Lessons · Lesson 3 of 6
Examination is a probability, not a disaster
Turn your broker's own examination history into a number you can plan with, and price the two decisions that really change your exposure.
Lesson 3 of 6 · 20 min
The eleventh consignment
2 April, Falkstead. A consolidated 40 ft container arrives. It carries eleven separate consignments from eight shippers. One of them is Marbeck Falkstead's 480 cartons of WJ-518. Consolidated means several customers' goods share one box.
Falkstead customs picks the container for a non-intrusive scan. That is an X-ray image of the sealed box. The scan finds nothing, as it almost always does. It costs the terminal's published scan charge, split between the eleven consignees, and Marbeck's share is small.
It costs everybody in the box two days.
Marbeck Falkstead's merchandiser writes it up as bad luck. It was not bad luck. It was the predictable result of a decision made in February about how to ship 480 cartons. This lesson is the arithmetic that decision needed.
Selection is not random, and it is not personal
Customs administrations do not open containers at random, and they do not open all of them. They use risk management. A consignment gets a score based on what is declared, who is declaring it, where it came from, how it moves and what the authority is looking for at the time. A small share is then picked for a check. The check has levels, and they differ enormously in time and money.
| Level | What happens | Corbray's charge | Days it adds |
|---|---|---|---|
| Documentary query | Customs asks a question about the entry; the broker answers from the file | No charge | 0.5 to 2 |
| Non-intrusive scan | The sealed container is imaged at the terminal | USD 145.00 | 1 to 2 |
| Partial devanning | A share of the cartons is unloaded, opened and restowed | USD 690.00 | 3 to 5 |
| Full devanning | The whole container is stripped, checked and repacked | USD 1,840.00 | 5 to 8 |
A merchandiser's instinct is to treat all four as one event — "we got stopped" — and to have no plan for any of them. But a documentary query answered from a complete file costs half a day and no money at all. A full devanning costs more than a week, and the cartons come back repacked by somebody who has never seen your packing standard.
Your broker already has your distribution
You do not have to estimate any of this. Your broker keeps a file per consignment with a timestamp on every event. The only reason nobody uses it is that nobody has asked for it as a table.
Corbray Brokerage cleared 96 consignments for Marbeck Vantoria in the year to March.
| Outcome | Consignments | Share | Charge each | Days added, midpoint |
|---|---|---|---|---|
| No intervention | 71 | 73.96% | USD 0.00 | 0.00 |
| Documentary query | 14 | 14.58% | USD 0.00 | 1.25 |
| Non-intrusive scan | 8 | 8.33% | USD 145.00 | 1.50 |
| Partial devanning | 2 | 2.08% | USD 690.00 | 4.00 |
| Full devanning | 1 | 1.04% | USD 1,840.00 | 6.50 |
Two numbers fall straight out of that table.
Expected charge per consignment. Over the year, the scans cost USD 1,160.00, the partial devannings USD 1,380.00 and the full devanning USD 1,840.00. That is USD 4,380.00 across 96 consignments, or USD 45.63 each. Put that number in a landed-cost build as the examination provision. It is about the price of a taxi ride. Examination is not expensive.
Expected days per consignment. Weighting the midpoints gives 44 consignment-days across 96, which is 0.46 days.
Now the trap. Neither number is the one to plan the calendar with. An expected value of 0.46 days describes a consignment that does not exist. 73.96% of shipments add nothing at all, and one shipment in a hundred adds six and a half days. Put 0.46 days in a calendar and you have budgeted for an event that never happens, in the wrong amount, on every single shipment.
Sort the 96 outcomes by days added and read the distribution instead. Ninety per cent of consignments add 1.50 days or less. Ninety-five per cent add 1.50 days or less too, because the curve is flat there. The last one per cent adds 6.50 days.
So a day and a half of allowance covers nineteen consignments in twenty. The twentieth needs a different kind of answer, and it is not a bigger allowance. It is a response you have prepared.
