Sourcing talk fixates on unit price and lead time. The question that actually decides what a bad order costs you is quieter: when the goods are wrong, whose money is already gone, and who do you have to chase to get it back? That is settled the moment you choose between buying direct from each factory and buying through a partner who contracts on your behalf — long before the first defect appears. This guide walks the payment and liability mechanics in plain terms: how the money moves, what shifts the risk from you to the other side, and the paperwork to hold before you release the balance.
The short answer
- In a direct-to-factory deal you sign with each factory, wire the deposit and the balance, and carry the counterparty risk alone. In a managed model one partner is the contracting party between you and the factories — so you enforce one agreement, not several across a language barrier.
- The payment shape is broadly standard in this trade — a deposit up front, the balance before the goods ship — but the detail that protects you is tying that balance to a passed inspection, not to “goods ready”. Exact splits and triggers are agreed per program, never fixed in the abstract.
- When a batch fails QC the real question is who absorbs the rework, discount or rejection. With one contracting counterparty you hold a single party to the standard; with four factory contracts you negotiate four times, each against a supplier already holding your deposit.
- A partner is not an insurer. It cannot cover your own spec errors, force majeure, or a decision to chase rock-bottom price against advice. What it changes is who carries the ordinary counterparty and coordination risk — and how few parties you must enforce against when something goes wrong.
Two ways to buy, and who you are actually paying
In a direct-to-factory purchase you place an order with each factory and, in almost every case, pay a deposit to start production and the balance before the goods leave the country. From the moment that deposit clears, the counterparty risk is yours and yours alone. If that particular factory underdelivers, ships short or goes quiet, you are the one trying to recover money from a company in another jurisdiction and another language, with production finished and your leverage mostly spent.
A managed or trading-partner model changes one structural thing: the party you contract with. The partner becomes your single counterparty — it places and pays the factory orders, and you place and pay one order with it. You are not signing four factory contracts and reconciling four deposits; you are enforcing one agreement against one company that has agreed to stand in front of the network for you. That coordinating party is what we are: we source and qualify independent factories, mills and dye houses and contract with you as one counterparty — we are not, and do not claim to be, the factory.
This is not automatically cheaper, and it is not a guarantee — later sections are blunt about where it stops. What it changes is the shape of your exposure. Direct buying spreads your risk across several suppliers you must each police from a distance; a managed model concentrates it into one accountable relationship. Which is better depends on your leverage, your language reach and how many evenings you can give to chasing factories. The point is to choose it knowingly, not to discover the difference when a payment has already gone out.
The payment structures you'll actually see
Across this trade the broad shape is remarkably consistent: a deposit to commit the order and reserve production, then the balance around the time the goods are ready to ship. The split you will typically be quoted sits somewhere near a third down and the remainder before shipment, though the exact figures move with order size, material outlay and how long the two sides have worked together. Treat any specific percentage as a starting point to negotiate and put in writing — and note that a partner's final terms are set per program, not fixed in advance of one.
Then the instrument. Most small and mid-sized activewear orders settle by TT — a telegraphic transfer, which is simply a bank wire — because it is cheap and quick. A letter of credit (L/C) is a bank's conditional promise to pay once the seller presents specified documents; it costs more and carries paperwork, so it tends to surface on larger orders, with brand-new counterparties, or where a buyer wants a bank sitting between the money and the documents. For most growing brands the useful question is not which instrument is “safest” in the abstract, but what the payment is actually conditioned on.
The timing detail that matters most is when the balance falls due relative to the bill of lading. In sea freight the bill of lading (B/L) is the document that controls the cargo: whoever holds the original can claim the goods. The balance is normally due before that original B/L is released — which is precisely the leverage point, the last moment your money and the goods sit on the table together. Pay earlier, against nothing firmer than “production finished”, and you have spent that leverage before anyone has confirmed the goods are right.
Why “payment against approved inspection” is the line that matters
“Goods ready” and “goods right” are different claims, and the gap between them is where buyers lose money. Payment against approved inspection means the balance is triggered by a passed quality inspection against an agreed standard — not by the factory's assurance that production is complete. It is the single most useful condition you can attach to the balance, because it keeps your remaining leverage alive until someone has actually looked at the goods.
For that to mean anything the standard has to be named in advance: an inspection level and defect limit both sides accepted before production — commonly an AQL such as 2.5 for major defects — together with the measurement, colour and labelling specs the inspector checks against. “We inspected it”, with no agreed standard behind it, is the same empty reassurance as “our fabric is safe” — it cannot be enforced and it cannot be forwarded. Settle who inspects, on whose behalf, and against exactly what, while you still hold the balance.
