changes.

The answer for most houses is fewer styles, developed better. Three styles specified properly inside a compressed calendar beat eight specified loosely, and the discipline that makes it work is the one described on our craftsmanship page: retain the reference, digitise the pattern with version control, and change one variable per revision.

It also argues for a core range that does not enter the calendar at all. Styles that repeat across seasons are specified once and re-run, which frees the compressed window for the genuinely seasonal pieces. A house with a stable core can absorb four to six drops a year; one without it cannot.

Cascade Five: Audit Findings Have Deadlines

An audit that produces a finding produces a deadline, and the deadline is short. Thirty days is common for a minor finding and sixty for a major one, with evidence of closure required rather than a promise. Houses that treat an audit as an event rather than a cycle discover this at exactly the wrong moment.

The management discipline is unglamorous: log every finding, name an owner, date the closure, and keep the evidence. A finding closed with a photograph and a date is closed. A finding closed with an explanation is not, and the second attempt is always more expensive than the first.

Then pre-empt the common ones. Most findings in this category concern records rather than conditions: a missing batch link, an expired test report, an undocumented process change, a training record that was never signed. All four are administrative, all four are preventable, and all four are what an unprepared house fails on.

Traceability response

Batch identified within 24 hours, under one hour as an internal bar

Traceability links

Delivery to batch, batch to panels, panels to finished piece

Record owner

One named person owning the joins between the three links

Self-test

Random finished piece traced once a quarter, timed

Substance testing

Per material delivery, report covering the batch actually delivered

Report retention

Five years from delivery, stated as a policy rather than a habit

Finished-piece test

Once per style per season, covering assembly as well as inputs

Capacity booking

90-120 days ahead, slot held before sample approval

Drop cadence

Four to six per year for seasonal styles

Core range

Styles repeating across seasons, specified once and re-run

Audit closure

30 days minor, 60 days major, with dated evidence

Sampling window

6-10 working days, 12-15 for complex builds

Bulk window

35-50 days from approval, contingent on a held slot

Cost anchor

USD 4.80 trim and hardware allowance before material

What an Independent House Has to Hold

Five records, maintained continuously rather than assembled on request. A material delivery log with batch references. A cutting record linking batches to panels. An assembly record linking panels to finished pieces. A test report file with retention dates. And an audit finding log with owners and closure dates. That is the whole list.

None of the five require software, and none require scale. They require one person who owns each and a quarterly check that they still work. Houses that have them describe audits as routine. Houses that do not describe them as disruptive, which is the same observation from two directions.

The point is not compliance for its own sake. A house that can trace a piece in twenty minutes can also answer a customer question in twenty minutes, which is a commercial advantage that appears long before an auditor does. Records kept for compliance turn out to be useful for everything else.

What It Costs

Testing is the visible cost. Per-delivery substance testing on five inputs across a season is a real cost line, and it scales with the number of materials rather than with the number of pieces, which is why a house with a narrow material palette pays less for the same coverage. Narrowing the palette is the least costly compliance move available.

Record-keeping is the invisible cost, and it is bench and office time rather than money. It is also where houses under-invest, because the cost is continuous and the benefit is occasional. Set it against a trim and hardware allowance of USD 4.80 before material and the comparison is clearer: the record is a fraction of the allowance and prevents the failure that costs the most.

The expensive cost is a failed audit that closes a client relationship. That cost does not appear on any sheet, and it is the reason to hold the five records as standing practice rather than as a response to a questionnaire. It is insurance with an unusually clear payout condition.

Where Independent Houses Have an Advantage

There is a counter-current worth noticing. The same cascades that raise the bar also create space at the top of the market for houses that can move faster and take smaller runs. A group programme is designed around volume; an independent house can take 500 pieces per style and colourway and still make the economics work.

Speed is the second advantage. A decision that takes a quarter inside a group structure takes a week in a house where the person who specifies the piece is the person who answers the question. On sampling at 6-10 working days and bulk at 35-50 days, that difference compounds across a season.

The third is candour. A house that can say precisely what it tested, what it measured and what it did not is able to make an argument that a larger organisation cannot make as quickly. Buyers increasingly respond to that, and the shift behind it is described in more depth on our luxury trends page.

How We Prepare for Group-Standard Audits

Our validated partner factory Quanzhou Yingming Bags, an SGS-validated site with 7 production lines and roughly 200,000 pcs per month of capacity, holding export experience since 2014, keeps the five records above as standing practice rather than as preparation, and works to 500 pieces per style and colourway with bulk in 35-50 days from approval and shipment on FOB Xiamen terms.

Alongside them, four checks run before shipment with attribute inspection at AQL 2.5 under ISO 2859-1: panels measured against ±1.5 mm, closures cycled about 200 times, hardware matched against the retained finish reference, and the batch link on the finished piece verified. The last one is what turns a traceability claim into a traceability answer.

What to Do Before the Next Questionnaire

Five things, in order, and none of them need a consultant. Write the delivery log with batch references. Link cutting to the batch. Link assembly to the panels. File the test reports with retention dates. Open the finding log with owners and closure dates. Each one is an afternoon of work.

Then test the chain with a real request: pick a finished piece, ask for its batch, and time the answer. If it takes longer than an hour, find out which link failed and fix that one rather than rebuilding all three. Most chains break at the same place, which is the join between cutting and assembly.

Finally, tell clients what you hold before they ask. A house that states its traceability response time, its testing cadence and its retention period in the first reply is answering three questions a client has not yet worked out how to phrase, and that is worth more than any certificate on the wall.

Frequently Asked Questions

People Also Ask