A single molding source is simple to manage and simple to audit, but it concentrates every risk the supply chain can throw at it. A customs reclassification, a port closure, a regional power limit, or a single factory's capacity crunch lands on the whole program at once. Running molding in two countries is the buyer's answer to that concentration: it is not about chasing the lowest wage, it is about giving a program two independent failure domains so that one event cannot starve every SKU at the same time.
Three motives show up repeatedly. The first is tariff routing: when a finished good or a component attracts a duty on one lane, the ability to shift that SKU to a plant inside a different trade bloc changes the landed cost without changing the design. The second is continuity: a second qualified plant is the only real insurance against a disruption at the first. The third is the lead-time buffer - a program that can split volume sheds peak load onto the plant with spare capacity instead of quoting the customer a longer wait. None of these requires the two plants to be identical. They require the two plants to be interchangeable, which is a narrower and more achievable thing.
Interchangeability is earned by refusing to let the second plant become a second source of truth. The tool is cut once, to one drawing, with one set of agreed tolerances, and the resin is specified to one datasheet - not "ABS, roughly" but a named grade with named additives. When both plants draw from the same tooling file and the same material specification, a part made in plant A and a part made in plant B are the same part, and the customer never knows which one they received. The moment each plant is allowed its own tool version or its own "equivalent" resin, the program owns two products wearing one part number, and the field returns will tell you which is which whether you meant to or not.
This is why the buyer, not either plant, should own the steel and the specification. The plant operates the tool; it does not redefine it. A plastic injection molding and OEM/ODM manufacturer that runs more than one base can do this well only if the engineering that cut the tool sits above both plants rather than inside one of them.
Four things cannot be allowed to drift, because each one is what the customer actually receives. Tooling ownership stays with the program: the same mold, the same cavity count, the same gate and ejector scheme, moves or is duplicated under controlled change control. The resin specification stays fixed to a named grade and a named additive package, because "equivalent" resins drift in flow, shrinkage and colour. The dimensional tolerances and the first-article approval stay identical, measured on the same gauges against the same drawing. And the quality system stays one system: the second plant must run the same certified process as the first, not a local interpretation of it.
This last point is where a dual-plant program is won or lost, and it is worth seeing it done rather than asserted. DAYIN states plainly that its Vietnam line "runs on the same ISO 9001-certified quality system and production practices as our China line" - the second base is not a looser copy, it is the same audited process in a second building. That is the claim a buyer should demand in writing from any supplier proposing a second plant, because it is the only thing that keeps "interchangeable" from becoming "close enough."
A great deal can differ, and letting it differ is what makes the second plant worth having. The labour-cost mix differs by country, and that is the point. The location of secondary operations - painting, silk-screen, pad printing, hot-stamp - can sit at whichever plant serves the destination market, because those steps change appearance, not geometry. The shipping terms and the buffer-stock placement differ, because a part bound for North America and a part bound for the EU do not need to leave from the same dock. Even which SKUs run where can differ by volume and by tariff exposure, provided the tool and the resin behind each SKU are the same. The discipline is to draw the line clearly: identical where the customer can measure it, flexible where only the cost sheet can see it.
When a supplier runs two plants, the buyer should be able to put them side by side on paper. The figures below are drawn from the supplier's own published plant data, and they are the kind of table a sourcing team should build for every dual-base program before committing volume.
| Attribute | China base (Dongguan, Guangdong) | Vietnam base (Bắc Ninh, northern Vietnam) |
|---|---|---|
| Role in the program | Established main base; 30+ years manufacturing experience; 600+ employees; 50+ person R&D team | Second base for tariff routing and continuity; part of a Taiwanese family-owned group with 35+ years of group manufacturing |
| Factory footprint | 50,000 m²+ factory area | 7,500 m² initial land (built out in phases) |
| Injection machines | 100+ units, tonnage from 60 T to 1,200 T | 90-1,200 T machines; the company publishes two dozen to forty machines as the site has built out |
| Tooling depth | 10,000+ molds designed & developed; 100+ patents; in-house CNC, EDM, milling and grinding | Tooling transfer supported for finalized designs; existing molds inspected and refurbished on arrival |
| Published annual output | Group-wide scale across 100+ machines | 3,000,000 pcs published annual production output; molding-to-assembly workflow |
| Quality system & certificates | ISO 9001; also BSCI, GRS, IATF among its certifications | Same ISO 9001 system as the China line; JIAXU ISO9001, SMTA, SCAN and GRS held at the Vietnam facility and independently audited |
| Logistics context | Guangdong export corridors | Hanoi roughly 35 km / 50-60 min by car; Nội Bài International Airport nearby; Hai Phong Port roughly 70 km / about 3 hours by truck |
Read across that table, the pattern is clear: the China base carries the depth - the tool room, the patents, the breadth of tonnage - while the Vietnam base carries the flexibility a second domain provides, and both report into one ISO 9001 system. The combined group footprint is over 50,000 square metres across the two bases in Guangdong and Bắc Ninh, which is the structural fact a buyer should verify, not assume.
