ERP for Chinese-Owned Factories in Vietnam: A Practical Setup Guide

The short answer, and what actually decides it

A Chinese-owned factory in Vietnam has to keep Vietnamese statutory books, issue Vietnamese invoices, and pay Vietnamese taxes, while simultaneously delivering a Chinese parent numbers it can consolidate into group accounts. No single off-the-shelf answer covers both sides cleanly. The real decision is therefore not which ERP is best in the abstract, but where you place the boundary between the local system and the group system.

Three variables decide that boundary. The first is the size of the Vietnamese entity: a hundred-person assembly plant and a two-thousand-person integrated factory need completely different answers. The second is who owns the finance function on the ground — a local chief accountant working in Vietnamese, a Chinese expatriate finance manager, or a shared-service team at the parent. The third is how much of the parent's reporting has to emerge from the same database rather than being assembled afterwards. Change any one of those and the right architecture changes with it.

Most groups get this wrong in the same direction. They treat Vietnam as a configuration exercise on the group template, discover that Vietnamese statutory output is not one configuration step away, and finish the project with a working group system and a parallel manual process for local compliance. That is expensive, because the manual side is precisely where filings and audits actually happen.

• Small entity, local accountant, parent needs mapped aggregates: a local system plus a disciplined reporting bridge is usually cheapest.
• Material entity, group consolidation in a single platform: the China-localization pattern repeats — group system for consolidation, local system for compliance, one owned bridge.
• Volatile scope, phased investment: a local cloud system that can carry manufacturing first, with group reporting added once volumes justify it.
• No answer works if the boundary has no named owner. That is the failure mode behind most of the rest.

Why Vietnam is not simply a lighter version of China

Teams that already run a China entity often assume Vietnam will be familiar territory at a smaller scale. It is not, and the differences are structural rather than cosmetic. Vietnamese is not readable by most expatriate managers, which means the system interface language is not a preference but an operating constraint: if the local bookkeeping team cannot read the screen, the data entry will be wrong regardless of how good the software is.

The accounting framework is also different. Vietnamese Accounting Standards are their own body of rules, not a translation of Chinese standards, and the statutory statements a factory has to file follow a prescribed Vietnamese format. Invoicing has moved through a mandatory electronic regime with its own validation and transmission rules. Add customs and investment-incentive reporting on top, and the compliance surface a factory has to cover is genuinely distinct from the equivalent Chinese workload.

Finally, the local implementation market is thinner. In China you can find a certified partner for almost any platform in any province. In Vietnam the pool of consultants who can both configure an ERP and read Vietnamese tax rules is smaller, which shifts the balance toward platforms whose local ecosystem is established. Group procurement teams that ignore this end up flying implementers in, and paying for it every time something changes.

The three-books problem, and why it is the core of the design

Every Chinese-owned factory in Vietnam ends up carrying three different views of the same business. There is the statutory book in Vietnamese, prepared under Vietnamese standards, which is what the tax authority and the auditor see. There is the group book, in the parent's accounting framework and reporting currency, which is what headquarters consolidates. And there is the management view, which is what the factory director uses to run production, costing, and inventory.

• Statutory book — Vietnamese language, Vietnamese standards, Vietnamese filing deadlines and formats.
• Group book — the parent's framework and currency, mapped chart of accounts, group reporting package.
• Management book — cost by work centre, yield, scrap, machine and labour efficiency, the numbers the plant is actually managed on.

The design question is whether these come from one transaction set or from three. The bad pattern — and it is common — is that the statutory book lives in a local accounting package, the group book is re-keyed at the parent, and the management numbers live in spreadsheets maintained by the plant manager. The three then drift apart, and the month-end reconciliation turns into a negotiation about which number is right rather than a check on whether the process worked.

The good pattern is a single transaction base with three reporting layers on top: one posting, then a statutory view, a group view, and a management view derived from it. That requires the ERP to store the mapping between the local chart of accounts and the group chart, to hold exchange rates at the transaction level, and to produce each view from stored data rather than from an export-and-adjust ritual. It is not exotic, but it has to be designed before go-live, not retrofitted. The underlying principle is the same one we describe for any China entity in our guide to HQ consolidation from a subsidiary.

What the ERP must handle on the Vietnamese side

Electronic invoicing

Vietnam's invoice regime is electronic and validated, which means an invoice is not simply a document your system prints — it has to be issued through the prescribed channel, carry the required content, and be reportable. For the ERP this is an integration requirement, not a formatting requirement. The practical questions to ask are whether the platform has a maintained e-invoice connector for Vietnam, who updates it when the format changes, and whether the connection covers credit notes and adjustments as well as ordinary sales invoices. A factory that can raise an invoice but cannot cleanly reverse one has not solved the problem.

Statutory reporting in Vietnamese

The statutory statements and periodic tax filings have to be produced in Vietnamese and in the prescribed format. This is where local platforms have an inherent advantage and where global platforms depend on a localisation layer. When you evaluate, ask for a live demonstration of the actual statutory output for the Vietnamese entity rather than a slide describing it. The same discipline applies here as we described for Chinese statutory reporting requirements: a printable report from live data is the only evidence that matters.

