Most foreign hardware companies negotiate the price of a Chinese manufacturing programme before they negotiate the risk. The price is visible, comparable across three quotations and easy to argue about. The payment schedule is where the risk actually sits, and it is usually accepted in whatever form the factory put on its proforma invoice.
That schedule is not an administrative detail. It is a credit agreement, a performance bond and an acceptance mechanism compressed into four or five lines of a document that nobody reads closely until the goods arrive wrong. A buyer who pays 40 per cent on signature and the balance before shipment has lent the factory six weeks of working capital and holds nothing in return except a promise and an email thread.
The conclusion worth carrying into every negotiation is this: build the schedule backwards from the evidence you can obtain. Each tranche should be released against something you can verify before the money leaves your account, not against a stage the factory reports by telephone. A payment term that cannot be evidenced is a payment term you will lose control of. Everything below elaborates that principle for production, tooling, acceptance, retention, bank instruments and the awkward case where the factory will not move without the whole sum in advance.
What a payment schedule is actually allocating
Four risks move between the parties every time a tranche is agreed, and they do not move together.
The first is performance risk: the chance that the factory takes the money and does not produce. The second is quality risk: the chance that it produces something that does not conform to the specification. The third is timing risk: the chance that it produces late, after your retail window or your funding milestone has passed. The fourth is recovery risk: the chance that, having lost on one of the first three, you cannot get the money back.
A deposit shifts the first three risks onto you and reduces the factory’s recovery risk, because it no longer has to chase payment. A balance paid before shipment shifts quality risk onto you, because the goods leave before you have seen them. A retention shifts a slice of quality risk back, but only for the duration of the retention. An instrument such as a letter of credit changes the mechanism without changing the allocation: the factory still gets paid on documents, and you still pay for goods you have not touched.
The mistake foreign buyers make is to treat the percentage split as the whole negotiation. Thirty/seventy is a convention, not a risk model. What matters is what each tranche is secured against, what evidence releases it, and what happens to the money if the evidence turns out to be false. A 30 per cent deposit released against a signed tooling report and a factory audit is safer than a 20 per cent deposit released on signature to an entity nobody has verified.
There is a second allocation hidden in the schedule, and it is the one finance teams notice. Between the deposit and the balance, somebody is funding the work in progress. If the deposit is smaller than the material cost of the order, the factory is financing your programme, and it will price that financing into the unit rate or quietly deprioritise your line when a better-funded customer arrives. If the deposit is much larger than the material cost, you are financing the factory’s working capital and carrying the exposure. Sizing the deposit against material content, rather than against habit, is the single most useful piece of arithmetic in the negotiation.
Chinese law supplies default rules for the arrangement, but they are thin where payment is concerned. Whether the deal is characterised as a contract for work or a sale of goods determines which default provisions on inspection, risk and payment fill the gaps, and most OEM programmes are a mixture of both. The contract should state which regime applies, because the defaults on inspection and acceptance are not identical and neither set of rules will guess what you intended.
The instruments, ranked by what they protect
The menu is small. What varies is how much protection each item buys and what it costs in fees, time and factory goodwill.
| Instrument | What it protects | What it does not protect | Practical constraint |
|---|---|---|---|
| T/T (bank transfer) | Nothing once sent | Quality, delivery, fraud | Irreversible; bank details must be verified out of band |
| T/T against documents | A little, if the documents are originals | Quality, if the documents do not prove it | Factory must agree to hold originals until payment |
| Letter of credit at sight | Payment against compliant documents | Quality, unless an inspection certificate is a required document | Fees, credit line, discrepancy risk, factory familiarity |
| Usance L/C | Same as sight, with deferred payment | Same as sight | Factory prices the deferral into the rate |
| Standby L/C or advance payment guarantee | The deposit, if the issuer is sound | Performance after the guarantee expires | Small factories cannot obtain one |
| Platform escrow | Both sides, within the platform’s rules | Disputes the platform does not hear well | Only for orders placed through the platform |
| Bank escrow | Both sides, on terms the bank will administer | Nothing beyond the stated release conditions | Banks dislike conditional release; documents must be simple |
| Documentary collection (D/P) | Title, in theory | Quality and timing | Goods can arrive before documents; port storage costs |
| Open account | Nothing | Everything | Only for established partners with a track record |
Telegraphic transfer remains the default for production in China, and for most repeat orders it is the right instrument. It settles in two to five business days through the SWIFT network, costs somewhere between thirty and eighty US dollars per leg depending on the banks involved, and is understood by every factory. Its weakness is that it is final. Once the funds are credited, the only remedy is a contractual claim, and a contractual claim against a factory in another province is a slow and expensive thing.
