Cancellation, claims and what comes after

Wednesday 25 November 2026

A rescue that works does not make the loss disappear. It turns it into a bill, and somebody then has to decide who carries it. Two instincts arrive at that moment, and neither is right on its own. Recover everything the contract allows, or absorb it quietly to keep a good factory willing. This lesson works out three things instead, starting with how much of an unfinished order a buyer actually owns.

Everything landed. The first air tranche reached the warehouse on 20 October, the second on 22 October, the sea balance on 20 November. Every shop received its goods inside its own stated limit. Nobody cancelled and nobody claimed a discount.

Extra freight against the plan: USD 76,393. Gross profit actually earned USD 305,147 against a planned USD 381,540.

Notice what that means, because it is worth saying plainly and it surprises people. The revenue was fully protected, so the cost of the rescue is the entire loss. A recovery that works does not reduce the damage. It turns it into a freight bill. That is a good trade and it is still a bill, and somebody now has to decide who pays it.

Three questions, in this order. What could Beckwold have cancelled, and at what price. What can it claim. And what it should actually do.

What you own, and from when

What a purchase order commits you to is course 27.2's subject: the clause, the acceptance, the amendment. This is the other half. What physically exists in the world with your name on it, by date, and what it is worth to anybody else.

Read the third-party column, because that is the column that decides everything. Undyed yarn is an ordinary commodity and somebody else will buy it at a discount. The moment it is dyed to your four colours it is worth nothing to anyone but you. That happened on 2 June, three and a half months before a garment existed and eleven weeks before the first panel was knitted.

On a yarn-dyed knit you own the order long before you own anything that looks like a product. On a woven fabric dyed after weaving you own very little until the cloth is dyed, and on a style dyed as a finished garment you own almost nothing until the end. The cancellation ladder is set by where colour enters the chain, and that is chosen at design time rather than at crisis time. Course 8.6 reaches the same lever from the direction of minimum order quantities. Here it is about keeping your choices open, and the arithmetic is different.

Price it on this style. Made from greige yarn, which is yarn in its natural undyed state, and dyed as a finished garment, the liability would have stayed at USD 0.48 a piece until late September instead of jumping to USD 4.40 on 2 June. That is USD 94,080 of choice held open for fifteen extra weeks, at roughly USD 0.34 a piece more to make, or USD 8,160.

Is that worth buying? On this evidence, no. Beckwold has cancelled 2 of 61 knit orders in three years, and USD 8,160 a season to protect against a two-in-sixty-one event is a poor trade. Say that plainly rather than repeating the general advice. The general advice is that late colour buys flexibility. The specific answer here is that Beckwold does not use enough of it to pay for it.

What you can claim

Beckwold's purchase-order terms give it two things. One is a 2% late-delivery discount on the invoice value. The other is recovery of extra freight reasonably spent to reduce a delay the supplier caused, with the whole of it capped at 5% of the order value. Those are Beckwold's own terms, invented for this course. Yours will differ. The arithmetic below does not.

• The late-delivery discount: 2% of USD 302,400 is USD 6,048.
• The cap on everything: 5% of USD 302,400 is USD 15,120.
• The extra freight actually spent: USD 76,393.

Recoverable: USD 15,120, which is 19.8% of the loss.

That is the first honest finding and most people meet it as a shock. Your claim is capped and your exposure is not. A liability cap is not a fault in the contract, though. It is the reason the FOB price is USD 12.60. A supplier who carries unlimited liability for a customer's order book that they cannot see prices that risk into every garment, and you would pay it on every order rather than on the one that went wrong. You bought the cap. It is worth knowing what you bought.

What a claim destroys

Now the other side of the ledger, and this is where the question in the course title lives.

Turag makes about 78,000 pieces a year for Beckwold. Its FOB price is USD 1.80 a piece below the next approved alternative, which is USD 140,400 a year of advantage.

Turag's own profit on this order was USD 0.41 a piece, which is USD 9,840. Take the full USD 15,120 and the USD 6,048 discount and Turag ends up losing USD 11,328 on the largest programme it runs for you.

A supplier losing money on you does not resign. It moves you down the queue. Your order gets the second slot in January rather than the first, your sample room queue lengthens, and next season's quotation carries a number you cannot argue with because you cannot see inside it.

Beckwold has the evidence for that in its own history, and it is the mistake nobody made. In 2025 it raised a USD 22,000 chargeback, a deduction from an invoice to recover a supplier's cost, against a different knitwear supplier. It was correct, it was within the contract, it was properly documented and it was recovered in full. The following season that supplier's quotations moved USD 0.65 a piece, while the other three knitwear suppliers on the same yarn moved USD 0.18. On 96,000 pieces the difference is USD 45,120, a year. The chargeback was recovered once.

