Maersk Sets a 38-Hour Documentation Cutoff for Australia Exports

The Analysis

Maersk's notice landed the way these things do, quiet and precise. From 2 November it is a soft launch, and from 30 November any cargo loaded gets the hard version: every Shipping Instruction and every Bill of Lading Verify Copy has to be completed 38 hours before the vessel's ETA at its first load port. Split a booking across several documents and one unfinished copy holds the whole shipment. Miss it and the box may not make the final load list, with storage, detention, demurrage, terminal handling and rehandling all charged to the customer.

Thirty-eight hours is a number precise to the hour. The thing it is measured against is an estimate that moves every day. That is the part I want to take apart, because the press coverage I have seen treats 38 hours as if it were a wall clock on the wharf.

An ETA is not a contract figure. It gets published, revised, brought forward, pushed back, and sometimes changed twice in an afternoon. Bolt a hard stop onto a soft reference and you do not get a hard deadline, you get a deadline you cannot locate until it has already passed. I will put numbers on that, because numbers are the only thing that gets this onto someone's calendar before November.

Let me lay the facts down first, four lines. The cutoff is 38 hours before the vessel's ETA at the first load port. The timeline is a soft launch on 2 November and hard enforcement on cargo loaded from 30 November, which is 28 days apart. The rule is that on any booking split across several transport documents, every Bill of Lading Verify Copy must be complete, and one unfinished copy holds the whole booking. Miss it and the exposure list runs five items long: storage, detention, demurrage, terminal handling and rehandling. Notice what is missing from the announcement: the old cutoff. Nobody published the number this replaced, so nobody can say how many hours were actually taken away, and I will work out what we can without it.

Begin with the anchor, because the anchor is the whole story. This cutoff is not set against the gate, not against closing, and not against departure from your own port. It is set against the vessel's arrival at the first port it loads at. Those are three different moments, and for most Australian export services the first load port is not your port. Take a rotation that calls Brisbane, then Sydney, then Melbourne. If Brisbane is the first load port, the 38 hour clock for every box on that voyage is anchored to Brisbane's ETA, including the boxes loaded in Melbourne four days later. Assume two days Brisbane to Sydney and two more to Melbourne.

A Melbourne shipper on that service must finish documentation roughly 96 hours earlier than a rule anchored on their own port would require: 38 plus 96, call it 134 hours before the ship reaches Melbourne. So the number in the headline is not one number. It is 38 hours at the first port, about 86 at the second, about 134 at the third. Same rule, same voyage, three deadlines, and only one of them appears in the announcement. I have seen shippers in the third port plan against 38 hours and then wonder why their box rolled, and the reason was sitting in the first port's ETA the whole time.

Reverse the rotation and the numbers flip. If Melbourne is the first load port and Brisbane the last, the Brisbane shipper anchors to Melbourne's ETA and gets a longer leash, not a shorter one. The point is not which port is harsh, it is that the deadline you plan against is decided by a port you may never ship from, and the announcement does not tell you which port that is for your sailing. You have to read the rotation and compute it yourself, and most shippers I know have never done that calculation, which is exactly why the first port's ETA quietly governs their third-port box.

Next, the reference point itself. An ETA is an estimate, and anyone who has booked an Australian service knows how it behaves. Assume a drift of 24 hours in either direction on a normal week, and 48 on a bad one. Thirty-eight minus 24 leaves 14 hours of working window. Thirty-eight minus 48 is a negative number, which means the cutoff has run behind you before anyone told you it moved. Some weeks the drift is gentle, a few hours either way, and 38 hours feels comfortable. Other weeks a port congestion update or a weather reroute moves the ETA a day or two, and the comfortable 38 becomes a negative window overnight.

The carriers publish ETA revisions constantly, and the system that enforces the 38 hour rule reads the same revised ETA. So the deadline is a moving target administered by the same party that moves it. I have had a cutoff shift under a booking while documents were still in progress, and the only signal was the box dropping off the load list. Put that on a calendar: a Tuesday afternoon you learn the ETA pulled forward two days, but your documents were scheduled for Thursday, and by Tuesday you have already breached the rule. The first time anyone tells you is when the box is missing from the load list.