A letter of credit converts the payment obligation into a documentary one. The issuing bank pays against a compliant presentation, which means the quality of your protection depends entirely on the document list you write into the credit. If the required documents are an invoice, a packing list and a bill of lading, the bank will pay for a container of bricks. If the required documents include an inspection certificate issued by an inspector you nominate, and a certificate of conformity signed by the factory, the bank will not pay until those documents exist. The credit is governed by the ICC’s uniform customs and practice, so the rules on discrepancies are predictable, but predictability cuts both ways: a missing comma in a description of goods is a discrepancy, and a discrepancy means the factory does not get paid on presentation.
Letters of credit are usually worth considering above roughly fifty thousand US dollars, and for first orders with unfamiliar suppliers. Fees commonly run from a quarter of one per cent to one and a half per cent of the value depending on the bank, the tenor and whether confirmation is required. Many small and medium factories will decline an L/C, not because they distrust you but because they have no export documentation function and would rather wait for a wire than assemble a compliant presentation.
Escrow splits the difference. The money sits with a third party and is released when the stated conditions are met. Platform escrow, of which Alibaba Trade Assurance is the most widely used example, works well for orders placed through that platform and poorly for anything outside it. Bank escrow is more flexible in theory and more rigid in practice, because banks prefer release conditions they can check mechanically, such as “on presentation of a signed acceptance certificate”, rather than conditions requiring judgment.
Documentary collection, where the bank releases documents against payment, sounds protective and usually is not. The goods travel independently of the documents. If you refuse to pay because the quality is wrong, the container sits at the port accruing storage and demurrage while you argue, and the factory still holds the goods. Documents against acceptance, where the bank releases documents against your promise to pay later, gives you the goods and the factory nothing but a trade debt. It is rare in China for good reason.
Western Union and money transfer services should not be used for supplier payments at any size. There is no buyer protection, no audit trail worth the name and no recourse. They are a standard vector for the impersonation fraud described later.
The standard production schedule, tranche by tranche
The market standard for production is a 30 per cent deposit on order confirmation and a 70 per cent balance before shipment, paid by T/T. That is where most negotiations start, and for a repeat order with a factory you know it is a reasonable place to end.
What the convention does not tell you is what each tranche is for, and that is what you should be negotiating.
The deposit exists to fund materials. A factory that receives 30 per cent on a product whose bill of materials is 55 per cent of the price is being asked to fund the difference until the balance arrives. Some will, at a price. Others will buy cheaper material, substitute a component, or schedule your line behind a customer who paid more. If your product is material-heavy, expect the factory to ask for 40 or 50 per cent, and understand that the request is commercial rather than predatory.
The balance exists to protect the factory against your non-payment after it has committed the work. That is why factories insist on payment before shipment: once the container is on the water, their leverage is gone and yours begins. The natural compromise, and the one worth pushing for, is payment against a pre-shipment inspection report rather than payment against a booking confirmation. The factory gets paid before the goods leave. You get a last look before the money moves.
A workable schedule for a first or second order looks like this.
| Tranche | Typical share | Release trigger | Evidence to hold |
|---|---|---|---|
| Tooling deposit | 50 per cent of tooling cost | Tooling purchase order signed, drawings approved | Signed drawing set, tooling PO, tooling register entry |
| Tooling acceptance | 40 to 50 per cent of tooling cost | First-shot samples measured and approved | Dimensional report, sample approval sheet, photographs |
| Production deposit | 30 to 40 per cent of order value | Production PO signed, material lead times confirmed | PO, material purchase evidence for critical components |
| Mid-production payment | 0 to 30 per cent | Agreed completion milestone verified on site | Inspection report at stated percentage complete |
| Pre-shipment balance | 50 to 60 per cent | Pre-shipment inspection passed | Third-party inspection report, packing list, photographs |
| Retention | 5 to 10 per cent | Fixed period after arrival, no claims | Arrival confirmation, defect log, release certificate |
The three-stage variant, 30 per cent deposit, 30 per cent at mid-production and 40 per cent after final inspection, is common on longer builds and on custom components with long material lead times. It costs more in inspection fees and administrative attention, and it is worth it when the order is large enough that a total loss would matter.
