A-Choice
物流行业亚甄研究院 · 2026年10月26日

舱位订到不等于锁住:马士基开了五天没装柜就自动清舱,订舱台账要改三处

马士基自 10 月 15 日起将未出运订舱的自动取消时限收紧到首程船 ETD 后 5 天,由系统直接回收舱位,不再等人工确认。叠加 9 月全球准班率 27%、远东-欧洲仅 4% 的现实,“订了舱没赶上船”已从偶发变日常。本文拆解两个常见误判——改航次不等于换船期、系统取消不等于免亏舱费,并给出订舱台账的三个字段改造与 ETD-3/ETD-5 预警节点。

舱位订到不等于锁住:马士基开了五天没装柜就自动清舱,订舱台账要改三处

10 月 15 日起,马士基对未实际出运的订舱改了时间口径:从首程船预计离港时间(ETD)起算,满 5 天货物仍未实际出运,符合条件的订舱可能被系统自动取消。公告用的是“符合条件的订舱”“可能自动取消”,不是所有订舱到第 5 天一律清掉,但触发之后执行的人是系统,不是客服。这一句改动看似只是把时限说清楚了,实际是把一条存在了很多年的灰色缓冲带收掉了。

过去的做法大家都熟悉:柜子没赶上船,订舱先挂着,人再去和船司沟通,能延就延,不能延再改航次。新规之后,这条缓冲带只剩 5 天,而且第 6 天不会有人打电话通知你。船司的目的写在公告里——更早释放没用上的舱位,让运力重新进入分配。

现在这个时点尤其不能大意。Xeneta 9 月准班率报告显示,全球班轮准班率降到 27%,比 8 月再低 2 个百分点,连续第四个月下滑,与今年 2 月的低点持平;船舶平均延误升到 6 天,比二季度初多了 2.4 天。分航线看更极端:远东到欧洲准班率只有 4%,平均延误 9.9 天,是 2021 年 9 月以来最差;远东到北美 20%,延误 6.9 天。上海锚地排队船数从 101 艘降到 86 艘,但等泊时间从 1.9 天延长到 2.7 天——队短了,等得更久了。这意味着“订了舱、柜子却没赶上这条船”已经从偶发事件变日常事件,而新规恰好卡在这个日常事件上。

两个误判最容易吃亏。

第一个是把“改航次”想得太简单。订舱一旦被系统取消,原订舱号就失效了,这不是后台点一下换个船期,而是重新占舱。在欧线计划取消约 17% 航次的背景下,重新占舱意味着重新面对当下的价格和当下的船期,运气不好还要面对没舱。

第二个是把“系统自动取消”当成“不用付钱”。自动取消回收的是舱位,不等于豁免亏舱费。属于未装船又未及时主动取消的,亏舱费仍可能照收。这一点在合同和确认条件里写得很清楚,账单来的时候不会因为“是系统干的”而网开一面。

对 A-Choice 的操作层面,订舱台账至少要改三个字段。第一,把 ETD 和 ETD+5 这条红线单独列出来,不要和“预计开船日”混在一列;建议在 ETD 前 3 天、前 5 天各设一个内部提醒节点,因为真正需要动作的时间在开船之前,不在之后。第二,把“已订舱”和“已装船”拆成两个状态,中间再加一个“已进港/已放行”,只有装船才算落地;只盯订舱状态,等于自我安慰。第三,加一列“责任人与处置动作”,明确这张单在红线前是“申请改航次”还是“正式取消”,并且必须走官网或 EDI 留痕——口头报备在事后争议里没有分量。

还有一个容易被忽略的分寸:确实要延后、但不想取消的订舱,在 ETD 后 5 天内主动向船司客服报备,是保留舱位有效性的唯一动作。这一步不做,系统按规则走,谁也没有裁量空间。

连带影响还有一层:清舱时点提前,意味着改配、改单、改报关资料的时间窗一起被压缩。柜子已经进港、报关单已经放行、船却走了,这类单子过去可以慢慢转到下一条船,现在要在 5 天内完成重新订舱、舱单更改和报关对接,任何一环卡住都会连累下一个航次。真正难的往往不是抢舱,而是舱抢到了、单证跟不上。

