Watch the work order, not the shipping calendar. A fabric pouch order at mid-October is usually not on a vessel yet, so ocean cut-off dates describe a decision you will face in three weeks, not the one you face now. The checkpoints that should trigger partial air freight are the ones upstream of the port: whether the approved proof is signed, whether the cloth is booked or still waiting on a dye batch, and whether the sewing line has a confirmed slot. Any one of those slipping past its date is the signal to move your launch quantity to air, and it arrives far earlier than any vessel delay.
Why the shipping calendar is the wrong first instrument
Ask how to split an order between sea and air and the answer comes back as vessel cut-off dates and transit times. That is the right frame when the goods are finished and sitting in a warehouse. It is the wrong frame in mid-October for a custom pouch, because at that point the goods do not exist, and the largest risks are in the factory rather than on the water.
The two failure modes also resolve differently. A vessel delay is a known event with a known remedy: you switch mode and pay the difference. A production slip discovered late is worse, because by the time you learn the sewing line has not started, the ocean option and possibly the air option have both narrowed, and you are choosing between expensive and impossible.
So the first instrument is a work order with dates on it, and the first job is comparing the date on each line with today.
The six checkpoints, in the order they fail
| Checkpoint | What it tells you | Air trigger |
|---|---|---|
| Proof signed | Whether the artwork is released for production at all | Unsigned at the agreed date: nothing downstream is real |
| Sample approved | Whether the physical pouch has been agreed | Not approved by the sample date: shift the whole plan or air the first portion |
| Cloth booked | Whether the fabric is reserved or still in a dye queue | Stock shade unavailable and a dye batch pending: the long pole, air whatever can be cut |
| Sewing slot confirmed | Whether the line has capacity in your window | Slot not confirmed for your week: split, air the launch quantity |
| Inspection booked | Whether the goods can be cleared in time to ship | Inspection date after the cut-off: re-sequence or air directly from a partial lot |
| Vessel booked | Whether an ocean option still exists for the remainder | No booking at cut-off: the remainder is air or nothing |
Two of those deserve more than a row. The cloth checkpoint is the one specific to this product and the one most often absent from a generic shipping plan. A stock shade can be cut in days. A shade that has to be dyed waits for a dye batch, and a batch waits for a number of metres and a slot in the dyehouse queue. That queue is not visible from your shipping tracker, and it is where October orders for December quietly die.
The inspection checkpoint is a sequencing problem rather than a capacity one. A third-party inspection needs the goods finished and packed, which means it sits after sewing and before loading, and if the inspection date falls after the vessel cut-off then the ocean option has already gone whatever the factory says about catching up.
Dividing the order rather than choosing a mode
The decision is rarely sea or air for everything. It is which portion goes by air, and the answer follows three questions: what has to be on shelf for the season’s first weeks, what a customer has already committed to, and what can arrive late without losing the sale.
A workable split, in order of what to air: the launch quantity for each colourway, because an empty shelf at the start of the season costs more than the freight; quantities already promised to specific accounts, because a missed delivery to a retailer is a penalty and a relationship cost as well as a lost sale; and any item whose margin covers the difference. What stays on the water is the replenishment behind the launch, slow-moving colours, and bulky items where air freight changes the economics.
Write the split as a quantity per item rather than as a fraction. “Air 30 percent” is not an instruction; “air 400 of the black, 300 of the navy, sea the balance” is, and it can be quoted against a forwarder the same day.
The arithmetic that makes the decision defensible
Two numbers decide it. The first is the cost of the goods arriving late, which is not a single figure but the sum of markdown on unsold stock, lost sales on the units you could have sold, and any chargeback a retailer applies. The second is the air premium, which is the difference between the air rate and the ocean rate for the same chargeable weight.
The comparison only becomes meaningful when both are stated per unit. Air premium divided by units gives you the extra cost per unit to protect the season. If that number is smaller than the margin you would lose by not having the unit on shelf, air is cheaper than sea and the argument is over. Air freight is priced on chargeable weight, which for a soft pouch means volume rather than mass, so ask the forwarder for the volumetric conversion they apply before you assume the pouch is light enough to be cheap.
Rate movement in the peak matters here. Air capacity in November and December is constrained and rates rise, so a decision deferred to late November costs more than the same decision taken in October. That is another reason the work order is the right instrument: it makes you decide while the cheaper option still exists.
What to do about a partial shipment once it lands
A split order creates two events where there was going to be one, and both need handling. The air portion has to be received, counted and put away as a separate consignment, with its own carton marks so that warehouse staff and the retailer can tell which units arrived by which route. The sea portion follows later and has to be reconciled to the same order, which means the packing list has to say clearly which units were in which shipment.
Mixing the two in a single record creates a quantity discrepancy that surfaces at invoice. Keeping them as two shipments against one order, with the ocean balance named in writing, is the arrangement that survives an audit.
There is also a documentation point that is easy to miss when a shipment is expedited. Cargo insurance on what all-risks cover on pouch shipments is usually bought per shipment, and an air consignment arranged in a hurry is a shipment. Check that cover exists for both parts before either moves.
Putting the decision into a weekly rhythm
A one-off calculation in October is not a decision system. What makes this work is a weekly comparison, done on a fixed day, with the same six checkpoints and a written outcome. The point of the rhythm is that the question changes from whether to act into whether anything moved.
The mechanism is short. Keep a one-page work order with a date against each of the six checkpoints. Each week, mark each line as on schedule, at risk, or slipped. If a line is slipped, decide in that meeting whether it affects the season and, if it does, which units move to air. Record the decision and the date, even when the decision is to do nothing, because a series of recorded “no change” entries is what turns an eventual move to air from a panic into a planned step.
Two people need to be in that meeting: whoever owns the supplier relationship and whoever owns the sales commitment. Without the first you cannot find out what the cloth queue is doing. Without the second you cannot price the cost of being late, and the cost of being late is the number the whole decision rests on.
What the order should say so this is possible
None of the above works if the order does not require the factory to report milestone dates. Add a short clause: the factory confirms dates for proof approval, sample approval, cloth readiness, sewing start and completion, and notifies the buyer in writing within a stated period of any of them slipping.
Pair it with the capacity commitment, which is the subject of what an order must say when a bigger client takes the line. A factory that has agreed a slot and has to notify a slip is a factory you can plan against. One that has agreed nothing will tell you about the delay when you ask, which is generally the week the vessel leaves.
Finally, agree in advance that a partial air shipment is an available remedy rather than a breach. Orders that treat air freight as an admission of failure push the decision to the last possible week, which is exactly when it costs the most and works the least.
The whole structure is cheap to set up and it is settled at order stage or not at all. A buyer who asks for six dated checkpoints and a notification clause has a planning instrument. A buyer who asks only for a delivery date has a promise, and the promise is not the thing that tells you when to book the air waybill.