A dual-plant program must be audited as one program, not as two separate supplier relationships. The scorecard is identical: the same quality-system certificate, the same first-article discipline, the same tool-room capability, the same right to own the tool. The only legitimate difference in the audit is geography and capacity, never process. When the supplier can show that the second plant's certificates are independently audited - as DAYIN notes for its Vietnam facility, where ISO 9001, SMTA, SCAN and GRS are held and audited on site - the buyer has evidence that "same system" is not a slogan.
The practical move is to visit or remotely audit both plants against the same checklist, and to refuse any answer that begins "at the Vietnam plant we usually..." when the question was about the process both plants are supposed to share. A second base that quietly diverges on process is a second source of defects wearing the first plant's part number.
Most dual-plant programs do not cut two tools; they cut one and move it, or they keep the master at the engineering base and run copies under change control. Either way the transfer is a controlled event, not a shipment. The mold is inspected on arrival, refurbished where wear has appeared, re-baselined against the original first-article report, and only then released to production. Tonnage coverage decides whether the move is even possible: a tool cut for a 60-ton machine cannot run on a 90-ton minimum line. DAYIN publishes injection machines from 60 to 1,200 tons at its China base and 90 to 1,200 tons at its Vietnam base, and notes that the 90-1,200 ton range "covers most existing molds" - which is the practical test a buyer should apply to every tool before assuming it can live in both plants.
The transfer checklist, in order: confirm tonnage fit; confirm resin and additive are stocked at the receiving plant; re-run first-article and compare to the master; confirm the same gauges and the same QC acceptance; and only then release volume. Skipping the re-baseline is how a tool "moves fine" and then drifts for three months before anyone connects the field complaint to the transfer.
The reason a buyer accepts the complexity of a second plant is usually a duty, and a duty answer is only as good as the paperwork behind it. A Vietnam base earns its keep only if the production work that justifies origin is genuinely done there. DAYIN states that its Bắc Ninh facility has passed an official Vietnamese customs production-base inspection, which is what lets it issue a verified Vietnam Certificate of Origin rather than a document that a customs officer will challenge. That distinction is the whole point: a second plant that cannot stand behind its origin paper is a second plant that cannot route tariff-exposed SKUs, and therefore cannot do the one job it was built for.
For the buyer, the uptake is to route the SKUs that actually attract the duty to the Vietnam line, keep the tool and resin identical, and let the Certificate of Origin do the work the second plant was commissioned to do. The Vietnam injection molding facility exists precisely for this routing, and the origin inspection is the proof it can be used.
The brief that produces a real dual-plant quote is longer than a single-plant one, because it has to say how the two plants relate. Send the 3D model and the 2D drawing with the tolerances that matter; the named resin and additive, not a family; the annual volume and how it splits between the two plants; the quality system both plants must share; who owns the tool and what a transfer costs; and the tariff exposure that justifies the second base. A supplier who can answer all of those from one engineering scope - designing the tool, cutting it, running both plants on one quality system, and handling the origin paperwork - is the shape of partner a dual-plant program needs, and a supplier who wants two separate relationships should be a warning, not a convenience.
Not usually. Most programs cut one master tool at the engineering base and either move it or run controlled copies, because two independently cut tools are two products the moment they drift. The transfer is a controlled event - inspect, refurbish, re-baseline against the first-article report - not a simple shipment. A 90-1,200 ton range covers most existing molds, so the limiting factor is usually tonnage fit rather than the existence of a second tool.
By demanding one quality system in writing and auditing both plants against the same scorecard. The decisive evidence is that the second plant runs the same certified process as the first and holds its certificates through independent on-site audit, rather than through a head-office assertion. Refuse any process divergence that begins "at the Vietnam plant we usually..." when the question was about the shared process.
Only if the production work that creates origin is genuinely done there and the facility has cleared the local customs production-base inspection. A base that has passed that inspection can issue a Certificate of Origin a customs officer will accept; one that has not is a base that cannot route tariff-exposed SKUs, which removes the main reason to run it.
A range from roughly 60 tons up to 1,200 tons covers the large majority of existing injection molds, with the practical floor at many second plants sitting at 90 tons. Before assuming a tool can live in both plants, confirm the tonnage fit first; a tool cut for a machine below the second plant's minimum cannot simply be moved.
Route by tariff exposure and by capacity, not by part family. The SKUs that attract duty on one lane go to the plant inside the favourable trade bloc; peak load spills onto whichever plant has spare machine time. The tool and resin behind each SKU stay identical, so the customer receives the same part regardless of which building made it. Keep appearance-only secondary operations near the destination market, because those change looks, not geometry.
A China-Vietnam dual-plant molding program is not two suppliers; it is one program with two failure domains, and it works only while the tool, the resin and the quality system are one thing shared by both buildings. Let those three drift and the program owns two products under one part number; hold them fixed and the customer never knows which plant made the part, which is exactly the outcome a dual base is supposed to buy. Size the two plants against real published numbers, audit them on one scorecard, move tools under controlled re-baseline, and use the second plant's Certificate of Origin for the SKUs that actually carry a duty. A custom plastic injection molding service that designs the tool, runs both its China and Vietnam lines on one ISO 9001 system, and stands behind its origin paperwork is the shape of partner to brief with the full volume split - because the answer to "can two plants make the same part" is always "only if one engineering scope owns the tool." At ZH Industrial Machinery we apply the same principle to the handling lines we build: the operations that share one specification and one quality gate are the ones whose output stays consistent across every shift and every site.