Higher deposits of 40 to 50 per cent are normal for custom components, tooling-intensive parts and very small quantities where the factory’s setup cost dominates. A 20 per cent deposit is achievable with an established relationship, a large order or a product with low material content. It is not a target to insist on in a first negotiation, because a factory that accepts a deposit below its material cost will recover the difference somewhere else.
Whatever the split, the schedule should be expressed as conditions rather than dates. “Seventy per cent before shipment” tells nobody what has to happen. “Seventy per cent within three business days of the buyer’s written acceptance of the pre-shipment inspection report, and in any event not later than the loading date stated in the shipping instruction” tells both parties, and a tribunal, exactly what was agreed.
Tooling: the schedule that runs before production
Tooling has its own economics and its own schedule, and mixing the two is a common source of dispute. A mould is not a batch of product. It is a capital asset that the buyer usually pays for and frequently does not end up owning, and the payment milestones should follow the engineering milestones rather than the production calendar.
The simple structure is 50 per cent on order and 50 per cent on sample approval. The more careful structure is 50 per cent on order, 40 per cent on first-shot samples and 10 per cent on production qualification, sometimes described as 50/40/10. The three-stage version exists because first-shot samples prove that the mould opens and closes and produces a part, while production qualification proves that the part can be made repeatedly to tolerance at rate. Those are different achievements, and the second is the one that matters.
The first payment should be released only after you have approved the mould design drawings, not merely the purchase order. Design approval is the last cheap moment to change the gate position, the ejector layout or the parting line. Once steel is cut, changes are expensive and slow, and the factory will charge for them.
The second payment should be released against a dimensional report on the first-shot samples, measured against the drawing, with the critical dimensions identified in advance. A sample that “looks right” is not evidence. A report showing which dimensions are in tolerance, which are marginal and which are out, signed by the factory’s quality function and checked by your engineer, is evidence.
The final payment should be released against a signed sample approval sheet and, where the volume justifies it, a short production run at rate. The run tests the mould, the process window, the fixture and the operator instructions together. A mould that produces one good part in an afternoon and a mould that produces four hundred good parts a day are different assets.
Tooling costs run from a few thousand dollars for a simple single-cavity part to hundreds of thousands for a complex multi-cavity tool with hot runners, slides and tight tolerances. The payment schedule should scale with the complexity of the acceptance evidence, not with the size of the invoice. A two hundred thousand dollar mould released on a photograph of a sample is a two hundred thousand dollar exposure taken on trust.
Two structural points belong here because they are paid for at the same time as the tooling. The first is the test programme: the test fixture, the test limits and the software that runs the sequence. A buyer who owns the mould but not the test programme has bought a shape, not a product. The second is the process documentation: the parameters, the gauges, the fixture drawings and the maintenance schedule. Payment for the final tooling tranche is the natural moment to require delivery of both, because it is the last moment at which you have leverage.
Tooling amortisation is the alternative to direct payment. The factory spreads the tooling cost across the unit price over an agreed volume, and ownership passes to you when the amortisation is complete. It reduces upfront capital and raises the unit price. The risk is that the amortisation schedule and the ownership trigger are never written down, and the factory treats the tool as its own equipment because it funded the toolmaker. If you amortise, the contract needs a schedule showing the amortised amount per unit, the cumulative total, the volume at which ownership transfers and what happens to the unrecovered balance if the programme ends early.
Acceptance: the trigger that decides when the money moves
Most payment disputes are acceptance disputes wearing a different hat. The buyer says the goods were defective. The factory says the buyer accepted them. Both are quoting the same clause, which says something like “balance due on acceptance”.
Acceptance needs three things defined, and usually none of them is.