再往上看一层,这会改变旺季的锁舱习惯。以前舱位紧张时多占几个舱做保险,成本很低;现在冗余订舱的持有成本从“占用信用”变成“5 天后清零 + 潜在亏舱费”,多锁反而更贵。更稳的做法是精准锁舱加备选船司:主选舱位对应确定的货盘,同时把第二家船司的船期与价格提前落成书面报价,而不是等出事再去找。

给客户的报价单上也值得加一句:舱位有效性以实际装船为准,ETD 后 5 天未出运的订舱存在被自动取消的可能。这句话不讨喜,但比交期塌了之后再解释要体面。

English Version / 英文版

A Booked Slot Is Not a Locked Slot: Maersk Clears Unshipped Bookings Five Days After ETD

From 15 October, Maersk changed the clock on bookings that never materialise: where five days or more have passed since the first vessel's estimated time of departure (ETD) and the cargo still has not shipped, eligible bookings may be cancelled automatically. The wording is careful — "eligible bookings", "may be cancelled" — so this is not a blanket sweep on day five. But once the trigger is met, the party executing the cancellation is the system, not a customer service desk. Read plainly, a grey buffer that had existed for years has just been removed.

The old routine was familiar to everyone in the trade: the container missed the vessel, the booking stayed open, someone spoke to the carrier, and a rollover was negotiated. That buffer is now five days, and nobody will call on day six. The carrier's stated purpose is in the notice itself — release unused capacity sooner so it can re-enter allocation.

The timing makes this hard to absorb. Xeneta's September scorecard put global schedule reliability at 27%, down two percentage points from August, the fourth consecutive monthly decline and level with the low set in February. Average vessel delay rose to six days, 2.4 days more than at the start of the second quarter. By trade the picture is starker: Far East to Europe fell to 4% on-time with average delays of 9.9 days, the weakest since September 2021; Far East to North America held at 20% with delays of 6.9 days. At Shanghai the anchorage queue shortened from 101 to 86 vessels while waiting time lengthened from 1.9 to 2.7 days — a shorter queue that takes longer to clear. Cargo booked but not loaded is no longer an exception; it is routine, and the new rule sits directly on top of that routine.

Two misreadings cost the most.

The first is treating a rollover as a formality. Once a booking is cancelled by the system, the booking number is dead. This is not an edit to the sailing date; it is a fresh claim on capacity at today's rate and today's schedule — and European services are planning to blank roughly 17% of sailings.

The second is reading "automatically cancelled" as "no charge". What the system reclaims is the slot, not the liability. Where cargo failed to load and no timely cancellation was filed, a no-show charge can still apply. The contract and confirmation terms are explicit on this, and the invoice will not care that a system did it.

For our own operations at A-Choice, the booking ledger needs three changes. First, carry ETD and the ETD+5 line as a field of their own rather than folding them into "scheduled departure", and set internal alerts at ETD-3 and ETD-5 — the action has to happen before the vessel sails, not after. Second, split "booked" from "loaded" and put "gated in / released" between them; only loading counts as landed. A ledger that tracks booking status is a ledger that reassures without informing. Third, add an owner-and-action column stating, before the deadline, whether the booking is being rolled to another sailing or formally cancelled — and filed through the portal or EDI, because a verbal request carries no weight in a later dispute.

One distinction is easy to miss: for cargo that genuinely slips but should keep its slot, notifying carrier customer service within the five days after ETD is the only action that preserves validity. Skip it and the system applies the rule; there is no discretion left to appeal to.

One level up, this changes peak-season habits. Reserving extra bookings as insurance used to be almost free. Now the carrying cost of a redundant booking is a hard reset after five days plus potential no-show charges, so over-booking has become more expensive than being caught short. The steadier approach is precise slotting plus a named alternative: a primary booking tied to confirmed cargo, and a second carrier's sailing and rate already quoted in writing rather than sourced in a crisis.

It is also worth adding one line to customer quotations: slot validity is confirmed on loading, and bookings unshipped five days past ETD may be cancelled automatically. Unpopular, but better said upfront than explained after a delivery date collapses.

需要这条航线的运价与舱位方案?
提交起运港与目的港,10 秒获取参考运价,专人跟进未命中航线
立即询价
返回行业洞察