Replenishment

What belongs on a purchase order when the supplier is overseas

A domestic reorder survives a sloppy purchase order, because the van comes on Thursday and someone spots the mistake. An import order is read once, months before anyone can check it.

Marcus VolstedCo-founder, Restocio Published 21 September 202612 min read

Every source on this page was opened and read, and each one is dated in the Sources section at the foot.

A purchase order going to an overseas supplier needs eight things: the SKU exactly as your store spells it, the quantity, the unit cost and the currency it is in, the incoterm with the place it names, which freight mode each part of the order travels on, the day the goods leave the supplier, the day you expect them sellable, and the payment terms. A domestic reorder gets away with half of that. The goods land on Thursday, somebody notices, and one phone call sorts it out.

On an import nobody notices. The order goes out, the factory reads it the way it is written, and the first evidence that anything was wrong turns up eleven weeks later on a pallet. Everyone in the chain followed the document. The document was the problem.

The eight lines that earn their place

Shopify's Help Center describes a purchase order as the record that lists “the products, quantities, costs, payment terms, and supplier details for an order”, read on 21 September 2026. That is the commercial core, and it is the same in any purchasing system. Importing adds four more that a general template has nowhere to put: the incoterm, the freight mode, and two different dates.

What goes on the order	What it has to say	What happens when it is missing
SKU	The code your store uses, not the factory's article number	Nothing can match the order back to a product, so the units never reduce your next order
Quantity, per variant	Per size and per colour, not per product	The size split becomes the factory's guess, and it will guess evenly
Unit cost and currency	The agreed price and the money it is stated in	A number with no currency on a document crossing a border is read as whatever the reader uses
Incoterm and named place	The rule you agreed, and the port or door it applies at	Who pays for the main carriage, and where your risk starts, gets settled after something has already gone wrong
Freight mode	Which units ride the boat, the train or the plane	The order has no arrival date anyone can plan against
Ship date	The day the goods are due to leave the supplier	You cannot tell a slow factory from a slow crossing, and you chase the wrong person
Expected arrival	The day you expect them countable on your own shelf	Nothing for a late order to be late against, so it is never late and never chased
Payment terms	The deposit share and when the balance falls due	Your cash plan for the quarter is a shrug, and the deposit lands as a surprise

The incoterm is the one most small importers skip, usually because it feels like paperwork for somebody larger. The ICC, which writes the rules, says on its own page that they exist to avoid costly misunderstandings “by clarifying the tasks, costs and risks involved in the delivery of goods from sellers to buyers”, and the edition in force is Incoterms 2020, read on 21 September 2026. Three letters and a place name on the order is the whole job. Leaving it off does not remove the question, it just postpones it until a container is sitting somewhere accruing storage.

Why the same mistake costs more when it crosses an ocean

Nothing on that list is exotic. What changes with distance is the gap between making a mistake and finding out, and that gap is where the money goes.

The mistake	Local supplier, three-day lead time	Imported, 90 days including production
Wrong quantity	Top it up next week	Either a quarter of dead stock or a quarter of stockouts, and the fix is air freight
Wrong variant on the line	Swapped in a day	A pallet of the wrong colour, and the right one is still 90 days away
Freight mode never written down	Irrelevant, the van comes when it comes	You cannot tell whether the order lands before or after you run out
Price agreed verbally	Corrected on the next invoice	A wrong cost sits inside every margin you look at for a quarter
Order never closed off	Obvious, the shelf is full	Units count twice, on the shelf and on the water, and the next order comes out too small

Read the right-hand column again and notice what the entries have in common. None of them is a big dramatic failure. They are all small clerical things that had a long time to compound.

Put the freight mode on the order, not in your head

This is the field most purchasing templates do not have, and for an importer it decides more than the quantity does. An order of 3,000 units is one event on paper and often two or three shipments in practice. Send 300 by air to cover the weeks before the boat lands and the rest by sea, and that single order now has two arrival dates. Write down one of them and you have lost the one you were relying on.

