China Factory Setup Advisory for Foreign Manufacturers

CHAPTER 01

The investment case: what should China production solve?

A manufacturing investment begins with a commercial hypothesis: producing in China will improve access to customers, shorten delivery, reduce a specific cost, improve engineering responsiveness or secure a production capability. Each proposition needs a measurable comparison with the existing operation. “China has a large market” is context; it is not evidence that a particular factory will earn an acceptable return.

Separate domestic demand from export demand

For a domestic-market plant, build demand from identifiable customers, product approvals, qualification cycles and realistic share of wallet. For an export plant, start with destination-market requirements, freight, tariffs, origin rules, delivery promises and working capital. A combined plant needs separate demand and margin assumptions for each channel. Domestic and export sales can require different specifications, packaging, certifications and service arrangements.

Organize research into three evidence levels. Confirmed purchase commitments are stronger than customer forecasts; customer forecasts are stronger than a general industry growth estimate. Record who provided the evidence, when it was collected, what assumptions it relies on and when it expires. A customer letter of interest should not silently become a committed order in the investment model.

Translate the market into a plant-loading plan

Build a product-by-customer forecast with volume, price, launch date and qualification status. Then convert it into machine hours, labor hours and material demand. Demand for several products cannot simply be added if they compete for the same bottleneck machine or require lengthy changeovers. The first commercial question is whether the proposed asset base can be loaded profitably during ramp-up, not merely whether nameplate capacity will eventually be needed.

Decision gate: proceed to detailed feasibility when the commercial case can be expressed as testable volumes, margins, operating requirements and an investment ceiling. If customer demand remains uncertain, a staged entry may preserve more options than a fully integrated plant.

CHAPTER 02

Choose the operating model before selecting a property

Owning a Chinese manufacturing company and owning its factory building are different decisions. An investor may operate a wholly owned business from a leased industrial building, purchase an existing business, construct a new facility or outsource production. Treat ownership, property and production scope as separate variables.

Compare the routes over a common planning horizon. Include transfer costs, qualification, management attention, minimum orders, inventory, warranty exposure and exit costs. A contract-manufacturing quote contains a supplier margin, but it may also avoid underused equipment and a permanent management team. An owned plant can deliver greater control while becoming more expensive at low utilization.

Use staged commitment when uncertainty is high

A practical sequence might start with outsourced production, establish a local technical or quality team, introduce a leased assembly operation and localize selected processes only after demand is demonstrated. This is an analytical option, not a universal recommendation: regulated production, IP sensitivity or specialized process control may require an owned operation earlier.

Write down the trigger for moving to the next stage. Examples include signed customer programs, stable process yield, a verified order pipeline or reaching a defined capacity threshold. Also define the conditions under which the project will stop. Without those conditions, a “pilot” can become an open-ended investment.

CHAPTER 03

Foreign-investment policy: distinguish access, eligibility and permission

Three questions must be answered independently: can the foreign investor enter the activity; can this project be established in the proposed location; and can the finished operation legally manufacture and sell its products? A positive answer to one does not settle the others.

Foreign-investment access

The 2024 national foreign-investment negative list took effect on 1 November 2024. NDRC stated that the remaining manufacturing-specific foreign-investment access restrictions were removed. This concerns foreign-investment access; it does not remove general industrial, environmental, safety or product requirements. Read the actual activity against the applicable list and check associated business lines as well as the main manufacturing process. NDRC announcement; official order.

General market access and industrial policy

The separate Market Access Negative List applies to market-entry questions beyond foreign ownership. Its 2025 edition contains prohibited and permission-based activities. A company registration alone is not evidence that every proposed process or product is authorized. Review the production activity, capacity, technology and location against applicable industry and project rules. Official 2025 Market Access Negative List.

Encouraged investment is a separate classification

The Encouraged Catalogue for Foreign Investment (2025 edition) took effect on 1 February 2026. It includes national and regional provisions; a project needs to match the relevant description and conditions rather than rely on a broad industry label. Catalogue inclusion and an actual incentive entitlement are different findings. Identify the specific measure, competent authority, required confirmation and conditions before including a benefit in a financial model. MOFCOM catalogue.

Convert policy research into a usable register

Create one row per issue: proposed activity; rule and version; applicability; responsible authority; evidence required; project owner; unresolved question; and decision affected. For example, an unresolved wastewater acceptance condition belongs in both the compliance register and the site-selection decision. It should not remain a footnote in a legal memo while the property team signs a lease.

The implementing regulations of the Foreign Investment Law provide the broader framework, including investment promotion and protection. They are a starting reference, not a substitute for checking the specific project and local implementation. Official implementing regulations. Policy references here were checked on 22 September 2026; applicability should be reconfirmed at the investment decision.

CHAPTER 04

Define the process and size the factory from the bottleneck

A useful technical brief describes how the product is made, inspected, stored and moved. Begin with a process-flow diagram and material balance. Identify which activities are internal and which are subcontracted, including surface treatment, heat treatment, cleaning, testing and waste handling. These choices influence location eligibility, utility needs and supplier dependency.

