Lessons · Lesson 5 of 6
Agreeing in advance what a screen decides
The document that turns a virtual sample from an extra step into a replaced round, and the price of every decision it fails to name.
Lesson 5 of 6 · 18 min
The comment sheet that was never written
Go back to the first week of this course. Sanna Vollmar opened a virtual sample of KA-7318 and wrote nothing on it.
She was not being obstructive. She had no idea whether an opinion she gave on a screen would still be hers in week fourteen, whether her director would honour it, whether the factory would treat it as an approval, or what would happen if she changed her mind when cloth arrived. Faced with that, any sensible person waits for the garment.
A virtual sample that decides nothing is an extra step. Nothing about the software determines which of those two it is. A document does, and it has to exist before the season starts.
The matrix
It is one page. Kalthoum and Aldemere wrote it in a ninety-minute call, and it is the highest-return ninety minutes in this course.
| Decision | Decided on | Who signs | What reopens it |
|---|---|---|---|
| Proportion, silhouette, panel breaks | Screen | Aldemere design and the buying director together | A pattern change altering any length or width beyond an agreed threshold |
| Pocket, placket, label and trim placement | Screen | Aldemere design | A pattern change in the panel that carries it |
| Print scale and placement | Screen, with the mill's own repeat entered | Aldemere design | A change of mill or of repeat |
| How the garment hangs and where the hem breaks | Screen, only against a measured record of the buying cloth | Aldemere design | A change of cloth, of weight, or of finishing route |
| Fit on the body | Screen for intent, cloth for sign-off | Aldemere technologist | Any graded measurement change |
| Colour | Cloth. Never screen | Aldemere colour | Reopens on any dye-lot change |
| Hand and surface | Cloth | Aldemere design | Reopens on any cloth change |
| Seam appearance and construction | Cloth, at the pre-production sample | Kalthoum quality and Aldemere technologist | Any operation or machine change |
| Grade across the range | Cloth, at the size set | Aldemere technologist | Any block change |
Four things make that table work. A matrix missing any of them is decoration.
It names what is being decided, not what is being looked at. "Review the 3D" is not a decision. "Proportion and silhouette are settled" is.
It names the evidence. The row about how the garment hangs carries this course's whole argument: it is a screen decision only against a measured record of the cloth being bought. Without that clause the row is a lie waiting to happen, and lesson 6 is what happens.
It names a person. Not a company, not a department. The individual whose approval binds. This is the row Kalthoum got wrong, and it cost more than the programme saved.
It names what reopens it. An approval is not permanent. It is conditional on the thing it was given about. Writing the condition down is what stops a reopening from becoming a fight.
The mistake nobody made
Sanna approved the proportion of three styles on screen in week three. She was right to. She was the person the factory had been dealing with, and the proportions were good.
Margit Sablon, Aldemere's buying director, saw the physical pre-production samples in week fourteen. On three styles she said the skirt read short against the rest of the range.
She was also right. She is paid to see a range together rather than a style alone, she had not been in the screen reviews, and she had never agreed to be bound by one. Nobody in this story did anything wrong.
The bill:
- Three styles reopened at the pre-production stage: three rounds at USD 87.63, or USD 262.89, plus nine days each.
- Two of those three were tight-calendar styles from lesson 4. Nine days put 2,700 pieces back on the plane at USD 1.94: USD 5,238.00.
- USD 5,500.89 in total.
The whole programme's gross money saving across eleven styles was USD 2,546.50. The reopened decisions cost 2.16 times everything the programme saved, and the correction was free: put the person whose approval actually binds into the screen review, and write her name in the matrix.
The other way a programme fails, and it is the commonest one
On six of the eleven styles, Kalthoum's first season ran the virtual round and all five physical rounds.
Nobody decided this. It happened because the matrix arrived late, the buyer's calendar had already been published with five rounds in it, and no individual felt able to be the one who cancelled a garment. Six styles at USD 117.30 of digital work is USD 703.80 of pure addition: real cost, no round removed, no day saved.
This is what most 3D programmes look like in their first two seasons, and it is why they are reported as expensive. The software did not fail. A round was added and none was taken away, which is not a technology outcome at all. It is a calendar that nobody rewrote.
