Last updated: 2026-09-04 · 6 min read

How Do You Vet a Chinese Mold Factory Before the Deposit? — Evidence Checks That Replace Trust

Vetting an overseas mold supplier is not about finding one that promises to be good — it is about collecting 5 verifiable artifacts that make a bad outcome expensive for them and easy for you to catch. Words cost nothing; a named engineer’s DFM signature, a live video tour, a full-dimension report with gauge R&R, a staged payment tied to measurement and a trial video with a mold number are evidence you can independently check.

Key Facts

A live video shop tour at a scheduled hour — not a pre-recorded loop — separates a working mold shop from a trading office, because a middleman cannot show live clamping units, an EDM row and a CMM on request.

A comparable full-dimension report with English GD&T and actual readings is the single strongest signal of quality culture — ask them to redact the customer name while leaving the measurement columns and pass/fail flags intact.

A staged payment tied to dimensional acceptance means the supplier’s incentive is to pass your measurement, not to charm the sales call; a meaningful final tranche released only on a passing report aligns their cash flow with your outcome.

Warpage — Part deformation after ejection caused by uneven shrinkage. Amorphous resins warp less (0.4–0.7% shrinkage). Semi-crystalline resins warp more (1.0–2.5% shrinkage). Fix: uniform wall thickness, balanced cooling, optimized gate location, adequate packing pressure. The related guide covers adjacent topics.

Which Evidence Actually Matters?

Five artifacts, each with what it proves and how to verify it:

Ask For	What It Proves	How to Verify
Live video shop tour	They own the machines, not broker the job	Schedule a time, ask to pan to nameplates and read tonnage
Machine and gauge list	Real clamping range, EDM wire, CMM capacity	Cross-check make and tonnage on the call
Comparable full-dimension report	They measure to drawing and can document it	Redact client, keep GD&T and readings visible
Signed DFM on your part	A named engineer reviewed your actual geometry	Prefer a report over a verbal reassurance
Trial video with mold number	Their claimed shot actually ran in their tool	Timestamp and mold ID visible in frame

The order is deliberate: start with what is hardest to fake and cheapest to request. A live video tour at an agreed time instantly separates a working shop from a trading office renting credibility — you can ask the camera to walk to a specific machine, read the tonnage plate, then show the CMM and confirm it is calibrated and used rather than decorative.

The comparable full-dimension report is the strongest signal of quality culture. A factory that routinely issues an itemized, English-GD&T report with actual readings against the drawing is one that measures parts as a habit. You do not need their past client’s name to learn from it — ask them to redact the customer and the proprietary geometry while leaving the measurement columns, the tolerance stack and the pass/fail flags intact.

How Do You Structure the Deal to Protect Yourself?

Four rules, in order of importance:

Tie payments to verifiable milestones, not the calendar. The classic exposure is a front-loaded deposit that leaves the supplier with little reason to fight for your dimensional acceptance. Structure the money so a meaningful final tranche is released only when a full-dimension report passes against your drawing, and an intermediate one only at agreed gates like DFM sign-off and T1. A supplier who genuinely builds good parts will accept evidence-gated payments; the ones who resist are telling you where their real risk lies.
Put ownership and data rights in writing before the first cut. The mold is your asset even when it sits in their building. The agreement must state that the tool, its drawings, its CNC programs and its spare-parts list are yours, and that they will not use your tool or geometry for another customer. Specify what happens to the mold if the relationship ends — release terms, condition standards, and a documented storage-and-maintenance log you can audit remotely.
Define acceptance as measurement, not impression. Before money moves, agree exactly what “done” means: which dimensions are critical, to what tolerance, measured on what gauge, reported in what format, and how many shots must pass consecutively. A vague “customer satisfied” clause lets a supplier ship marginal parts and argue aesthetics; a numeric acceptance standard lets a third party — or you — settle any dispute over a CMM file.
Run a small paid test project before the real tool. Commission an inexpensive soft tool or a single-cavity insert for one representative part and take the whole relationship through a full cycle — DFM, T1, dimensional report, change order, shipping. You will learn more about their responsiveness, documentation discipline and honesty in that one project than in a hundred sales calls.

Runner system — The channel network that delivers melt from the machine nozzle to each cavity. Cold runner: solidifies between shots, produces runner waste. Hot runner: stays molten, eliminates runner waste but adds $15k–$40k to mold cost. The related guide covers adjacent topics.

You are not assessing whether a supplier seems honest — you are assembling a record that makes dishonesty expensive to sustain. Evidence, not promises, is what crosses an ocean.

Key parameter identified — critical variable confirmed against spec.

Tolerance within achievable range — commercial ±0.15mm or tighter as needed.

Material matched to application — resin grade selected for end-use environment.

Process window validated at T1 — parameters locked after first trial.

Inspection method defined — CMM, visual, or functional test specified.

Frequently Asked Questions

How do I tell a real factory from a trading company?

Ask for a live video tour at a specific time and spot-check details only an operator would have. A genuine mold shop can walk the camera to a named machine, show you the clamping tonnage and the EDM and CMM actually in use, and answer questions about their own workflow without hesitation.

A broker can send photos and polished documents but balks at unscheduled live walks, gives generic answers about machine specifics, or “can’t get to the floor right now.” That does not mean brokers are always bad — some manage real shops well — but you should know who holds your tool and who you actually hold accountable.

Is a low quote a red flag on its own?

Not by itself, but an unusually low quote against a detailed spec is a strong signal to read the scope line by line. China’s tooling genuinely costs less than a Western local toolroom, so a competitive number can be honest. The concern is the quote that is low because it quietly assumed a simpler steel class, fewer cavities, no side-actions your geometry needs, or measurement that stops at “looks fine.”

Compare quotes on identical assumptions — same SPI class, same steel, same acceptance standard — and the gap between an inexpensive-but-complete quote and an attractive-but-under-scoped one becomes obvious. Price only means something when the deliverable behind each number is spelled out.

What is the one check if I only have time for one?

Commission a small paid test project and watch how they handle its first problem. Every other artifact — tours, lists, reports — is them presenting their best case; a live project is where habits surface. You will see whether the DFM engages your real geometry or rubber-stamps it, whether the dimensional report shows genuine readings or a formality, and how fast and honestly they respond when a feature comes in out of tolerance.

One modest project exposes the operating culture you will live with on the real program, which is precisely the thing a sales pitch is engineered to hide. Spend a little to learn a lot before you commit a lot.

Key Takeaways

Five verifiable artifacts — live video tour, machine list, comparable dimension report, signed DFM, trial video with mold number — replace promises with evidence you can independently check.
Tie payments to dimensional acceptance, not the calendar; a meaningful final tranche released only on a passing report aligns the supplier’s incentive with your outcome.
Run a small paid test project before the real tool — one full cycle through DFM, T1, report and change order exposes the operating culture a sales pitch is engineered to hide.

Vetting an overseas supplier is not a trust exercise — it is an evidence exercise.
Collect the artifacts, tie money to measurement, define acceptance numerically, and run a paid test before committing the real program. A supplier who can back their claims with verifiable evidence is worth a deposit; one who resists any of these checks has told you what they are before you wire anything.