Start a conversation

China sourcing

How a factory decides whose order runs first

A confirmed order is not a place in the queue. A factory allocates a finite capacity between customers who all have confirmed orders, and the ranking is made on criteria nobody writes in a quotation.

France → China

The experience is familiar and feels like a betrayal: the order was placed, the deposit was paid, the date was confirmed in writing, and the goods are late while another customer's goods are not. The natural reading is that the factory is unreliable or that the relationship has gone wrong. Usually neither is true. The factory is doing arithmetic, every week, on a capacity that does not stretch, and you were on the wrong side of it.

What capacity actually is

Capacity is not a factory's floor area or its machine list. It is the narrowest of several different limits, and which one binds changes through the year. It can be a machine — one press, one oven, one line that everything has to pass through. It can be people: a skilled operation that three employees can do and the rest cannot. It can be a component arriving from somewhere else, in which case the factory's own capacity is irrelevant. It can be cash, when material has to be bought before anyone pays. Asking which of these is the real constraint, on an ordinary visit when nothing is urgent, tells you more about your future lead times than any capacity figure in a presentation.

The second thing to understand is that capacity is consumed by changeovers as much as by production. Every switch from one reference to another costs setup time, a first-article check and some scrapped material, and none of it produces sellable units. A factory running many small orders for many customers loses a significant part of its day to this, which is why an order of a single reference in a long run is worth more to it than the same quantity spread over several — and why your order size is read alongside its complexity, not on its own.

Why a confirmed order is not a position

A confirmed order is a commitment to produce, not a commitment about when. The date in the acknowledgement was calculated from the workload known on the day it was written, and the workload has changed since. When it changes, the factory has to disappoint someone, and it chooses by what each customer represents over the year rather than by what each order represents this week. A large customer who places predictable volume every month is a line the factory plans around; a small customer who appears twice a year is a line it fits between the others. Nothing about that is personal, and it is not a breach of anything written.

What moves you up the queue

  • A forecast you keep updating. Not a promise: a rolling view of what you expect, revised downwards as honestly as upwards. It lets the factory buy material and book the constrained operation in advance, which is worth more to it than the order itself.
  • Regularity, even at modest volume. A small order every month is easier to plan than a large one twice a year, and it makes you a line in the plan rather than an interruption of it. This is the single thing a small customer can offer that a large irregular one cannot.
  • A deposit that arrives without being chased. Material is often bought before anyone is paid, so a deposit is not a formality — it is what lets the order start. An order whose deposit is still pending is an order the factory has every reason to schedule behind one whose deposit cleared.
  • Being simple to run. A frozen specification, approved artwork, no last-minute changes, a quality standard the factory already understands, and answers that come back the same day. A customer whose orders run without friction gets scheduled early because scheduling them early is the easy choice.
  • Being known by the person who builds the schedule. That is rarely the sales contact. The planner is who decides the week, and a customer whose name and constraints they recognise is treated differently from a reference number arriving through a mailbox.

Reading a delivery promise

A date given in a quotation and a date given by a planner are different objects. The first is a commercial answer to a commercial question, calculated from a standard lead time and an empty diary; the second is a position in a real schedule. The way to tell them apart is to ask what the date is made of: when the material arrives, when the constrained operation is booked, how long the finishing and the packing take, and what the date assumes about your own approvals. A supplier who can break the date into those pieces has planned your order. A supplier who repeats the total has quoted it.

Two habits make the answer more useful. Ask for the date of the step before the last one — the day production finishes, rather than the day of shipment — because that is the date the factory controls and the one whose slippage you can see coming. And ask what would have to be true for the date to move, in both directions: a supplier who names the risk honestly is a supplier who will tell you early when it materialises, which is worth more than a confident date.

When high season arrives

  1. Book capacity rather than order goods. Tell the factory, well before the season, which weeks you intend to occupy and roughly what for. A booking made early is a line in the plan; the same request made when the plan is full is an interruption of it.
  2. Split the order so that part of it can run early. Producing the half that does not depend on a late decision, in a quiet week, removes that half from the contested period entirely.
  3. Freeze everything on your side first. Specification, artwork, packaging, approvals. In a saturated week, an order waiting for one of your answers is the first one to be moved, and it will not be moved back when the answer arrives.
  4. Pay earlier than you have to, once, deliberately. A deposit that arrives before the season lets material be bought while it is still available, and it is the clearest signal a small customer can send that an order is real.
  5. Ask for the production finish date weekly, not the shipping date, and ask the planner rather than the sales contact. Weekly contact during a saturated period is not pressure; it is how you find out about a slip while there is still time to do something.

The way out of the queue is to be the customer whose order is easiest to plan, not the one who shouts loudest. Over a year that means a forecast you maintain, a specification you stop changing, a deposit that arrives on its own, and a relationship with the person who writes the schedule rather than the person who signs the quotation.

Questions we get on this.

We paid a deposit and the order is still late. What leverage do we have?

Less than it feels, and more than shouting will realise. The useful move is to find out where the order actually is — which operation it is waiting for and why — because the answer decides whether anything can be done. If the constraint is material that has not arrived, pressure changes nothing; if it is a scheduling choice, offering something concrete in exchange sometimes does: taking a partial shipment, accepting a simplified variant, paying the balance early. Threats you will not carry out cost you the next allocation.

Does a penalty clause get us produced on time?

Rarely, and it has a cost. A factory asked to accept a penalty usually responds by quoting a longer standard lead time, which protects it and makes you later on paper than you were before. Penalties also have to be enforced somewhere, and for a small order the enforcement costs more than the sum. What changes the date is being easy to plan and being visible to the planner; what a clause buys is an argument after the fact.

Is it better to be a small customer at a small factory?

Often, yes, and it is the most underused answer to this whole problem. At a factory where you represent a meaningful share of the year, your order is planned rather than fitted in, the owner knows your name, and a question reaches the person who decides. The trade-offs are real — less equipment, thinner engineering, more fragility if they lose another customer — so the check is whether the constrained operation is done in-house and what happens to you if their largest customer leaves.