Lessons · Lesson 4 of 6
The approval you cannot move on your own
Which sample rounds a factory owns and which belong to the buyer, what a written eligibility rule looks like, and how to price the risk you take on when a screen replaces a parcel.
Lesson 4 of 6 · 18 min
Six styles, nineteen days, and nobody made a mistake
In the second month of the pilot, Bezirhan's technical manager sent an internal note. First fit is now judged in 3D, so fit samples for the tailored jacket programme are no longer to be couriered. It was a reasonable note. The virtual samples were good, the blocks were rebuilt, the fabrics were measured, and the whole point of the programme was to stop shipping parcels.
Wraysbury's garment technologist, Ruth Kelsall, had not been asked. She did what a technologist does with a style whose fit sample has not arrived. She left it open and moved on. Bezirhan's merchandisers saw six styles sitting unapproved and assumed she was busy.
Nineteen days. Six styles. Two of them missed a dyeing slot at Ravenglass Mills, which pushed their bulk fabric by 12 days, which was recovered by air at USD 2,910.
Every person in that story acted correctly on the information they had. The note was right about the technology and wrong about who owns the decision. State the rule in one line before the arithmetic. You may retire a round you own. A round somebody else owns is a change to an agreement, and it takes their signature.
Two kinds of round, and only one of them is yours
| Round | Who owns it | What it is really for | Can a factory retire it alone? |
|---|---|---|---|
| Internal proto | The factory | Proving a marker, a method, an operation sequence | Yes |
| First fit | The buyer | The buyer's judgement of proportion and fit | No |
| Revised fit | The buyer | Confirming the buyer's own comments were executed | No |
| Size set | Shared | Proving the grade across the range | Only by agreement |
| Lab dip and colour | The buyer | A physical standard two parties can hold | No, and it should stay physical |
| Pre-production sample | The buyer | An attestation about the LINE, not about the fit | No, and it should stay physical |
The last row is the one most often argued about and the least arguable. A pre-production sample is not a fit round wearing a different hat. It is evidence that this factory, on this line, with this bulk cloth and these operators, can make this garment. That is a statement about a production system, and no simulation of a garment can make it. The programmes that go wrong are usually the ones that treated the pre-production sample as one more round to digitise, because it looked like the others on the calendar.
What Wraysbury actually signed
Not "we will accept 3D". A one-page eligibility rule, applied style by style, which either fires or does not:
- The block is a carry-over from an earlier season, documented and in the digital set.
- Every shell fabric is characterised, from mill data or Bezirhan's own measurement.
- No new construction detail against last season's approved style in that block.
- The buyer's own point-of-measure list is used, unchanged, with no re-lettering.
If all four hold, the first fit round is judged on screen. If any one fails, that style ships a fit sample as before. The pre-production sample stays physical, always. Colour stays physical, always.
Of the 23 styles in the pilot, 14 passed all four, which is 60.9%. The other 39.1% shipped parcels, and it is worth being precise about why. That share was decided by the buyer's rule, not by the software's limits. The system could have simulated all 23. What it could not do is make a style a carry-over.
Who carries the risk when the screen was wrong
Somebody has to, and the clause is short. If a style approved on screen fails at the pre-production sample for a fit reason, Bezirhan carries the re-cut and Wraysbury carries the calendar.
Bezirhan then did what almost nobody does. It priced the clause before signing it. One of the first 14 failed that way, which is 7.1%, so the next season was priced at 8%, on a re-cut costing USD 1,240:
- 14 styles at 8% at USD 1,240 = USD 1,388.80 a season of expected exposure
- 14 fit rounds not shipped, at USD 155.58 each = USD 2,178.12 saved
The risk Bezirhan agreed to carry is 63.8% of the money it saved. That number should change how you talk about the whole programme. If you are moving approvals to a screen to save the cost of sample parcels, the arithmetic barely works, and one bad season makes it negative. The reason to do it is the calendar, and lesson 6 is about measuring that honestly rather than assuming it.
Check yourselfYour buyer's technologist says she is willing to try, but her manager wants no change to the contract. Is that a yes or a no?Show the answer
It is a yes to a bounded experiment and a no to a permanent policy, and the two need different paperwork. Write the eligibility rule as a trial covering a named list of styles for one season, with a review date and a clause on who carries a re-cut. That is not a contract change. It is a documented variation for a defined scope, and it is exactly the shape most buyers can approve without going near their legal team. Ask for the policy in season three, with the trial's own numbers on the page.
What to take away
- You may retire a round you own. A round the buyer owns changes what they are approving and needs their agreement in writing.
- A pre-production sample is an attestation about the line, not a fit round. It stays physical.
- Write an eligibility rule that either fires or does not, style by style. Bezirhan's four conditions passed 14 of 23 styles.
- The share that stayed physical was set by the buyer's rule, not by the software.
- Price the risk clause before you sign it. Bezirhan's expected exposure was 63.8% of the money it saved.
- If the case for moving an approval is the sample cost, the case is thin. The case is the calendar.