So the mode belongs on the line, not on the order. In Restocio a purchase order line carries its own quantity per transport, so one product can hold sea units, rail units and air units at once, and the supplier copy of the document prints that split per line. When the order is placed, its expected date is seeded from the slowest leg that actually carries units, plus the supplier's production days, plus the delay that supplier usually runs.

The slowest leg is the right default and it is also the honest weakness of it. An order whose air portion lands in a fortnight still shows the boat's date, because that is the day the order is finished rather than the day the first box arrives. If you need the early part tracked separately, split it into its own deliveries, which is what the Goods in page is for. Choosing the split in the first place is a separate question, covered in air or sea freight and how to work out the split.

Two dates, and neither one is the day you pressed send

Most spreadsheets hold a single date column, and it is usually the day the order went out. That date tells you nothing you can act on later. The two that matter are the day the goods leave the supplier and the day they are countable on your shelf, and the distance between them is your freight, your clearance and your own receiving.

Keeping both apart is what lets you chase the right party. An order that shipped on time and landed late is a freight problem. An order that never shipped is a factory problem, and no amount of chasing the forwarder moves it. That is also the difference between a lead time you can plan with and one that quietly runs short, which supplier lead time and why yours is probably too short takes apart segment by segment.

One practical detail for anyone starting a new system mid-flight. Orders already placed need entering with their real order date, not today's, or the app dates them from the day you typed them and every arrival estimate sits weeks too late. Restocio lets you move the order date back for exactly this reason, and the expected date follows from it.

The SKU has to match your store, letter for letter

Suppliers keep their own article numbers, and it is normal for a purchase order to grow a second set of codes that only the factory understands. That is fine on paper and fatal in software. Any system that nets what is already coming against what you are about to buy has to match the order line to a product, and it matches on the SKU.

A line carrying a code your store does not know is not a small formatting issue. It is a quantity that exists on your order and nowhere in your planning. Restocio flags those lines on the order row with a count and a hover that says plainly they “are not counted as incoming stock for any product”, and a draft offers the closest real SKUs as buttons so the line can be corrected in place. The reason this matters at all is the one Shopify's own incoming inventory has already taught most importers the hard way: what the system cannot see, it will happily buy again.

An order nobody closes keeps lying to your next order

The second half of purchase order hygiene is not about writing the document. It is about ending it. An order that was placed, arrived, got unpacked and was never marked received goes on counting as stock in transit forever. Those units then get counted twice, once on your shelf where they really are and once on the water where the record says they are, and the next recommendation comes out too small on exactly the product that just sold well enough to need topping up.

The mirror case is worse because it feels responsible. An order the supplier quietly cancelled, or one stuck at a port with no date, keeps its units in the plan and stops you reordering something you genuinely do not have.

Restocio asks about this rather than guessing. From the day after an order's expected date, plus however late that particular supplier usually runs, the order shows up with a straight question and five answers: it arrived, it is on its way with a new date, it was cancelled, order a replacement instead, or ask me again in three days. There is a second prompt for the opposite gap, when stock for products on an open order rises by at least half of what that order still has on the way, which usually means the goods landed and nobody wrote it down. It offers to open the counting screen and nothing more. Marking goods received is a person's job, not something we infer from a stock movement we did not witness.

If you run this on paper instead, the same discipline is one weekly pass: every open order older than its expected date gets an answer, even if the answer is that you do not know yet.

Counting in is where the document earns its keep

Count what arrived line by line, against what the order says, on the day it lands. Not the number of cartons on the pallet, the units per SKU. This is the only moment when the order document and physical reality are in the same room, and it is the last chance to catch a short shipment while the supplier still remembers the batch.

While you are there, keep three things apart rather than netting them into a single figure. What arrived, what is still coming on a later leg, and what came in damaged. Those are different states, and collapsing them loses the claim you were entitled to make.

In Restocio a receipt prefills what is still outstanding for the transport you picked, a line that comes up short asks whether the rest is still coming or is never coming, and the order's status follows the counts rather than the other way round. You cannot mark an order partially received by hand, because a partial with no shortfall recorded anywhere is a claim the line data contradicts. Where counts come up short, Goods in groups the missing and damaged rows by supplier and gives you a copy button, so chasing them is a paste into your own email rather than an afternoon in a spreadsheet. The chase itself stays yours, which is the right place for it.

