Ocean Alliance Extends Europe-Far East Service Network to Gdansk, Poland

The Analysis

The Ocean Alliance, which is CMA CGM, COSCO Shipping, OOCL and Evergreen, has extended its Europe to Far East service network to the Port of Gdansk in Poland. The change adds direct deep-sea calls into the Baltic. For importers in Poland, the Czech Republic and Slovakia, the alliance puts the transit time improvement at 5 to 7 days against the current pattern of transshipping through Hamburg or Rotterdam.

That is the entire announcement, and it contains exactly one number worth arguing about. The number is 5 to 7 days. The rest is context.

Most of what has been written about this treats the 5 to 7 days as a fact about the future. It is not. It is an estimate, issued by the party selling the service, measured against a specific alternative, on a specific set of lanes, before a single box has moved on the new rotation. All three of those qualifiers matter, and each one can move the answer enough to change whether you should rebuild your routing.

Start with what the 5 to 7 days is measured against. It is measured against transshipment through Hamburg or Rotterdam, not against another direct call. That distinction decides everything that follows. If you already have a direct service into the Baltic from another carrier, this announcement changes nothing for you, because your base is not a transshipment routing. If you are transshipping, the comparison is real and the saving is real, but it is a saving against your current design, not against the market. When a number is quoted against your own current practice, the first question is whether your current practice was ever the right one. Some shippers are transshipping through the North Range because it is genuinely cheaper for their door, and some are doing it because that is how the booking was set up five years ago and nobody has revisited it.

The second qualifier is who produced the estimate. A carrier putting a number on its own new service is not lying, and there is no reason to assume bad faith. There is every reason to check the base, because the seller chooses which alternative to compare against and which lane to quote. The most favourable comparison for a new direct call is against the slowest transshipment routing in the market. The least favourable is against a well-run feeder connection through a hub where your cargo already clears fast. The published 5 to 7 days sits somewhere between those two, and you do not know where.

The third qualifier is the range itself, and this is the part I find most useful. A range of 5 to 7 days is not sloppiness, it is an admission that the answer depends on the inland origin or destination. If the midpoint is six days and the range spans two, the spread around the midpoint is a third of the value. That is a wide band for a decision about rebuilding a supply chain. If your specific lane sits at the 5 end, the annualised case is a sixth smaller than the headline suggests, and if it sits at the 7 end it is a sixth larger. Before anything else, find out which end you are on.

Now the money, because that is where this either justifies the disruption or does not. The saving lands in three separate buckets and most people only count the first. Bucket one is the time value of the goods in the pipeline. Assume an importer moving 40 FEU a month with an average cargo value of $25,000 per FEU, so about $1 million of goods a month. Six days less transit removes roughly $200,000 of goods from the pipeline at any given moment, because $1 million spread across 30 days is about $33,000 a day and six days of that is $200,000. At an assumed 6% annual financing cost, that is about $12,000 a year of working capital released, per trade, per shipper. Every figure in that sentence is an assumption I have stated, so put your own numbers in before you rely on the output.

Bucket two is the transshipment cost that disappears rather than moves. A hub transshipment carries terminal handling at the hub, a documentation charge, and usually a feeder leg. Assume $200 a box for the handling and documentation combined, which is an assumption and not a published figure. On a 40 FEU monthly programme that is $8,000 a month, or $96,000 a year. Compare that with the $12,000 from bucket one. The cost bucket is eight times the capital bucket, and it is the one people forget, because it arrives as a line item buried in a forwarder invoice rather than as a number in a treasury report.

Bucket three is the one nobody puts in the business case, and it is usually the largest. A direct call removes a handoff. Handoffs are where the tail of your arrival distribution lives, not the mean. A transshipment routing with a six-day mean might carry a standard deviation of four or five days, because the connection can miss. A direct call with the same mean carries a much tighter distribution, because there is no connection to miss. Safety stock is priced off the tail, not the mean, so a reduction in variance is worth money even when the average transit time does not move at all.

Assume your daily cost of goods is $33,000 and the direct call cuts the standard deviation of arrival by two days. At a service level factor of 1.65, the safety stock release is roughly $110,000 of inventory, and at an assumed 25% annual carrying cost that is about $27,000 a year. That is more than both other buckets combined, and it appears in no press release anywhere.