Multi-currency, VND, and exchange differences

A Chinese-owned factory typically earns in USD or another hard currency from export customers, pays in VND for labour and local materials, and buys equipment and some raw materials in CNY or USD. That is a three-currency operating reality, and the ERP has to hold rates at transaction level, revalue monetary balances at period end, and produce exchange differences that flow correctly into both the statutory and group views. Groups that handle this in spreadsheets discover the cost during their first audit. The wider mechanics are covered in our guide to multi-currency ERP for trading companies.

Payroll, personal income tax, and social insurance

Vietnamese payroll has its own personal income tax structure and its own social, health, and unemployment insurance contributions, with rates and caps that change. Whether the factory runs payroll in the ERP or in a local service depends on headcount, but either way the cost has to land in the right cost centres and reach the general ledger without a re-keying step. The maintenance question matters more than the feature question: ask who keeps the local rules current, and what happens in the month after a change.

The export-processing question

Many Chinese-owned factories in Vietnam operate inside an export-processing or industrial-zone regime, which changes the customs and tax treatment of goods moving in and out. Materials may enter duty-suspended, finished goods may have to be exported rather than sold domestically, and the boundary between the factory and the domestic market becomes a compliance line rather than a commercial choice.

This matters for ERP design because it affects the inventory model. If the system cannot distinguish regime-specific stock, track the linkage between imported materials and exported finished goods, and produce the reconciliation the customs authority expects, that reconciliation will be done by hand. For a factory under this regime, that is one of the largest single sources of compliance effort in the business, and it is exactly the kind of flow that a China-side manufacturing-and-trading company already understands — see our note on the manufacturing and trading hybrid model.

Language, staffing, and the six-hour gap

Three operational realities shape the implementation more than any feature list. The first is that the local team works in Vietnamese and the parent works in Chinese, so the system has to serve at least two interface languages on the same data. The second is turnover: local accounting staff move, and a system that only two people understand is a business risk. The third is the working-day gap, which means that anything requiring a synchronous conversation between the plant and headquarters loses half a day.

• Interface language: bilingual or multilingual UI on one database, not two systems kept in sync by hand.
• Documentation: standard operating procedures written in both languages, owned locally rather than held at the parent.
• Approvals: delegated authority designed for an asynchronous working day, not a model that assumes the parent can approve within the hour.
• Training: budget for a second and third wave as staff turn over; the first go-live training is never enough.

How the projects that work are sequenced

The sequencing matters more than the platform choice, because most of the expensive mistakes are ordering mistakes.

• Decide the architecture boundary first — what lives in the local system and what lives at the parent — and write it down before any vendor demo.
• Confirm the statutory and invoicing requirements with the local tax office and a Vietnamese-qualified adviser, in writing, before configuration begins.
• Stand up the local statutory and inventory core first; add group reporting once the local close is stable.
• Run one full parallel statutory close before switching off the old process, and reconcile it line by line.
• Name the owner of the boundary between local and group systems, and put that name in the project plan.

The parallel close is the step teams most often skip and most often regret. It is the only thing that proves the new pipeline produces the same numbers the tax authority already has on file, and it is far cheaper than discovering the difference during an inspection.

Where Vietnam setups usually break

• The group template is treated as a starting point for Vietnam, and the statutory gap is discovered late.
• Invoice integration is assumed rather than demonstrated, and a manual workaround becomes permanent.
• Exchange differences are handled outside the system, producing statutory figures that do not tie.
• The regime-specific inventory reconciliation lives in a spreadsheet owned by one person.
• The local chart of accounts is designed by the parent without input from the local accountant.
• No one owns the local-versus-group boundary, so every period-end becomes a negotiation.

None of these is a software defect. Every one of them is a design decision that was deferred, and each is cheap to fix before go-live and expensive to fix afterwards.

What to ask a vendor

• Show a live Vietnamese statutory statement produced from live data, not a template in a slide deck.
• Describe the e-invoice integration, who maintains it, and what happens when the format changes.
• Name the Vietnamese-speaking implementation and support resource, and say where they are based.
• Explain how one posting produces the statutory, group, and management views without re-keying.
• Show the exchange-difference treatment end to end, including period-end revaluation.
• Give a reference for a Chinese-owned factory in Vietnam of comparable size and sector.

Any vendor with real Vietnam maturity answers these in specifics. Any vendor without it answers them in adjectives, and that difference is more informative than any feature matrix.

Verdict

Vietnam is one of the most common first stops for Chinese manufacturers moving production abroad, and it is a genuinely different compliance environment rather than a smaller version of China. The groups that handle it well are not the ones that pick the most impressive platform; they are the ones that decide the local-versus-group boundary deliberately, prove the statutory output on real data before cutover, and name a single owner for the seam between the two systems. For a group building several overseas entities at once, the same design logic extends across countries — see our framework for multi-entity consolidation for Chinese groups overseas. Groups comparing a Chinese platform with a local Vietnamese one should also read Chinese ERP versus local ERP in Vietnam and Thailand. If your team would rather discuss the specifics in Chinese, the Chinese-language overview is a faster route.