If you want the order document, the arrival date and the reorder plan to be the same object rather than three files that disagree, that is the part of Shopify replenishment we built the Purchase Orders page for.

What a purchase order still will not do for you

Three limits, two of them ours.

There is no incoterm field in Restocio today. The supplier document carries the order number, the supplier, both dates, the lines with their transport split, optional prices and product links, and a notes block. Neither the incoterm nor the payment terms print on it. Payment terms do at least exist inside the app, as a deposit share on the supplier card that pre-fills the first payment against an order. The incoterm has nowhere to live but your own agreement with the supplier, and deeper supplier profiles, incoterms included, sit on the roadmap rather than in the product.
Counting goods in does not change your stock level in Shopify. That is deliberate rather than unfinished. The permission exists and the endpoint exists, and we are not calling it until we are certain it cannot corrupt a live catalogue, so receiving in Restocio updates the purchase order and the plan while Shopify stays whatever your own process makes it. Worth knowing before you build a receiving routine around it.
A purchase order cannot make a supplier honest about dates. It records what was promised, which is exactly what makes the pattern visible after a few orders. That is worth more than any single date on any single order, and it only works if the orders get closed.

Common questions

What should a purchase order include for an overseas supplier?

The SKU as your own store spells it, the quantity per variant, the unit cost with its currency, the incoterm and the place it names, the freight mode each part of the order travels on, the day the goods leave the supplier, the day you expect them sellable, and the payment terms. The last four are the ones a general purchasing template usually has no field for, and they are the ones that decide whether an import lands on time.

Do I need to put the incoterm on the purchase order?

Yes, and it costs you three letters and a place name. The ICC describes the Incoterms rules as clarifying the tasks, costs and risks involved in delivering goods from seller to buyer, and Incoterms 2020 is the edition currently in force. Leaving it off does not remove the question of who pays for the main carriage and where your risk begins. It postpones it until a shipment is already stuck somewhere.

What is the difference between the ship date and the expected arrival date?

The ship date is when the goods leave the supplier, the arrival date is when they are countable on your shelf, and the gap between them is freight, customs clearance and your own putaway. Keeping both lets you tell a slow factory from a slow crossing, which decides who you chase. A single date column, usually the day you placed the order, tells you neither.

What happens if a purchase order never gets marked as received?

Its units keep counting as stock in transit, so they are counted twice: once on the shelf where they actually are, and once on the water where the record still says they are. The next reorder then comes out too small, and it does so on the products that are selling. The cure is a weekly pass over every open order past its expected date, giving each one an answer even when the answer is that you do not know yet.

Should one purchase order carry more than one freight mode?

Often yes, and the split belongs on the line rather than on the order. A single order might send a few hundred units by air to cover the weeks before the container lands and the rest by sea, which gives that order two arrival dates. Writing only one of them down loses the one you were planning against. In Restocio each line holds its own quantity per transport, and the order's expected date follows the slowest leg that carries units.

Does receiving a purchase order in Restocio update my Shopify stock?

No. Receiving updates the purchase order and the replenishment plan, and it writes nothing to Shopify. We hold the permission and the endpoint exists, and we are deliberately not calling it until we are sure it cannot corrupt a live catalogue. Your Shopify stock level stays whatever your normal process makes it, so build your receiving routine with that in mind.

Sources

Every source below was opened and read on the date shown against it. Rules, rates and schedules change, so check the current page before you act on a number that matters. If anything here is out of date or wrong, tell us and we will correct it.

Shopify Help Center, Purchase orders (read 21 September 2026)
International Chamber of Commerce, Incoterms rules (read 21 September 2026)

Marcus Volsted

Co-founder, Restocio

Marcus co-founded Restocio and works on it daily with a Swedish importer who plans their purchasing in it every working day. Restocio is built in Sweden by two founders, one Swedish and one Danish, and Marcus is the Danish one. Why we are building it.