The decision you actually own
You cannot make customs less likely to select you. You can make being selected cost less, and there are exactly two levers.
Lever one: the quality of the entry. Fourteen of the twenty-five interventions were documentary queries. That means customs asked a question the entry did not answer. Every one of them is a document, a description or a value that did not line up. Lesson 4 is a whole lesson on what that costs.
Lever two: readiness. Corbray splits its own documentary queries by when the full document set reached it.
| Documents in the broker's hands | Queries | Median days to close |
|---|---|---|
| Five or more working days before arrival | 9 | 0.50 |
| After the vessel had arrived | 5 | 3.50 |
The same query, seven times the delay. Nothing about the query changed. What changed is whether the person answering had the file in front of them, or had to go and ask a factory in another time zone for a document that already existed.
A merchandiser owns that decision completely. It costs nothing. And it is worth more than any amount of arguing with a customs officer.
Eleven shippers' risk, in one box
Now back to Falkstead's 480 cartons.
When you ship less than a container load, your consignment travels inside somebody else's container alongside other people's goods. Your goods are only your own risk once they are in your own box. In a consolidated container, the whole box is stopped if any one consignment in it is stopped, and everybody shares the delay.
Take the intervention rate from Corbray's file: 26.04% of consignments get some intervention. If eleven consignments travel together and each is scored on its own, the chance that the container passes untouched is 73.96% multiplied by itself eleven times, which is 3.6%.
So the chance that a consolidated container of eleven consignments is stopped at least once is 96.4%.
Pricing the alternative
Ambergate Freight, Marbeck's forwarder, quoted the Falkstead leg both ways. Less-than-container-load freight is charged on whichever is greater, the weight or the volume. At 46.08 cubic metres against 3,888 kg, the volume wins. So the consignment pays for 46.08 revenue tons at Ambergate's quoted USD 38.00 each, which is USD 1,751.04. A part-loaded 40 ft container of its own, door to door, was quoted at USD 2,480.00.
The container costs USD 728.96 more. Here is what that buys:
- Four days. No deconsolidation at a container freight station. That step is what the consolidated route added between customs release on 6 April and availability on 10 April.
- A drop from 96.4% to 26.04% in the chance of an examination delay, because the box now holds only Marbeck's own declaration.
- Its own seal, so nobody else's cartons are opened next to yours.
Four days for USD 728.96 is USD 182.24 a day. Marbeck's own provision for a day past a floor-set date is USD 0.09 a unit a day, which on 5,760 jackets is USD 518.40 a day. The container is worth buying by a factor of nearly three. It was never considered, because the freight quote was compared against the freight quote.
Check yourselfYour consignment's own chance of intervention is 26.04% however it ships. Why does consolidating it with ten others make an examination delay far more likely?Show the answer
Because the customs authority selects the container, and any one of the eleven declarations inside it can trigger the selection. Your own probability has not changed, but you now sit behind eleven independent chances at once. The container goes untouched only if all eleven pass, which is 73.96% raised to the eleventh power, or 3.6%. The other side of that, 96.4%, is your real exposure to a delay. This is the one place in clearance planning where the arithmetic really is counter-intuitive, and it is why an LCL calendar needs a standing examination allowance where an FCL calendar needs a probability.
What to have ready before the vessel arrives
An examination is not a negotiation. Nothing you say changes the selection, and the officer in front of you did not make it. What shortens it is having the answer already assembled:
- The full document set in the broker's hands five working days before arrival, not on the day.
- A named person at the factory who can produce a photograph of a packed carton, a copy of the packing standard and the production records, on the same working day, with the time-zone difference already allowed for.
- Your own carton and shipping marks matching the packing list exactly, so a physical count adds up without anybody having to interpret anything.
- A ready answer to "what is this, physically" — composition, construction, end use — that nobody has to invent under pressure. Classification is track 12.2's subject, and the entry your broker filed is where it lives.
None of that lowers your chance of being selected. All of it turns a five-day intervention into a two-day one, on the few occasions when it matters.