The standard comes first. Acceptance should be measured against a specification that is a controlled document, with tolerances, test limits and a defect classification. The classification matters more than the standard itself, because it decides what happens to a batch with a cosmetic flaw. A three-tier classification into critical, major and minor defects, with an acceptable quality limit set for each, converts an argument into an arithmetic exercise. Sampling plans of the kind published as international standards allow a defined number of minor defects per sample size; the contract does not need to reproduce the tables, but it should adopt one by reference.
The method comes next. Who inspects, where, with what equipment and against what sample size. A pre-shipment inspection by an independent third party at the factory, five to seven working days before the planned departure, with a written report issued within twenty-four hours, is the practical standard. The inspector needs the specification, the approved sample and the packing specification before arriving, and the factory needs notice of the inspection window.
The consequence is the part that is usually missing. What happens when the inspection fails. The contract should give the factory a cure period, require re-inspection at the factory’s cost where the failure is the factory’s, and state what happens if the cure period expires without a pass: rejection of the lot, rework at the factory’s cost, or termination of the order with refund of the deposit. Without a stated consequence, a failed inspection produces a negotiation rather than a remedy.
There is a structural trap here that catches buyers who draft from a Western template. Under the usual shipping terms, risk passes when the goods are loaded, not when they arrive. A payment condition tied to “acceptance on arrival” therefore asks the factory to carry a risk it did not agree to, and it will either refuse or price the risk in. The cleaner structure separates the two questions. Risk transfers on shipment, as the shipping term provides. Payment is conditional on a documentary event that happens before shipment, namely a passed inspection. If you genuinely need protection after arrival, buy it explicitly, through a retention or a warranty holdback, rather than by stretching the meaning of acceptance.
Deemed acceptance clauses deserve a look for the same reason. A clause that says goods are deemed accepted if the buyer does not object within a stated period protects the factory against a buyer who sits on a defect list. It also, if the period is short and the notice requirements are strict, allows a defect that appears in the field after thirty days to fall outside acceptance. The period should be long enough to cover installation and first use, and the notice requirement should be a written notice identifying the defect, not a formal claim in a prescribed form.
Retention and warranty holdback
A retention is the simplest quality instrument available and the one most often left out. The buyer withholds a percentage of the price, typically five to ten per cent, for a fixed period after arrival, usually thirty to sixty days, and releases it if no claim is made.
It works because it keeps money in the relationship after the goods have shipped. The factory has an incentive to answer the telephone when a defect appears in the first month, because it is still owed something. The amounts are small enough that the factory will usually accept the principle, and large enough to fund a replacement batch of the defective units.
Three details decide whether a retention functions or becomes a permanent argument.
Whether the retained sum is a separate invoice or a deduction matters more than the percentage. If it is a deduction, the factory’s accounts team will treat the balance as paid and the retention as a receivable it chases. If it is a separate invoice payable on a stated date subject to claims, both sides’ records agree on what is outstanding.
The release mechanism decides whether the retention works. “Released on expiry of the retention period” is not enough. It should be released on expiry unless the buyer has notified a claim, in which case the undisputed portion is released and the disputed portion is held until resolution. A clause that lets the factory sit on the whole retention because of a two hundred dollar dispute is a clause that will be used.
Set-off is the clause that gives the retention teeth. The contract should give the buyer an express right to set off warranty costs, rework costs and the cost of substitute supply against sums otherwise payable, including the retention and any later order. Without a set-off clause, the buyer’s remedy for a defective batch is a separate claim, and separate claims are what factories expect to negotiate away.
A warranty holdback is a longer version of the same idea, running for the warranty period rather than thirty or sixty days. It is harder to negotiate and more appropriate to capital equipment, where the failure mode appears after months of use rather than days. On consumable hardware the retention is usually enough.
When the factory insists on full prepayment
Sooner or later a factory will ask for the whole sum before it starts. Sometimes the request is reasonable. Often it is a signal.
The first step is to find out which of five situations you are in, because each has a different answer.
The factory may be cash-constrained. Small factories in the component supply chain run on thin working capital and finance themselves through customer deposits. If the factory genuinely cannot buy material without your money, the fix is not to refuse but to reduce what you are funding. Pay the material cost directly to the material supplier under a tripartite arrangement, with the factory’s purchase order as the basis, and pay the processing element on delivery. You fund what the factory cannot fund and you do not fund its margin in advance.
