The registry entry on a Chinese die maker comes back clean: name matches, credit code matches, status active, business scope covers mold manufacturing and includes an import-and-export clause. Then the same registry data shows a micro-enterprise. If you are about to put a rolling annual tooling programme through that company, those two facts sit awkwardly together, and the awkwardness is legitimate — but not for the reason most buyers assume.
The reason is not quality. Small tool shops routinely produce dies that are better than anything a large general-purpose factory would make, because die work is skill-dense rather than capital-dense and because a shop that does one thing for years accumulates judgement that does not scale by hiring. Verifying capacity at a small Chinese tooling workshop is not an exercise in catching a fraud. It is an exercise in mapping a single point of failure — and then deciding, with your eyes open, how much programme you are willing to route through it.
What the registry tells you, and what it stops short of
Registration facts are the cheapest verification available and worth having first. For the worked example in this article — ZLD Precision Mold, a Dongguan drill point die maker listed on CMH — the checkable set looks like this: incorporated September 2019 in Dalang, Dongguan; registered scope covering manufacture and processing of molds, hardware and machinery parts, with an import-and-export clause, which is what lets a company contract, invoice and ship under its own name rather than through a third party; and a registry check in August 2026 in which name, credit code, status and registered address all matched.
That set answers questions about legal existence and about who you are actually contracting with. It answers nothing about throughput. Registry data is not a capacity document, and the honest reading of a micro-enterprise flag is narrow: it tells you that the organisation is small and that scale should be confirmed independently — through a live video walkthrough or an audit — before a large programme is scheduled. It does not tell you the shop cannot do the work.
One practical detail while you are checking: older industry-directory listings for this company still show a previous address. Stale directory entries are extremely common for small Chinese manufacturers, who move premises more often than they update third-party listings, and they generate false alarms in due diligence. Treat the registry entry as the authority and directory sites as unreliable secondary sources — but do ask the supplier directly to confirm the current operating address, because an out-of-date listing and an undisclosed relocation look identical from a desk.
The company also confirms it currently holds no quality-system or inspection certificates, which the CMH profile states rather than papers over. For a shop of this size that is unremarkable and, arguably, more useful than a certificate whose scope you cannot verify. It does mean your assurance has to come from the tooling itself and from the records around it rather than from a document.
The real exposure is continuity, not competence
Reframe the question. Instead of "can this shop make good dies", ask "what happens to my programme in the four scenarios below".
The principal is unavailable. Illness, family obligation, a two-week absence. In a skill-dense shop, one person may hold the grinding judgement that makes the difference between an acceptable die and a good one. Ask directly: if the principal is away for three weeks mid-order, who completes the work, and has that person completed a die of this code before?
A machine goes down. EDM and grinding are sequential; a single machine outage in a small shop is not absorbed by a parallel cell. Ask what the fallback is — a partner shop, a rented slot, a second machine — and whether the shop has used that fallback before or is describing it hypothetically.
Demand spikes. A stated peak monthly capacity above one thousand sets is a company-stated figure, and peak is not sustained. Ask what the sustained monthly figure looks like across a normal quarter, and what share of it your programme would represent. A supplier where you would be forty per cent of output is a different risk profile from one where you would be five per cent, in both directions.
The relationship ends. Whether by your choice or theirs. This is where drawing retention and design ownership stop being paperwork and become the thing that decides whether you can move a die code to a second shop in six weeks or twelve months.
The track record the company reports — a longest customer relationship of over seven years with uninterrupted reorders, and thirty-plus domestic fastener makers on long-term reorders with export deliveries to Eastern Europe, the Middle East, South and Southeast Asia — speaks to the first three scenarios, and it speaks well. Repeat business over years is the best available proxy for a small shop's ability to hold a standard. But these are company-stated figures with case names withheld, no independent capacity audit exists, and they should be treated as a reason to test rather than a substitute for testing. Use them as a hypothesis and check it with the cheapest instrument available, which here is a single-set trial order followed by a repeat of the same code.
Drawings, programmes and who owns the die design
Two questions belong in the contract, and both are open on the published record for this supplier — which makes them questions to ask rather than facts to rely on.
How long are drawings, models and EDM programmes retained after an order? This matters more for dies than for most purchased parts. Reorder number seven has to be dimensionally identical to reorder number one, and what makes that true is not the drawing alone but the electrode design and the machining programme derived from it. A shop that retains the programme per die code can reproduce your die from a code. A shop that retains only the drawing will ask you to ship a worn die back as a reference — and a worn die reproduces the wear. Ask for a stated retention period in writing, in years, and ask whether the retained item is the programme or only the drawing.
Who owns a custom die design built to your drawing? Also unpublished here, and worth settling before you send anything. The commercial questions are: does the design remain yours; may the shop reuse a geometry developed for you on another customer's order; is there exclusivity on a point profile you paid to develop; and can you require the drawings, models and programmes to be transferred or destroyed at the end of the relationship. None of that is unusual to ask, and none of it is expensive to agree at quotation stage — it becomes expensive when raised after the fact. If you paid for the development of a point geometry, exclusivity is reasonable to ask for — but be specific about what it covers, the exact geometry you specified rather than a whole size range, because a demand that is too broad tends to be refused outright. The broader mechanics are covered in protecting IP with a China factory.