Nobody at that supplier ever said the two things were connected, and nobody ever will. What you can do is exactly what Beckwold did too late: compare one supplier's price movement against the others buying the same yarn in the same season. It is the only evidence available and it takes twenty minutes.

When you claim anyway

The answer is not "never claim". Four conditions, and each is a real one.

1. When not claiming gets the supplier's price wrong in your own books. A loss you never claimed is a subsidy you have hidden from yourself. It will be missing from the scorecard in course 27.5 as well, so next season's work goes to this supplier on economics that never existed.
2. When the failure is conduct rather than capacity: a hidden problem, undisclosed subcontracting, a falsified report. A capacity failure is forgivable and it recurs. A conduct failure tells you the reporting itself is unreliable, and everything you believe about that supplier rests on the reporting.
3. When somebody else requires it: an insurer, your own customer, an auditor. A loss quietly absorbed is a loss you cannot recover from anyone else either.
4. When you are leaving the supplier anyway, in which case the relationship is worth nothing and the arithmetic is just the claim.

The tool that beats both

Record the claim at its full value. Agree it. Then spend it on the cause instead of keeping the money.

The cause is on the record, and course 27.5 is where the record is built. This is the second miss in two years and both had the same cause: one nominated spinner, whose dye-house queue Beckwold does not control and cannot see. Two misses with one cause are not two misses. They are one cause nobody has fixed, and every claim raised against the symptom leaves it exactly where it was.

The fix is dull and cheap. Approve a second spinner and split the yarn 60/40. Approving one costs USD 4,200 once, for lab dips, a trial lot and an audit. A lab dip is a small sample of yarn or fabric dyed to check the shade before the bulk is dyed. The second spinner is USD 0.35 a kilo dearer on its share, and Beckwold buys about 24,960 kg of yarn a year through Turag, so the running cost is USD 3,494 a year.

USD 4,200 once and USD 3,494 a year buys out an exposure that cost USD 76,393 this season alone, and the agreed claim of USD 15,120 pays for the first year of it nearly twice over.

Then waive the USD 6,048 late-delivery discount, which Beckwold is plainly entitled to. Not out of sentiment. Taking it leaves Turag USD 11,328 down on the year, and waiving it buys Beckwold the second slot next January.

And write the waiver down at its full value. This is the part people skip, and it is the part that matters most. A waiver nobody records is not generosity. It is a hole in your own cost of ownership, which is what this supplier really costs you once every consequence is counted. It is precisely the hole that makes the next order look cheaper than it is.

There is the answer to the question this course exists to make askable. A claim is worth less than the supplier when the supplier's advantage, measured after the failure has been fully counted, is still positive. Here it is USD 0.94 a piece, or USD 73,320 a year. The right decision is to take the capped claim, spend it on a second spinner, and waive the discount. Record every one of those numbers, so that the next person to look can see what actually happened.

Settle it in one sentence, in writing, naming everything it covers: the freight, the discount, the delay, and any deduction arising from it. Course 7.5 shows what that sentence looks like from the other side of the table, and the reason it matters is the same in both chairs. Without it, the figure agreed in November and the debit note arriving in March are two separate payments.

Act as a category director who owns both the claim and the supplier relationship and has to live with next season's consequences. Help me decide what to claim on a late order and what to do with it. The failure: [WHAT HAPPENED], supplier [FACTORY], root cause as far as I know it [CAUSE], my direct loss [AMOUNT] broken into [FREIGHT / DISCOUNTS / CANCELLED ORDERS / OTHER]. My purchase-order terms: late-delivery discount [PERCENT], recoverable heads [LIST THEM], liability cap [PERCENT OR AMOUNT]. The relationship: [PIECES] a year at FOB [PRICE], next qualified alternative FOB [PRICE], this supplier's delivery record over the last [NUMBER] orders [SUMMARY], and the supplier's own margin on this order if I know it [AMOUNT]. Do the following. First, tell me exactly what is recoverable under my own terms and what is not, and give me the recoverable figure as an amount and as a percentage of my actual loss. Second, tell me whether this failure is a CAPACITY failure or a CONDUCT failure, and say which of the four conditions for claiming anyway apply here. Third, cost the relationship: what this supplier's price advantage is worth a year, and what my cost of ownership becomes once the failure is fully carried, including anything I waive, expressed per piece. Then tell me whether the remaining advantage is positive. Fourth, identify the ROOT CAUSE and price the fix — a second qualified source, a buffer, a changed nomination, a different point at which colour enters the chain — as a one-off and an annual figure, and compare it with the claim. Fifth, draft the settlement in one sentence that closes every head of loss at once. Sixth, tell me exactly what to record even if I waive it, and why. Be blunt if the answer is that I should claim and leave.

What you own at the end of it

A settled claim, a second spinner being approved, a delivery window your customers agreed to in writing, and a cost of ownership for this supplier that is honest to the cent. Only the last one changes anything next year.