Two routes, one promise: choosing by the tail, not the average

The same box, two ways to Mississauga

The three containers of PO BM-4412 are going to a distribution centre outside Toronto. Thornbeck offers two routings and quotes both.

Route B is five days quicker on this table, and USD 310 a container cheaper. Across three containers that is USD 930. On the quotation it is not a close decision.

Ex-factory is Sunday 6 September. Brantmere needs the goods at Mississauga by Friday 23 October, which is day 47. Both routings clear it with room. The merchandiser takes Route B.

That is the mistake. And there is nothing wrong with any number on that table.

An average is not a promise

A transit time on a quotation is a typical duration. What you give your buyer is a date. A date is kept or missed. There is no partial credit for arriving four days late on average.

So ask the only question that connects the two: out of your last shipments on this routing, how many arrived by day 47? Thornbeck keeps the record and hands it over.

Route B is 4.2 days faster on average. It misses the date nearly three times as often: four shipments in thirty against eight. The mean got better and the promise got worse. Only one of those two numbers was on the quotation.

Where a tail comes from

A tail is not bad luck. It is made of connections.

Route A has one transhipment. If the box misses its mainline connection at Colombo, it waits for the next sailing. On this service that is about a week, and that is the whole of Route A's downside. Its slowest shipment in thirty was 52 days, eight days over the median.

Route B has a transhipment and a rail connection, one after the other. The Toronto ramp is served three times a week, not daily, so a box that misses its slot loses two or three days before it has moved at all. The rail leg crosses mountain passes that close for weather in winter. And a container coming off a ship at Vancouver competes for rail capacity with every other container on that ship. Three separate queues, each with its own frequency. They multiply rather than add. Route B's slowest shipment was 61 days, twenty-two days over its own median.

Every connection you add gives the routing another way to be late. The frequency of the connecting leg sets how late.

Pricing the difference

The sea shipment carries everything except the 200 trouser cartons that lesson 3 put into a consolidation. That is USD 616,000 of invoice value. PO BM-4412's late-delivery clause is 10%, so being late is worth USD 61,600.

Route B saves USD 930 of freight. It costs USD 9,336.67 more once lateness is priced. The cheaper, faster routing is worse by about ten times the saving that made it look attractive.

One document, several carriers, and no convention

Route B has an inland leg inside it. That changes the paperwork as well as the clock.

Thornbeck offers it under a single multimodal transport document covering sea and rail from Chattogram to the Toronto ramp. That is convenient, and it is worth knowing what is underneath it. Carriage by sea has its own international conventions. Carriage by air has one. Carriage by road and by rail have regional ones. There is no international convention in force governing multimodal carriage as a whole. One was concluded at Geneva in 1980 and never came into force. What fills the gap is contract: a set of standard rules the trade has agreed to adopt, written into the document by its own terms.

The practical consequence is one sentence long. On a through document, what you can recover depends on the clause in that document, and on which leg the goods were on when the loss happened. An inland leg may carry a much lower limit than the sea leg beside it. So when you add a rail or road leg to a routing, read the liability clause of the document that covers it before you sign the booking, not after a claim. Course 12.4 owns the difference between a forwarder acting as agent and as carrier. Course 12.8 owns insurance and claims. This lesson only insists on one thing: adding a leg is a legal decision as well as a scheduling one.

And a different clearance

The two routings also clear customs in different places. Route A's containers are declared and released at the port of arrival. Route B's move inland from Vancouver under a transit procedure and are declared at an inland office. That needs a bond, a different instruction to the broker, and a different set of dates in your file. Course 12.3 owns import clearance itself. Note here only that a routing decision quietly changes who clears your goods, where, and on what timetable. The person who chose the routing usually did not know that.

Act as a logistics planner who has been burned by transit estimates. I want a door-to-door clock for one shipment, built leg by leg, and an honest read on where the risk is. Facts: origin [PLACE], destination delivery point [PLACE], goods ready date [DATE], the date the consignee must have the goods [DATE], mode or modes offered [LIST], and for each routing the forwarder's leg-by-leg durations if they gave them. Working week at origin [DAYS], at destination [DAYS], and any public holidays in the window [LIST]. Value of the shipment [AMOUNT], and the penalty for missing the consignee's date [CLAUSE OR AMOUNT]. Do the following. First, lay out every leg with its duration, and say which are transit and which are dwell, queue or waiting. Second, for EVERY leg give me the FREQUENCY of the service: daily, three times a week, weekly, fortnightly. Next to it, put what one missed cut-off on that leg costs in days, because that is the interval, not the transit. Third, list every cut-off I own with the hour it falls, including the documentary and advance-filing ones, and mark the earliest. That is my real ship date. Fourth, tell me which legs connect to each other, and where a missed connection on a frequent leg can spend a whole interval of an infrequent one. Fifth, ask me for the forwarder's own arrival record on these routings: fastest, median, ninetieth percentile, slowest, and how many of the last thirty landed inside my required day. If I do not have it, tell me to go and get it before deciding, and say why the mean will not answer my question. Sixth, price being late using my penalty and each routing's hit rate, and put that next to the freight difference in one table. Do not present an average transit as a delivery date.

What to take away

1. A transit time is an average. A delivery date is a promise. Choose the routing on the distribution, not the mean.
2. Ask for the forwarder's own arrival record, and count how many landed inside your required day.
3. Tails are made of connections. Every extra connection adds a whole service interval to the downside.
4. Price lateness explicitly and put it next to the freight saving. A USD 930 saving does not survive a USD 61,600 exposure.
5. Adding an inland leg changes the liability regime and the place of clearance. Read the through document's clause, and tell your broker before the box moves.