Which bucket dominates depends entirely on what you ship, and this is where a generic reading of the announcement fails. High value density cargo, meaning a lot of money per cubic metre, is dominated by bucket one, because the capital tied up per box is large relative to the freight. Low value density cargo, furniture, building materials, basic components, is dominated by bucket two, because the handling cost is fixed per box while the goods are cheap. Anything with a fixed shelf date or a seasonal window is dominated by bucket three, because for that cargo the tail is not a cost, it is a write-off. Three shippers reading the same announcement should reach three different conclusions, and any analysis that reaches one conclusion for everybody has not done the work.

Why this is happening now has two parts and neither is generous. Demand in Central and Eastern Europe has grown to the point where a direct call can fill itself, which is the precondition for any new string. The North Range hubs have become expensive and congested enough that shippers are actively shopping for alternatives, which is the demand signal. The alliance is not running a direct call to Gdansk as a public service. It is running it because the corridor can now support a string and because whoever puts the first direct calls on a corridor takes the accounts that come with it. Read the service announcement as a competitive move wearing a timetable.

What the alliance is actually buying with this is not the 5 to 7 days. It is the account. Once an importer rebuilds its inland contracts, its customs filings and its warehouse slotting around a Gdansk arrival, the switching cost to move back is real, and it is measured in months rather than in dollars per box. The 5 to 7 days is the marketing for that lock-in. That is not a criticism, it is how every network change works, and it is the reason you should treat the first two rotations as a trial rather than as a migration.

On who benefits, the split is clean. Polish importers benefit directly and most, because the port is in their market and the inland leg is short and contestable. Czech and Slovak importers benefit through the inland leg out of Gdansk, which is real but conditional: it depends on rail and road capacity from the port, and on how the inland transit from Gdansk compares with the inland transit from Hamburg for their specific factory, not for their country. Baltic importers get an option they did not previously have, but their volumes are thin, so the constraint on them is service frequency rather than transit time, and frequency is what gets cut first when a string underperforms. German importers notice nothing, because the North Range calls are unchanged. Hamburg and Rotterdam lose a slice of transshipment volume that is small in absolute terms and is exactly the kind of slow drip that shows up in a hub port's statistics a year later.

Timing is where the announcement misleads by omission. A network extension is a schedule change, so it lands with the next rotation rather than phasing in gradually. That means the useful question is not when the alliance announced it. The useful question is which day of the week the vessel calls at Gdansk, because the call day sets your documentation cutoff and your inland booking window, and those two clocks govern whether the six days survives contact with your warehouse. Get the port rotation and the cutoff times from your carrier in writing before you rebuild a single arrival forecast. A forecast built on an assumed call day is a forecast with a hidden error term.

There is a planning consequence that follows from this and it catches people out every time a schedule changes. When transit falls by six days, the number that moves is not your arrival date, it is your purchase order date. Your requirement date is set by the customer or the production line and it does not move, so a shorter transit pushes the order release later, and later order release means less time to change quantity, which is the whole point of a shorter pipeline. Shippers who take a transit improvement and leave the order date where it was have bought nothing except a longer dwell in their own warehouse. The saving is only real if you move the trigger date on the purchase order at the same time you move the routing.

The arithmetic you need is not the announced saving, it is the announced saving minus the inland delta. The 5 to 7 days applies to the sea leg. Your door-to-door saving is that number, less whatever the Gdansk inland leg costs you in time compared with the hub you use today. If the inland leg from Gdansk to your door is two days slower than the inland leg from Hamburg, a six-day sea saving becomes a four-day door saving, and the entire business case has to be rebuilt on four. If the inland leg is four days slower, you are down to two days and the disruption is probably not worth it. Price the full inland leg from Gdansk to your warehouse before you price anything else.

Here is the part that decides whether this works, and it is not in the announcement. The saving has a ceiling set by the inland connection behind the port, and that ceiling can be close to zero. A direct deep-sea call backed by a thin rail schedule or a customs bottleneck at the port converts a six-day sea saving into a two-day door saving, and sometimes into nothing at all. The port can be excellent and the corridor can still fail, because the constraint is the last hundred kilometres rather than the ocean leg. Ask for the rail departure frequency out of Gdansk to your region, not the transit time on the best day, and ask what happens to that frequency in the week after a public holiday.

