Lessons · Lesson 6 of 6
The file, and what would have to move
How to write an origin decision down so it can be re-run, which of its inputs is fragile, and why changing the destination changes the answer more than anything else.
Lesson 6 of 6 · 16 min
A decision nobody can reconstruct is a decision that gets re-argued
The output of this work is not a country. It is a short file that lets somebody else — or you, next February — reach the same answer, or a different one, on purpose.
It has one property that matters more than its contents: every input carries a date and says where it came from. Most of the mistakes in origin comparison are not arithmetic. They are a freight rate from a quotation that expired in November sitting in a sheet dated March, a lead time somebody typed in from a factory's website, and an exchange rate nobody can source at all.
So separate the two kinds of input, because they go stale in different ways:
- Quotes expire. An FOB, a freight rate, a mill's price. They have a validity date printed on them, and you should copy it across.
- Measurements drift. A lead time, a transit time, a full-price sell-through, a shade minimum. A quoted lead time is not a measurement; it is a hope. Measure the last three deliveries from that factory and use what you find.
The commonest error in the whole exercise is putting a quoted lead time in the measurement column. That is how a comparison ends up rewarding whichever factory is most optimistic on the telephone.
Solve for the move that changes the answer
A ranking is more useful when you also know how hard it is to overturn. Egypt beats Bangladesh on the polo by 0.1071 a unit. How far is that?
Bangladesh's landed cost moves by 1.1276 for every 1.00 that its FOB or its freight moves, because duty at 12.0% rides on top and the cash column rides on both. So the flip needs a fall of 0.0950 a unit — which is almost exactly half the 0.19 Osbern pays today to move a polo from Bangladesh. Ocean freight rates halve. They have done it inside a single year more than once.
Or take it from the other side: Egypt's FOB rising by 0.1069, to 4.7269, does the same thing. That is 2.3% on a quotation, which is a normal outcome of a cotton move or an annual wage settlement.
Now compare that with the fashion order. Turkey wins FW-812 unless the markdown gets shallower than 19.66%, and how deep a printed blouse is marked down is a property of the category and the retailer, not a price that moves on a Tuesday.
Change the destination and the map changes
Everything in this course has been priced into Lyddenmark. Osbern also supplies Ostrany, whose schedule is different: a general rate of 9.6%, and duty-free access granted to a different set of origins. Same four factories, same four FOBs, same polo.
| Bangladesh | Vietnam | Egypt | Turkey | |
|---|---|---|---|---|
| FOB a unit | 4.10 | 4.28 | 4.62 | 5.05 |
| Freight and insurance | 0.21 | 0.23 | 0.16 | 0.14 |
| Customs value | 4.31 | 4.51 | 4.78 | 5.19 |
| Rate on Ostrany's schedule | duty-free | duty-free | 9.6% | 9.6% |
| Duty a unit | 0 | 0 | 0.4589 | 0.4982 |
| Cash in transit a unit | 0.0276 | 0.0311 | 0.0165 | 0.0154 |
| Landed a unit | 4.3376 | 4.5411 | 5.2554 | 5.7036 |
Into Lyddenmark the order was 4.7305, 4.8376, 5.0777, 5.1263 and Egypt won. Into Ostrany it is Bangladesh, Vietnam, Egypt, Turkey, and the top of the table and the bottom have changed places.
Nothing about any factory changed. The destination did. An origin is not good or bad; it is good or bad into a market, for a product, at a quantity, this season. That is why a company-wide approved-origin list is a decision to be wrong somewhere, and why this course names six countries and argues for none of them.
The things that cannot be computed, and must not be blank
Three questions decide origin as firmly as any of the arithmetic above, and none of them has a number:
- Can you actually get the booking? Space and equipment on a route are not guaranteed by a rate.
- Will the factory's compliance status hold through the whole production window, rather than through the day you checked it?
- Will the rule still be there? Trade arrangements are renewed, amended and allowed to lapse, and the notice is rarely long.
Write each one into the file as an open question with the name of the person who can answer it. Do not leave the cell empty, and do not write a zero. An empty cell is read as nothing to worry about, and a zero is a claim you have not earned. It is the same discipline as refusing to call an origin "green" because nobody has yet reported a problem with it.
Check yourselfA colleague says the polo comparison should be redone every year at budget time. What does the fragility analysis say instead?Show the answer
That the interval is set by the input that moves, not by the finance calendar. The polo answer turns over on a 0.0950 move in one freight rate — roughly half of today's — so it should be re-run whenever the freight market moves materially, which may be twice in one year and not at all in the next. Reviewing it every January and ignoring it in between guarantees both a wasted review and a missed one. The blouse, whose answer needs the category's markdown depth to change, does not need an annual review at all.
The rest of this track is the detail this lesson keeps pointing at. Courses 26.2 and 26.3 take the real preference conditions for two large markets. Course 26.4 takes the customs and export procedure that turns a claim into an accepted one. Course 26.5 takes the banking and the incentives. Course 26.6 puts the origins side by side as competitors rather than as candidates on one buyer's list.