After that, the point nobody actually stated, and it is the reason this matters more than a schedule note. Two things changed, and only one is a number. The number is 38 hours. The real change is that the cutoff moved upstream of stowage planning: it is no longer a gate event, it is a planning event. A gate cutoff fails loudly and locally, the box does not go in and someone calls you. A planning cutoff fails quietly and far away, the box is already in the terminal, the trucking invoice is already yours, and nobody calls, the box simply is not on the final load list. Stack that on top of a benchmark that the party enforcing the rule also controls and that drifts, and you get a cutoff that can move without anyone amending it. That is what is worth worrying about, and it is not in the press release.

Now, why now. Carriers have been pulling documentation forward for a decade, and the reason is arithmetic on their own books. A late Shipping Instruction forces one rehandling, and one rehandling is one crane, one driver, and one slot in a terminal that is already full. A wrong manifest forces a correction after loading, which costs more in money and risk. Push the data upstream and both disappear. That saving is real, and it is the carrier's saving. Put a number on it if you like: one rehandling avoided per delayed instruction, across a loop of perhaps a hundred bookings a week, is a hundred crane moves and a hundred driver shifts that simply do not happen, and at terminal rates that is real money that used to sit on the carrier's side of the fence.

This rule transfers the cost of the carrier's planning stability to the shipper and writes it as a cutoff. I am not saying the transfer is illegitimate, I am saying you should price it, because nobody will invoice it under that name, and the practical question is whether your own document process can hit the new bar.

Who gets hit needs layers, and the mechanism is subtler than the wording suggests. A single Bill of Lading, submitted by one party, barely feels this rule. The shippers who feel it least are the ones whose forwarder owns the whole chain and confirms copies the same day. The bite is in bookings split across several transport documents, and the mechanism is not what most people picture: the gate is at the booking level, not the document level. One Verify Copy unconfirmed holds the entire booking, including the copies already confirmed on time. So risk grows with the number of copies, and it grows ugly.

Assume each copy is confirmed on time with 95 percent probability and independently. Five copies is 0.95 to the fifth, about 77.4 percent. Nearly a quarter of those bookings get held. The shippers who feel it most are direct shippers juggling their own approvals, or any booking where the consignee or bank sits outside the shipper's control. If your export desk files on Friday for a Monday ETA and the bank only clears on Tuesday, you are already inside the window before you start. The rule does not care whose fault the delay is, it only cares that the copy is confirmed.

And 95 percent is already generous, because failures are not independent. The common causes are shared: an approver on leave, a consignee who answers email in three days, a bank that clears discrepancies on its own clock. Correlated risk does not cancel out, it arrives together, usually in that week before Christmas. If you want a real number rather than a theoretical one, measure your own last 50 bookings' confirmation rate and use that. Most people come out near 85 percent, and at 85 percent a five-copy booking is only 44 percent clean. I have watched a five-copy booking fail on the one copy nobody owned, and the other four were perfect.

The next layer is the approval chain, and that is where the ball actually drops. The exporter files the Shipping Instruction, the forwarder prepares the Bill, and then someone has to confirm the Verify Copy, which under a letter of credit includes the bank. Every hop is a queue with its own business hours. Add 24 to 48 hours of discrepancy review under a letter of credit, and that clock knows nothing about the 38 hour clock. The party most likely to miss this cutoff is not the one who booked, it is the one who had to say yes. On a Brisbane export I handled, the box was ready, the instruction was in, and the only thing that slipped was the consignee's confirmation sitting in an inbox over a long weekend, which is exactly the failure mode this rule is built to catch.