The second thing the announcement does not say is that a direct call concentrates risk, and concentration is a cost even when it never materialises. When your cargo transshipped, a problem at the hub had workarounds: a different feeder, a different sailing, a delay that could be recovered. When your cargo rides a single direct string, a missed call at Gdansk is a week, not a rebooking, because the next vessel on that string is a week away and the one after that is two. That is the trade you are making. You are selling routing flexibility for transit time, and the price of the flexibility you sold is only visible in the weeks when something goes wrong.

There is a third consequence that almost nobody prices, and it is about where your buffer sits. At a hub, you had a place where cargo could be held, split or consolidated. A direct call removes that place, so every box is committed to the full rotation from the moment it is loaded. If your warehouse relied on the hub as a de facto buffer, moving to a direct call does not just change the transit time, it removes a piece of your inventory architecture and hands the function to your own distribution centre. That cost shows up as warehouse space and as working capital, and it never appears on a freight comparison.

The reverse case deserves the same rigour as the case for moving. If Central and Eastern European demand softens, or if Gdansk's inland performance disappoints, the direct call becomes a candidate for removal at the next network review, and the shippers who moved their routings pay the cost of moving back. The conditions that flip this are visible and specific: loadings on the string, and port productivity data. Both are readable within two or three rotations. Keep the Hamburg or Rotterdam routing alive as a live option through the first two quarters of the new schedule, and close it on the data rather than on the announcement. A fallback you have unwound is not a fallback.

One thing to be clear about before you build a plan on six days: the saving is not yours to keep. A direct call is available to every importer on that corridor at the same time, which means the six days does not make you faster than your competitor for long. It makes you as fast as your competitor, and the ones who moved first get a season of advantage before the corridor fills and the transit time stops being a differentiator and becomes a baseline. That is worth having, and it is worth exactly one season, not a strategy. The shippers who treat a network change as a permanent edge are the ones who are still paying for the last one.

The second thing to be clear about is the difference between a transit time and a schedule. A transit time is an average over the sailings that went well. A schedule is a promise about the ones that do not. Ask your carrier for the on-time performance of the specific string, not the design transit time, because the number you have been quoted of 5 to 7 days is a design figure and design figures are computed on a route that has not yet met a winter, a congestion event or a missed connection. If the string runs at 70% on-time, then roughly one box in three arrives on a different day than your forecast assumed, and your safety stock has to cover that, which takes back a large part of the bucket three saving.

Design the trial properly or the trial teaches you nothing. Run one shipment and record four timestamps: the documentation cutoff as it was actually applied, the vessel departure, the Gdansk discharge, and the arrival at your warehouse door. Those four give you the sea leg, the port dwell and the inland leg as three separate numbers, and only then can you compare them with the same four timestamps from your current Hamburg or Rotterdam routing. Most trials record the departure and the arrival and skip the two middle points, which is precisely where the difference between a six-day sea saving and a two-day door saving is decided.

What I will watch, in order of how early it tells me something. Loadings on the first sailings out of Gdansk, because an empty first call is the clearest signal that the corridor cannot fill a string. Inland rail and road capacity announcements from the port, because that is where the saving is either realised or lost. Whether competing alliances put their own direct calls into the Baltic or elsewhere in Eastern Europe, which would confirm this is a network war rather than a one-off, and would compress the window in which the first mover keeps the account. And the cutoff times on the new rotation, because a cutoff that moves earlier can eat a day of the saving before the vessel has left.

The number to keep in front of you is not 5 to 7. It is 5 to 7 minus your inland delta, adjusted for which end of the range your lane sits on, and then checked against the cost of losing your hub buffer. Most of the shippers who get this right will find their answer somewhere between two days and five, which is a real saving and a much less exciting sentence than the one in the announcement. The ones who get it wrong will quote the seven.

One last thing, and it is the least comfortable. A network extension is a promise about a schedule that has not run yet. The alliance will be right about the sea leg, because that is arithmetic. The question is whether the port and the railway behind it are ready on the same date, and that is not arithmetic, that is capacity, and capacity is announced in press releases long before it is available in the yard.

↑ Back to the key points

— By Vivian Zhao