The factory may be a trading company rather than a manufacturer. A trading company that takes 100 per cent upfront is using your money to pay its own supplier, and it carries no balance sheet risk. Verify the business licence, the registered address, the export licence and the factory audit. If the entity that invoices is not the entity that manufactures, the credit risk is the trading company’s, not the factory’s, and the pricing should reflect that.
The factory may be pricing risk rather than cash. A new customer with an unfamiliar specification, an unusual material or a tight tolerance is a risk to the factory too, and a full prepayment is how it avoids carrying that risk. The answer is to reduce the factory’s risk in a different way: a smaller first order, a simpler first article, a longer lead time, or an inspection regime the factory trusts.
The request may be a fraud indicator. If the factory wants the whole sum to a personal account, or to an account whose name does not match the contracting entity, or to a “collection agent” in a third country, the correct response is to stop. The same applies where the bank details change by email shortly before payment, which is the most common impersonation pattern in China trade.
The factory may simply be testing you. A demand for 100 per cent that softens to 50 per cent when you say no is a negotiating position, not a policy.
If you decide to accept a full prepayment, accept it for a defined exposure rather than for the programme. Pay 100 per cent for a pilot lot of two hundred units, inspect it, and move to a staged schedule for the production order. The pilot lot costs you the exposure on two hundred units, which is a number you can absorb, and it buys you the evidence you need to argue for better terms. What you should not do is pay 100 per cent on a first production order of twenty thousand units because the factory says that is how it works with new customers. It is not how it works with customers it expects to keep.
There is a middle path that is underused. A letter of credit at sight gives the factory the certainty of payment it says it needs, because a compliant presentation is as good as cash from a bank it recognises, while giving you documentary control over what is shipped. The factory’s objection is usually administrative rather than commercial: it has never done one, or it had a bad experience with a discrepancy. Offering to pay the issuing bank charges, or to use a bank with a presence in its city, removes most of the objection. Where the factory cannot obtain an advance payment guarantee from its own bank, an L/C is often the only instrument that gives both sides something.
Walk away when the request comes with any of the following: refusal to sign a written contract, refusal to allow pre-shipment inspection, refusal to identify the manufacturing site, insistence on a personal or third-party account, or a price materially below every other quotation. Any one of those is a reason to slow down. Two together are a reason to stop.
The sequence, and who does what
Payment sits inside a longer sequence, and most payment failures are failures of sequencing rather than of drafting. The table below is the order in which the steps should happen, who owns each one, and the point at which foreign buyers typically skip ahead.
| Step | Responsible party | Typical timing | What foreign companies miss |
|---|---|---|---|
| Supplier verification | Buyer, with local counsel and an audit firm | One to three weeks, before any payment | The invoicing entity is not the manufacturing entity |
| Contract and payment schedule signed | Both parties, with counsel | Before the first transfer | A proforma invoice is treated as the contract |
| Tooling order and deposit | Buyer’s finance team | On tooling PO | No tooling register, no title clause |
| First-shot samples and dimensional report | Factory quality, buyer engineering | Three to six weeks after steel cut | Samples approved verbally, report never signed |
| Production qualification run | Factory, buyer engineering | One to two weeks | Acceptance criteria not written before the run |
| Production deposit | Buyer | On production PO | Deposit sized by habit, not by material content |
| Material and sub-tier verification | Buyer or inspector | Before assembly starts | Substitutions discovered at pre-shipment |
| Mid-production check | Third-party inspector | At the agreed percentage complete | Inspection booked after the line has moved on |
| Pre-shipment inspection | Third-party inspector | Five to seven days before departure | Inspection booked too late to cure a failure |
| Balance payment | Buyer | After the inspection report is accepted | Paid against a booking confirmation or a photograph |
| Shipment and documents | Factory and forwarder | On confirmation of payment | Consignee and description errors in the bill of lading |
| Retention release | Buyer | Thirty to sixty days after arrival | No set-off right, no claim log |
Two sequencing points are worth separating from the table. The first is that the contract must be signed before the deposit, not after. A factory that asks for a deposit to “hold the slot” while the contract is with its lawyer is asking you to fund an obligation that does not yet exist. The second is that the inspection must be booked early enough that a failure can be cured before the vessel sails. An inspection two days before departure finds defects that cannot be fixed and converts a quality problem into a commercial argument about who pays for air freight.
