An order placed in Yiwu, China in the second week of August doesn’t meet the same factory that quoted it in June.
Nothing has changed about the equipment or the people. What has changed is the queue. Chinese plants run toward two fixed walls in the autumn, and both of them are already visible. The Golden Week holiday closes production for the first week of October, and before that, every buyer with Q4 shelf dates is trying to get goods onto water. The result is an orderbook that fills from the top down, and a factory floor where the sequence of work stops being negotiable during some time in August.
Buyers experience this as a delay, but it’s more useful to understand it as a reallocation. Line time in this season goes to whoever committed volume earliest. A plant manager planning a six-week run of one product against 40 small runs of 40 products will take the six-week option, because changeovers cost him hours he no longer has. Your order loses, not to a bigger customer, but to the arithmetic of setup time.
What happens to the small order is the part that matters for quality. It gets slotted into gaps. A run that would normally occupy a clean block gets split across the ends of other jobs. That means more changeovers. It also means more first-article checks that nobody has time to do properly. The operator running your goods on Thursday might not be the one who ran them on Tuesday. Or the order goes out to a subcontractor — and this is where buyers get surprised. The subcontractor is a real factory with real capability. It’s just not the factory that anyone audited. It’s the one with capacity in week 36.
Neither of those is concealed. Producers simply aren’t reporting upward, because from the supplier’s side nothing unusual has happened. He committed to a date, and he’s meeting it.
Quality assurance thins at the same moment, and for the same reason. A plant that runs four lines in June and eight in September hasn’t usually doubled its quality-check headcount. The same inspectors cover more product, and the checks that get compressed first are the ones that feel routine. Incoming material verification goes early. So does the first-article sign off, and the visual pass slides from the first carton of a run to the third.
Then the buyer’s own window closes.
Third-party inspection slots in this period get booked out. An inspection ordered on the usual notice can fall after the goods are already scheduled onto a truck. The buyer then chooses between a delayed shipment and an unverified one. Most buyers in that position ship, because the shelf date is fixed and the alternative is missing the season entirely. The decision feels reasonable at the time because nothing has visibly gone wrong.
The failures from this period tend to surface in November and December, in the returns data, long after anyone connects them to a production date in September.
There’s a version of this that goes better, and none of it requires a different supplier. It involves moving the purchase order earlier than feels necessary. The buyers who come through this season cleanly aren’t the ones who negotiated hardest. They’re the ones whose orders were in the book before the allocation happened, which in practice means committing in June or early July for goods that need to sail in September.
Book the inspection when you place the order, rather than when the goods are ready. Inspection capacity runs short in exactly the weeks production capacity does. A slot reserved in advance costs nothing to move. A slot requested at the last minute might not exist.
Ask which line will run the goods and whether any part of the order is going outside the plant. Suppliers answer this far more often than buyers expect. In their frame, it’s a scheduling question rather than an accusation. Asked in June it is easy to answer; asked in September ,it becomes a conversation nobody wants, because by then the answer is a subcontractor two towns away.
Accept a later date rather than a compressed one, where the calendar allows it. A buyer who says the goods can ship the week after Golden Week rather than the week before will usually get a cleaner run, because the factory restarting after a shutdown has an empty floor and a full crew.
The last point is the one that gets missed: Get the supplier’s own holiday schedule in writing, covering the days either side of the official closure, because Chinese factories don’t stop and restart cleanly on the published dates. Workers travelling long distances leave early and return late, and a plant that’s officially open on the eighth of October may be running at partial crew for another week or more. A ready date that assumes a full restart on the official first working day is one that’s built on a calendar that nobody on the factory floor is following.
None of this reflects badly on Chinese manufacturing, and it’s not a story about suppliers cutting corners. It’s about a fixed shutdown date, a demand peak that lands immediately before it, and a buyer whose purchase order arrived after the sequence was already set. The suppliers are managing a hard constraint. Plan around it and you get the same factory in September that you got in June. Miss it, and you get a factory doing its honest best inside a queue it can’t expand.
Liam Cai is the founder of Supplymo (United Profit Import and Export Co., Ltd.), in Yiwu, China.