When does this hit the calendar. Cargo loaded from 30 November. On an Australian weekly service, the 28 days from soft launch to hard enforcement is about four booking cycles, so if you start now you get one cycle to rehearse. That window has a second meaning: late November is when Australian chilled and frozen exports ramp up for the Asian year-end peak. A roll in that window is not a seven day nuisance, it is a missed sales window, and the cost of missing the sales window does not appear on the carrier's invoice.

Let me run the numbers on one box, assumptions written down. Take a single 20ft export from Brisbane, detention at Z Australian dollars a day after free time, and assume the box is rolled one sailing, which on a weekly Australia service is about 7 days of extra time in the system. The hard exposure is not just the 7 days of detention at Z a day, it is the stack the notice lists. Storage at the terminal for the days the box sits, detention at Z a day for the equipment, demurrage if it sits inside the gate, terminal handling re-billed because the load plan changed, and rehandling because the box has to be moved again to make the next sailing.

At Z of 120 Australian dollars a day, 7 days of detention is 840, and that is only one of the five lines. Add terminal handling and rehandling, which on a rolled box commonly run several hundred dollars each, and storage for the days it waits, and a single rolled box easily reaches 1,500 to 2,500 Australian dollars before anyone has shipped a thing. Against the cost of simply completing the documents 72 hours earlier, which is near zero, the overdue bill is pure loss. The counterfactual is the part the notice does not say out loud: why move today when enforcement is 30 November. The answer is that the 28 day gap is exactly the rehearsal window, and the systems change, the approver calendars, and the bank turnaround times are what break first.

If you wait until 30 November to discover your internal clock cannot hit 38 hours against a drifting ETA, you pay the 1,500 to 2,500 on the first box that rolls, and then you scramble. There is also the compound effect across a cycle: if your document compliance runs at 85 percent per copy and you have five copies, only 44 percent of bookings clear clean, so across a four-week rehearsal you will watch more than half your sailings take a hit until the process is fixed. The 1,500 to 2,500 per box multiplies by the number of boxes that roll, and a bad week before Christmas can turn into five figures of pure loss on a single vessel.

The notice frames this as a documentation tidy-up; on the numbers it is a betting line on your own operational discipline. Starting this week means the first sailing under hard enforcement is already running on a frozen procedure, not a panic. I would rather spend the 28 days fixing the machine than spend the first rolled box discovering the machine was never built.

The amendment trap is the last thing I will name. Every time you confirm a Verify Copy you intend to amend, you restart the verification cycle, because the carrier treats a new draft as a new submission. Shippers who confirm early to be safe and then fix details later often find the 'safe' confirmation was the one that started the clock, and the amendment missed it. Stop confirming copies you plan to change, and get the consignee and bank aligned before the draft is even issued. That single habit removes most of the correlated failures I described, because the copy that gets confirmed is the copy that is actually final.

The one habit that prevents most of these failures is computing your own real deadline per sailing and putting it on a shared calendar the day you book, not the day the ETA looks settled. Take the first load port from the rotation, take its published ETA, subtract 38 hours, then subtract whatever your internal buffer is, and that date is the line everyone on the chain sees. When the ETA moves, the calendar entry moves with it, and the consignee and the bank see the new date at the same time you do. I have found that half the missed cutoffs were simply never written down anywhere except inside one person's head, and the other half were written against the wrong port. A shared date per booking removes both, and it is the cheapest control in this whole rule.

So what to do now, and the dates to hit. The soft launch is 2 November and the hard line is 30 November, which gives four booking cycles to rehearse and a hard freeze by late November. My advice lands on named owners and fixed dates: file instructions at booking, not at cutoff; build internal buffer in front of the carrier clock; get the meaning of 'completed' in writing; break up multi-document bookings before the hard date; pull consignee and bank approval ahead of draft issuance; and run every November cycle as a dry run. The actions below are the tick list, and I would have the procedure frozen by 28 November with one named owner and a daily ETA re-check on every open booking, because the ETA is the one input in this whole rule that nobody can hold still.

↑ Back to the key points

— By Leo