If you take one operational instruction from this course: when the matrix is signed, go and delete the round from the published development calendar in the same week. A round that is still in the calendar will be made.
Colour is the row people argue about, and they should not
Every season somebody asks why colour cannot be approved on screen when the render looks so convincing.
Because a colour on a display is defined only in relation to a specified display and a specified viewing condition. Change the panel, the brightness or the room, and the same file is a different colour. Cloth is judged against a physical standard in a controlled light, because that is the only way two people in two countries can be looking at the same thing.
Screens have a real job in colour work. They are excellent for placement, for the proportion of a colour within a print, and for seeing a range together. None of that is approving a shade. Keep the row and lose the argument.
Check yourselfYour buyer's merchandiser approves a virtual sample by email. What do you need before you cancel the physical prototype?Show the answer
Four things, and they are all in the matrix. That the decision being approved is named: proportion and placement, not "the 3D". That the evidence supports it, so if the approval covers how it hangs, the fabric record came from the cloth being bought. That the person approving is the person who can overturn it, or that the person who can was in the review. And that you both know what reopens it. If all four hold, cancel the round in the calendar the same day. If any one of them does not, you have a very useful preview and you still owe somebody a garment.
Prompt · Draft the approval matrix before the season opens
In the planning call before a range starts, or the week after a screen approval was overturned by somebody who had not been in the review.
Act as a technologist who has seen digital approvals fall apart and knows exactly where. Help me draft a one-page approval matrix between my factory and my buyer, covering which decisions are made on screen and which wait for cloth. My situation: [FACTORY, BUYER, PRODUCT TYPE, NUMBER OF STYLES IN THE RANGE, HOW MANY SEASONS WE HAVE WORKED TOGETHER]. My 3D maturity: [HOW MANY STYLES PUT THROUGH, WHETHER FABRICS ARE MECHANICALLY CHARACTERISED AND HOW, WHETHER THE AVATAR MATCHES THE BUYER'S SIZE SPECIFICATION, WHETHER SCREEN JUDGEMENTS HAVE EVER BEEN CHECKED AGAINST THE CLOTH THAT FOLLOWED]. The people: [EVERY PERSON ON THE BUYER'S SIDE WHO CAN APPROVE OR OVERTURN A DECISION, WITH JOB TITLE, AND WHO ACTUALLY ATTENDS DEVELOPMENT REVIEWS]. Do the following. First, produce the matrix as a table with four columns: the decision, where it is decided, who signs it by name, and what reopens it. Cover at least proportion and silhouette, placement of pockets and plackets and labels, print scale and placement, how the garment hangs, fit, colour, hand, seam appearance and construction, and grade. Second, for every row you have put on the screen, state the evidence that row depends on, and where it depends on a fabric record say explicitly that the record must come from the cloth being bought. Third, look at my list of people and tell me which rows are exposed because the person who can overturn the decision does not attend the review — that is the failure that costs more than the programme saves. Fourth, tell me which physical rounds this matrix lets me delete from the published development calendar, and warn me that if I do not delete them in the same week they will be made anyway. Fifth, list what you would refuse to put on a screen at my stated maturity, and what would have to change for each. Keep colour in cloth and explain why in one sentence.
AI can make mistakes — check anything you act on.
What to take away
- A virtual sample decides nothing until somebody has written down what it is allowed to decide. Without that document it is an extra step, and the software cannot supply it.
- The matrix has four columns that matter: the decision, the evidence required, the person who signs, and what reopens it.
- The row about how a garment hangs is a screen decision only against a measured record of the cloth you are buying. Everything else in this course depends on that clause.
- Kalthoum's reopened decisions cost 2.16 times everything the programme saved, and every person involved acted correctly. The fix was to put the binding approver in the room.
- On six of eleven styles the virtual round was added and no physical round removed, at USD 703.80 of pure cost. Delete the round from the published calendar in the same week the matrix is signed.
- Colour is never a screen approval, because a colour on a display is defined only against a specified display and viewing condition. Screens are for placement and proportion, not shade.