In a managed model this condition sits in two places at once. You hold it against the partner — your balance is due on an inspection you accept — and the partner in turn holds it against each factory before releasing their money. That is the concrete value of a counterparty positioned above the factories: it can refuse to pass a failed batch down to you and settle the problem at the factory level, instead of leaving you to argue remedies with a supplier that already has your deposit.
When a batch fails QC, who carries it
A failed inspection has three ordinary outcomes: the factory reworks the goods, the price is discounted to reflect the defect, or the batch is rejected outright. Which one happens is a commercial negotiation, and the party still holding unpaid balance holds the leverage in it. That is the whole reason the timing matters — you want the remedy decided before the balance moves, not after it has gone.
Here the counterparty structure shows its teeth. Buying direct, you negotiate that remedy with the very factory that under-delivered — one that already has your deposit and sits in another legal system and language. Buying through a partner, you negotiate with a single party that agreed to hold the standard, and a competent one will absorb a problem it caused in coordination or push a factory's defect back to that factory, rather than onto you. Ask a prospective partner — us included — how rework, discount and rejection are handled and whose cost they are; a clear, specific answer is itself a signal.
Be realistic about what “the partner carries it” means. It does not mean every loss lands on the partner regardless of cause — a serious partner separates a defect it is accountable for from one written into your own spec. It means you have one accountable party to resolve it with, and a future stream of orders as leverage, rather than a stand-alone factory whose only remaining interest in you is collecting the balance it is owed.
Currency, and who is holding the FX risk
The factories cost in RMB; most US, UK and EU brands pay in USD or euros. Someone carries the gap between those currencies from the day a price is quoted to the day each payment clears, and it is worth knowing who. When you are invoiced and pay in USD, the RMB movement between quote and payment sits on the supplier or partner side; when a supplier prices in RMB, or asks you to settle in it, that exposure shifts onto you.
On a single quick order the swing is usually small. It stops being small on a program that reorders across a year, or on long-lead custom development where months separate the quoted price from the final balance. On those, ask how long a quote is held, what happens to the price if the rate moves materially before you place the order, and which currency the balance is actually payable in. These are ordinary questions, and a supplier who cannot answer them plainly has told you something.
None of this is unique to the managed model, but the model concentrates it usefully: one counterparty, one invoice currency, one place to pin down the currency and quote-validity terms — rather than four factories each pricing, and re-pricing, in their own way. As with the deposit split, treat the specific currency terms as something agreed per program and written down, not quietly assumed.
What a partner cannot protect you from
A partner is not an insurer, and one that talks like an insurer should worry you. Concentrating your counterparty risk into a single accountable relationship changes who carries the ordinary failure; it does not abolish risk, and several whole categories sit outside what any counterparty can honestly absorb.
Three matter most. Your own specification: if the tech pack is wrong, the measurements ambiguous or the approved sample not quite what you meant, the goods can pass every inspection and still be wrong — that is a spec error, not a QC failure, and no payment structure reverses it. Force majeure: port congestion, weather, a factory fire, a raw-material shortage or a policy change can delay or break an order whoever contracts with whom; the paperwork decides who bears the delay, but the event is no one's fault to eat. And price against advice: push past a qualified supplier to a rock-bottom quote a partner has flagged, and you have chosen to inherit that risk — no counterparty can carry a decision you overrode.
Apply that standard to any partner you are weighing, us included. The honest version of this pitch is not “nothing can go wrong”. It is narrower and more useful: fewer parties to enforce against, a balance held until an inspection you agreed to actually passes, and someone accountable for the space between factories. Any claim larger than that is marketing, and you should read it as marketing rather than protection.
The checkpoints to insist on before releasing balance
The balance is your last piece of leverage, so treat releasing it as a checklist rather than a formality. Before the wire goes out, have in hand: an inspection report against the standard you agreed, showing the AQL result and a defect breakdown rather than a bare pass/fail word; dated photographs of the actual goods, cartons and labelling; and the packing list with real carton counts and measurements, since your freight and duty are calculated on them.
Then the shipping and compliance layer: confirmation of the booking or the bill of lading detail, so you know the goods are genuinely moving and to where; and any test or compliance document your market requires — fibre content, restricted-substance or care-labelling evidence in your own name, held by whoever actually produced the article rather than summarised for you second-hand. Where a certificate is involved, verify the holder, scope and validity rather than the logo on the cover.
Finally, the line most buyers skip: who bears the cost if a defect only surfaces after the goods land. A claims or latent-defect window, agreed before the balance moves, is the difference between a conversation you can win and one you have already lost. None of this makes an order risk-free. It makes your position defensible — which, when you do not own the factory, is the part you can actually control.