Where foreign companies get this wrong
Five failure modes account for most of the losses, and each has a recognisable consequence.
Treating the proforma invoice as the contract. The proforma is a quotation with bank details attached. It usually says nothing about acceptance, defect classification, retention, title to tooling, late delivery or governing law. When the goods are wrong, the buyer discovers that the only document it has describes the goods and the price, and the factory’s standard terms, printed on the back of the order confirmation, govern everything else. The consequence is a dispute conducted on the factory’s paper.
Paying to an account that does not match the contracting party. A Hong Kong account in the name of a group company, a “collection agent” in Shenzhen, or an account opened last week in the name of a person are all common. They may be legitimate, particularly where the factory uses a Hong Kong entity for dollar receipts, but they must be documented and verified out of band. The consequence of getting this wrong is that the money reaches a party that is not liable under the contract, and recovery requires a claim against a different entity in a different jurisdiction.
Releasing the balance against shipment evidence rather than quality evidence. A booking confirmation, a photograph of a loaded container or a forwarder’s email proves that something is moving. It does not prove that the right thing is moving, or that it works. The consequence is that the buyer’s only leverage is spent before the first unit is tested.
Writing a letter of credit without a quality document. The credit is only as good as its document list. If the list does not include an inspection certificate from an inspector the buyer nominates, the bank will pay against a clean bill of lading for a container of non-conforming goods, and the buyer will have paid for both the goods and the L/C fees.
Sizing the deposit by convention. A 30 per cent deposit on a product with 60 per cent material content leaves the factory funding your order. It will recover the cost, either by substituting materials or by scheduling your line second. The consequence is a quality or timing problem that looks like bad luck and is in fact a financing structure.
A sixth, less dramatic, failure is worth naming because it is the most common of all: paying late. A buyer who pays on day nine when the contract says day five, twice, will find that the factory’s next quotation is higher and its lead time longer. Payment reliability is the cheapest form of leverage a foreign buyer has, and it is squandered more often than the deposit is.
Cost and commercial impact
The cost of a payment structure is not the sum of the bank charges. It is the cost of the instrument, the cost of the capital tied up, the cost of the inspection regime and the price the factory adds for the risk it carries.
Bank charges are the visible part. A telegraphic transfer costs tens of dollars per leg, with intermediary banks sometimes deducting a further amount, which is why the instruction should state whether charges are borne by the sender, the beneficiary or shared. A letter of credit costs a fraction of a per cent of the value, plus amendment fees, plus the cost of the credit line it consumes. An escrow service charges a percentage or a flat fee depending on the platform. None of these is large relative to the order, and all of them are trivial relative to the loss they are designed to prevent.
The cost of capital is less visible and usually larger. A deposit paid eight weeks before shipment is eight weeks of working capital that the buyer cannot deploy elsewhere. A retention held for sixty days is two months of receivables. On a programme with a twenty per cent gross margin, the financing cost of an aggressive deposit schedule can consume a meaningful share of the margin, which is why the schedule is a commercial term and not a treasury detail.
The price the factory adds for risk is the part nobody quotes. A factory asked to accept a twenty per cent deposit, a sixty-day retention and payment after arrival will quote a higher unit price than the same factory asked for thirty/seventy before shipment. The difference is not greed. It is the cost of carrying your risk, and it is usually cheaper for the buyer to carry the risk itself through an instrument than to buy it back through the price.
Inspection costs scale with the regime. A single pre-shipment inspection on a small order may cost a few hundred dollars; a programme with mid-production checks, first-article inspection and a qualification run costs more, and the cost is fixed while the order value varies. On small orders, the inspection can approach the value of the goods, which is why the regime should be proportionate: full inspection on the first order and the first run of a new revision, sampling thereafter.
Currency adds a line. If the contract is in dollars and the factory’s costs are in renminbi, the factory carries the exchange risk and will either price a buffer or ask for a currency adjustment clause. A clause that shares the movement inside a band and adjusts outside it is usually acceptable to both sides and cheaper than a buffer. Where the buyer has no natural renminbi revenue, forward cover for the payment dates is a treasury decision that should be taken when the schedule is agreed, not when the invoice arrives.
