What a factory's capacity claim is actually worth

We can produce two hundred thousand pieces a month. It opens most factory profiles and it is one of the least informative sentences in sourcing. It is usually true. It is also a ceiling, measured under conditions that will not apply to your order, and it says nothing about whether your goods will exist on the date you need them.

Most buyers do not have a sourcing problem. Finding factories has never been easier, and a supplier list can be assembled in an afternoon. What they have is a verification problem — and it is the part that gets skipped precisely when the date is tight and skipping it costs the most.

Nameplate capacity and effective capacity

A stated capacity is usually nameplate: lines multiplied by the hours they could run, multiplied by an output rate, at full staffing and perfect yield. It describes a good month that has never happened.

Effective capacity is what emerges after the things that always occur: changeovers between styles, first-pass yield below a hundred percent, absenteeism, a machine down, materials arriving late. The gap between the two is commonly twenty to thirty percent, and it widens for complex products with frequent style changes relative to a single line running one item continuously.

Neither figure is the one you need. The relevant number is a third: uncommitted effective capacity in your production window, on a line capable of your specification, after the orders already accepted.

The questions that turn a claim into a number

• How many lines are capable of this specification — not how many lines exist.
• What is the output per line per day for this product, at the quality level in the sample?
• What is the first-pass yield on this construction, and what happens to the rejects?
• How long is a changeover, and how many different styles will be running in my production month?
• What proportion of that month's capacity on those lines is already committed?
• Where does my order sit in the sequence, and what would move it?
• What is the materials lead time behind the line, and are those materials in stock or still to be bought?

A supplier who intends to deliver will answer most of these in a single call. The answers do not have to be flattering — a factory saying it can give you sixty percent of what you asked for, in the week after you wanted it, is more useful than one agreeing to everything. What the answers have to be is specific.

Capability is a separate failure

Capacity and capability produce the same symptom from different causes. A factory can have the volume free and still be the wrong choice: no machinery for a particular construction, no experience with a finish or a substrate, no certification the destination market requires, no laboratory relationship to support a test report.

This is what samples are for, and it is why sample approval belongs before the purchase order rather than after it. A counter-sample made from the actual materials, on the actual line, at the actual quality level is the only evidence that survives contact with production. A golden sample produced by the development team on a bench proves that somebody in the building can make one.

Verifying without standing on the floor

A factory visit is the strongest evidence available and the least available evidence. Most of what a visit would tell you can be obtained from documents, and should be, whether or not anybody travels.

• The business licence and its registered scope. A trading company presenting a factory's photographs is common, and it is immediately visible in the registration.
• The machine list, with counts and ages, read against the specification you intend to run.
• Certifications relevant to the market and the product, verified with the issuing body rather than accepted as a PDF.
• Recent test reports for comparable goods, issued to that factory rather than to one of its customers.
• Two references in your own market — and the questions to ask them are about a date and a defect rate, not about whether the factory was pleasant to deal with.
• A third-party audit where the order value justifies it. It costs a fraction of one container and it is the only independent evidence on this list.

The queue matters more than the ceiling

The most common way a capacity claim fails is not that the factory could not make the goods. It is that it could, and it made somebody else's first.

Position in a production queue is decided by the size of the customer, the length of the relationship, whether the deposit has actually landed, and how much noise a delay will generate. A first order of moderate size from a new buyer sits at the bottom of every one of those lists, and no contract clause changes that on its own.

One factory is a single point of failure

Everything above assumes the question is whether this supplier is the right one. There is a separate question worth asking alongside it: what happens if the supplier is right and something goes wrong anyway — a fire, a compliance suspension, a labour dispute, an export restriction, or a factory that simply stops replying.

For a first order, single sourcing is reasonable; the volume does not justify anything else. For a line that becomes recurring, it stops being reasonable quickly. The usual answer is not to split every order down the middle, which throws away the leverage that volume buys. It is to qualify a second factory against the same specification — sample approved, documentation held, terms agreed — and give it a small share, enough to keep the relationship alive and the process proven.

That costs a few percent on the blended price. Not doing it costs nothing at all until the day the primary supplier cannot ship, at which point qualifying a replacement takes the twelve weeks you do not have.

None of this requires leverage or a travel budget. It is a set of questions asked before money moves rather than after a shipment is late. The suppliers worth having answer them readily. The ones that do not have told you something useful, and they told you while it was still cheap to hear.