Insider detail: ask what the shop stamps on the die. A shop that marks each die pair with a traceable code and serial can tell you, three years later, exactly which drawing revision produced it. A shop that ships unmarked dies is asking you to run your own identification scheme, and most screw plants discover that gap during their first quality investigation rather than during qualification.
Second insider detail: keep a golden die. On any small-shop programme, hold back one unused die pair from an accepted lot as a physical reference standard, in oil, labelled with the code and the acceptance date. It costs one set. It converts every future dispute from an argument into a comparison.
A staged programme that proves capacity before it commits it
Do not decide the whole question at once. Buy the answer in stages, each of which costs little and rules something out.
| Stage | What it costs you | What it proves | What it does not prove |
|---|---|---|---|
| Registry and scope check | An hour | Legal existence, who you contract with, export capability | Nothing about throughput |
| One-set trial order | One die pair, one machine slot | Real geometry, real records practice, real communication | Nothing about volume or repeatability |
| Repeat order of the same code | A second die pair | Whether die two matches die one — the core reorder question | Nothing about parallel capacity |
| Live video walkthrough | A scheduled call | Machines, work in progress, whether the shop is where it says it is | Nothing you cannot see on camera |
| On-site or commissioned audit | A day and a fee | Process, records, redundancy, actual load | Only valid as of the audit date |
| Staged volume ramp | Time | Sustained output under your load | — |
The second row does most of the work and costs the least. The company states it accepts trial orders from a single set with no minimum, which makes a one-set trial the cheapest real test available in this category. Build the stages into the calendar rather than bolting them on — each one consumes lead time, which is why they belong in the backward schedule described in sampling and lead-time planning for a tooling changeover. The third row is the one buyers skip and the one that actually predicts a programme: order the same die code twice, separated by a few weeks, and compare the two sets against each other and against your drawing. Consistency between reorders is the property you are buying from a tooling supplier. A shop can hit a drawing once by effort; hitting it identically on the second try is what a retained programme and a stable process look like from outside.
For the video walkthrough, prepare it like an audit rather than a courtesy call. Ask to see work in progress on other customers' orders — not a staged clean bench — the EDM and grinding machines running, the metrology equipment being used on a real part, and the storage where drawings and programmes live. The one-day inspection logic in what to inspect in a single factory visit transfers almost unchanged to a camera, and the scoring habit in building a supplier scorecard is what stops a good video call from becoming an unstructured impression.
Sizing the programme to the supplier, not the other way round
The last decision is yours alone, and it is the one that actually manages the risk: how much of your tooling demand should sit with a single small shop.
A defensible pattern is to dual-source at code level rather than vendor level. Qualify a second maker on your two or three highest-volume die codes only — the codes where an outage would stop a line — and leave the long tail of low-volume codes with the shop that does them well and cheaply. That gives you continuity where it matters without diluting the relationship into irrelevance, and small suppliers generally understand and accept it when it is stated openly.
Then hold three things: a golden die per critical code, your own copy of every drawing and revision, and a written retention commitment from the shop. With those in hand, a supplier disruption becomes a sourcing task with a known duration rather than an open-ended reverse-engineering project. The general framing sits in China supply chain risk management.
None of this argues against buying from a small workshop. On this factory's own account, its founding team has worked in drill-die tooling since 2000 while the company itself was registered in 2019 — a company-stated position that describes exactly the pattern under discussion: long individual craft experience inside a young, small legal entity. That combination can be an excellent thing to buy. It just needs to be bought with the continuity questions answered rather than assumed.
Common questions
Does a micro-enterprise registry flag mean the supplier is a trading company?
No. The two questions are separate. Scope and shipping behaviour tell you whether a company manufactures and exports under its own name; size tells you how much it can carry. A small registered manufacturer with mold-making and import-export both in scope is a manufacturer that is small — which is a capacity question, not an authenticity one.
How do I verify capacity without flying to China?
A structured live video walkthrough, with an agenda sent in advance and a request to see work in progress rather than a prepared display, gets you most of the way. Follow it with a repeat order of the same die code and compare the two sets. If the programme value justifies it, commission a local audit — but do the cheap steps first, because they often settle the question.
What is the single most useful clause to add to a tooling order with a small shop?
A drawing and programme retention clause with a stated period in years, plus a right to receive the drawings, models and machining programmes on request. It costs nothing to agree at quotation and it is the difference between moving a die code to another shop in weeks rather than months.
What to ask the supplier next
Send these before the programme, not after the first problem:
- If the principal is unavailable for three weeks, who completes an in-progress die, and have they built this code before?
- What is the fallback if the EDM or grinding machine goes down — and has it been used?
- What is the sustained monthly output in a normal quarter, as distinct from peak?
- How long are drawings, 3D models and machining programmes retained after an order closes?
- Who owns a die design developed to my drawing, and may that geometry be reused for another customer?
- What identification is marked on each die pair, and can it be traced to a drawing revision?
- Can we schedule a live video walkthrough with work in progress visible?
The company's registration status, stated terms and the scale flag are all set out together on the ZLD Precision Mold factory profile, which is the right place to start before you send the list above. The die-specific version of the vetting sequence is in how to vet a Chinese drill point die supplier, and what to demand from the shipment itself in pre-shipment trial runs on dies.