Practical checklist
- Verify the contracting entity, the manufacturing site and the bank account separately, and confirm the account details by telephone to a number you already hold.
- Sign the contract before the first transfer, and attach the specification, the defect classification and the payment schedule as controlled documents.
- Size the deposit against material content, not against the 30 per cent convention.
- Express every tranche as a condition with an evidence standard, and name the document that releases it.
- Put a tooling register, a title clause and a removal right in the tooling agreement, and pay the final tooling tranche against the test programme and process documentation as well as the sample.
- Book pre-shipment inspection five to seven working days before departure, and make the inspection report the release condition for the balance.
- Include a retention of five to ten per cent with a separate invoice, a claim mechanism and an express set-off right.
- Where an L/C is used, make the inspection certificate a required document and check the draft presentation against the credit before the factory ships.
- Refuse personal accounts, third-party collection agents and mid-stream bank detail changes, whatever the explanation.
- Pay on time, every time. It is the cheapest negotiating leverage available and the easiest to lose.
Payment terms in Chinese manufacturing are a risk allocation exercise dressed as an accounting convention. The 30/70 split is a starting point, not a standard, and the numbers matter far less than the evidence that releases them. A buyer who can show what each tranche is for, what document releases it and what happens when the document is missing will win most of these arguments before they start.
For the contract architecture that holds the schedule together, see our commercial contracts service and the guide to manufacturing contracts in China. Tooling payment and ownership are treated in more detail in Tooling Cost in China, and the quality regime that the acceptance clause depends on is set out in Warranty, Quality Guarantee and Defect Liability. Buyers still selecting a supplier should start with How to Select and Vet Factories in China.
Frequently Asked Questions↑ Back to article
What are the most common payment terms with Chinese factories?
The most common payment method is Telegraphic Transfer (T/T), typically structured as 30% deposit upon order confirmation and 70% balance before shipment. For larger orders or new relationships, Letters of Credit (L/C) at sight are also used. Escrow services like Alibaba Trade Assurance are common for smaller orders or first-time transactions.
Is a 30% deposit standard for manufacturing in China?
Yes, a 30% deposit is the industry standard for production orders in China. This deposit covers raw material procurement and initial production costs. For tooling and molds, a 50% upfront deposit is more typical, with the remaining 50% paid upon sample approval. Deposits can range from 20-50% depending on product complexity, order size, and relationship with the factory.
Are letters of credit commonly used in China?
Letters of Credit (L/C) are used in China manufacturing, primarily for larger orders (typically above $50,000) or when working with new suppliers for the first time. Sight L/Cs are more common than usance L/Cs. However, many smaller and medium-sized factories prefer T/T over L/C due to the simpler process and faster access to funds.
How can I protect my deposit with a Chinese factory?
To protect your deposit, verify the factory's legitimacy through business license checks and third-party audits, use a detailed written contract with clear milestones and quality specifications, tie payments to inspection-verified milestones, consider using escrow or L/C for first orders, start with a smaller trial order, and conduct pre-production and mid-production inspections before releasing final payment.
What payment method is safest for new suppliers?
For new suppliers, the safest payment methods are Letters of Credit (L/C) at sight or reputable escrow services like Alibaba Trade Assurance. L/Cs provide bank-level security but can be complex and expensive. Escrow services are simpler and protect both buyer and seller. If using T/T, keep the deposit at or below 30% and always inspect goods before releasing the balance payment.
What is the typical payment structure for tooling in China?
For tooling and molds, the most common payment structure is 50% upfront deposit upon order placement, with the remaining 50% paid upon sample approval. Some factories use a three-stage structure of 50/40/10 — deposit, after T1 samples, and after production qualification. Tooling deposits are higher than production deposits because the factory must purchase materials and commit engineering resources before any parts are produced.
How can I negotiate better payment terms with Chinese factories?
Better payment terms come with relationship building and volume. Start with a smaller trial order on standard terms, pay on time consistently, and gradually negotiate as trust builds. Larger order volumes give you more leverage to negotiate lower deposits or extended payment terms. Other concessions like longer lead times, upfront tooling payment, or annual volume commitments can also be traded for better payment terms. Never push too hard on payment terms at the expense of quality or relationship.