The schedule is a plan, not a promise
When you book ocean freight, you get a schedule: a sailing date, a transit time, an arrival date. It looks precise. But a container ship's schedule is a plan drawn up before the voyage, and the ocean rarely cooperates with the plan. Weather, port congestion, blank sailings, and reroutes all bend it. The result shows up in one industry statistic that every importer should know, and most have never seen.
That statistic is schedule reliability: the share of vessel arrivals that land on time. Sea-Intelligence tracks it every month across the major carriers in its Global Liner Performance report, and through 2025 it hovered in the low-to-mid 60s percent, ranging from around 58.7% in April to a high of 67.4% in June. In plain terms, roughly one in three container ships arrived late, month after month, even in a relatively calm year for the industry.
Roughly 1 in 3 ships arrives late
In 2025, global schedule reliability sat around 60 to 67%, according to Sea-Intelligence, so about a third of vessel arrivals missed their scheduled date. And when a ship was late, it was not a little late: the average delay for late arrivals ran close to 5 days. A five-day miss is enough to blow past a free-time window, miss a connecting service, or leave a production line waiting on parts.
How bad it can get, and how good
The low-60s figure is actually the calm end of the range. Reliability is deeply cyclical. Before the pandemic, in 2019, on-time performance sat around 78%. Then the port congestion of 2021 and 2022 sent it crashing to a record low near 34%, a point where more ships were late than on time and the average delay stretched beyond a week. It clawed back through 2023, took another hit when Red Sea reroutes scrambled schedules in 2024, and settled into the 60s through 2025. The lesson is that reliability is not a fixed number you can quietly design around. It moves with the state of the world.
The 2025 shake-up: a bet that reliability can be engineered
Reliability got a jolt in early 2025 when the ocean carrier alliances reshuffled. On February 1, 2025, Maersk and Hapag-Lloyd launched the Gemini Cooperation, built around a different idea: a hub-and-spoke network that runs a small number of tightly controlled mainline routes and feeds smaller ports with regional shuttles, instead of having every ship call at every port. The goal was blunt, an ambition of more than 90% schedule reliability once fully phased in.
It largely worked. Gemini reported 92% reliability in its first month, and across its first full year it averaged around 90%, with the Asia to North Europe and Transatlantic lanes reportedly clearing 95%, according to Hapag-Lloyd and Maersk. Set against an industry average in the low 60s, that is a striking gap, and it reset expectations for what the word reliable can mean at sea. The rest of the market reorganised over the same period into the Ocean Alliance and the new Premier Alliance, with MSC, now the largest carrier by capacity, running much of its network independently.
90% versus 60%: reliability is now a choice
Gemini's roughly 90% reliability against a low-60s industry average shows that on-time performance is not fixed. It is a design decision. That is encouraging, but it also means reliability now varies widely by carrier and by network, so which service you book genuinely changes the odds that your cargo arrives when it should.
Why a late ship costs more than lost time
A missed arrival date is not just an inconvenience. It sets off a chain of costs that land in different places:
- Demurrage and detention. A late arrival can compress or blow past your free-time window, so a delay you did not cause turns into storage fees you still have to pay.
- Missed connections. For cargo that transships, a late mainline vessel can miss its feeder, turning a few days' delay into a week or more while the box waits for the next sailing.
- Inventory and stockouts. Safety stock is sized around expected transit times. When a third of ships run days late, the buffer either has to grow, which ties up cash, or it runs out, which stops a line or empties a shelf.
- Broken customer promises. A delivery date you quoted off the schedule becomes a date you miss, and the cost of that lands on the relationship, not just the ledger.
Why you can't plan off the schedule alone
If even the industry's most reliable network misses about one sailing in ten, and the average carrier misses one in three, the published schedule cannot be the thing you plan around. It is the right starting point at booking. But the moment the ship is moving, the only date that matters is the one based on where the vessel actually is and how it is really sailing.
This is the case for live tracking paired with a predictive ETA. Rather than clinging to the arrival date printed at booking, a predictive ETA is recalculated from the vessel's real position, speed, and the conditions ahead, so when a ship falls behind or a port backs up, your expected arrival moves with it. You find out about a slip while there is still time to act, instead of when the box simply fails to show up.
The date worth planning around
A published schedule tells you what was supposed to happen. A predictive ETA tells you what is actually going to happen. When one in three ships is late by around five days, that difference is the whole space between reacting in time and explaining a miss after the fact.
How to work with unreliable schedules
You cannot make carriers run on time, but you can stop being caught out when they don't:
- Treat the booking date as an estimate, not a fact. Build your plans on a live ETA that updates as the voyage unfolds, and revisit them when it moves.
- Know your carriers' reliability. On-time performance varies widely by carrier and network, so weigh a service's track record at booking, not just its rate and headline transit time.
- Watch every shipment in one place. A delay on any line should surface the same way, so nothing hides in a separate portal until it is too late to react.
- Set alerts on the delays that matter. A meaningful ETA slip, a missed connection, or a long dwell at a congested port should reach you as a notification, not as a surprise at delivery.
- Build buffer where it counts. For time-critical cargo, size safety stock and lead times around real-world variability, not the optimistic number on the schedule.
The bottom line
A sailing schedule is one of the most confident-looking documents in freight, and one of the least reliable. Through 2025, about a third of container ships arrived late, typically by around five days, and only a redesigned network like Gemini pushed reliability toward 90%. The schedule still matters as a plan, but it is not a promise, and running a business as if it were is how good operations get blindsided. The importers who stay ahead treat the arrival date as a live number, watch where their ships actually are, and learn about a delay early enough to do something about it. In a world where the schedule regularly misses, knowing the real ETA is the advantage.