Build a capacity model before specifying floor area

For each major operation, record cycle time, available hours, changeover, uptime and expected yield. The slowest constrained process determines the practical output of the line. If a machine runs for 4,000 scheduled hours, processes 25 units per hour, achieves 85% availability and 95% yield, its illustrative good output is 80,750 units—not the theoretical 100,000. These are example assumptions, not industry benchmarks.

Check whether quoted productivity already includes downtime or scrap so the model does not deduct the same loss twice. Model shifts separately because recruitment, supervision, maintenance and overtime requirements can differ. Add capacity for inspection, packaging and material handling; a fast production machine does not compensate for an overloaded test station.

Translate the process into a facility schedule

Issue a controlled revision of this schedule to every park and landlord. When the process changes, update the site assessment and budget together. Otherwise, an early property shortlist can continue to be used after its technical assumptions have become invalid.

CHAPTER 05

Regional research: build an industry-specific geographic shortlist

There is no best Chinese province for all manufacturers. Regional research should identify where the project can obtain critical inputs, sell or dispatch its output, recruit its team and operate its process. Province-level averages are too coarse for the final decision: two industrial parks in the same city can have different utility constraints, environmental conditions and available buildings.

Start with cluster evidence, then test the individual location

MIIT publishes national advanced-manufacturing cluster lists, which provide a starting point for researching concentrated industrial capabilities. They do not certify that any particular property is available or suitable. Use official cluster descriptions to generate candidate areas, then verify the suppliers, institutions and facilities the project would actually use. MIIT 2024 cluster notice.

The following matrix is a research agenda, not a ranking or promise of suitability. It shows how the questions should change with the project rather than treating every region as interchangeable.

Translate an industry label into a process-specific location question

For electronics assembly, investigate component lead time, engineering change responsiveness, traceability and electrostatic-control requirements. For precision machinery, test access to machining, heat treatment, metrology and service engineers. For food processing, raw-material seasonality, cold chain and hygiene zoning may dominate. Chemical or surface-treatment projects require early investigation of process eligibility, hazardous inputs and environmental infrastructure. A medical-product project adds product-specific quality and regulatory requirements. These are screening questions; they do not establish approval in any candidate location.

Create an evidence folder for each region with supplier interviews, customer distances, labor-market checks, logistics quotations and property responses. A generic description of a city's “excellent business environment” should have no independent weight in the investment score.

CHAPTER 06

Site Selection: eliminate fatal flaws before scoring advantages

Use two stages. First, test mandatory conditions such as process acceptance, minimum power, required building geometry and a feasible start date. Second, compare qualified sites on cost, logistics, talent, supply-chain access and expansion. A location that fails a mandatory condition cannot become acceptable by scoring well on rent or incentives.

Design the scoring model around the operating model

A supplier-intensive assembly project and a high-energy continuous process need different weights. The illustrative weighting below demonstrates the method; it is not a standard to copy without adjustment. Score each factor against explicit anchors—for example, “power capacity confirmed with a dated connection plan” rather than “good utilities.” Attach evidence quality to every score.

Run sensitivity tests by changing weights and uncertain inputs. If a site wins only because of an unconfirmed subsidy or a speculative wage assumption, the ranking is fragile. Present both the numerical score and unresolved conditions to decision-makers. A score is a summary of evidence, not an alternative to due diligence.

Compare total occupancy cost

Normalize usable production area, office space, common areas, rent escalation, deposits, fit-out, electrical work, reinstatement and business interruption. A building with cheaper nominal rent may require substantial reinforcement or additional off-site storage. Compare costs over the intended occupancy period and separately show recoverable deposits, recurring expense and irreversible expenditure.

Maintain a shortlist of credible alternatives until the preferred site has passed verification. Losing all alternatives before lease negotiations weakens negotiating options and makes an unresolved technical issue harder to address.

CHAPTER 07

Property due diligence and lease negotiation

Property review should reconcile documents with the actual building and intended process. Obtain ownership and authorized-leasing evidence, permitted use, relevant building records, existing approvals and utility information. Commission qualified specialists for legal title, structural, environmental and engineering questions where required. A landlord presentation does not establish these facts.

Inspect the production journey through the building

Walk the route of incoming materials, large equipment, operators, finished goods and waste. Check loading areas, lift capacity, turning space, separation requirements and emergency routes. Compare machine point loads with structural evidence. Confirm whether ceiling services, sprinkler systems or columns reduce the advertised clear height. “Industrial use” alone does not demonstrate suitability for every manufacturing activity.

For utilities, request written capacity, delivery date, connection responsibility and capital contribution. Distinguish a utility line near the property from capacity available to the tenant. For wastewater, assess both flow and characteristics. A park treatment plant's existence does not prove it will accept the proposed discharge.

Turn unresolved conditions into contract decisions

Have legal advisers translate agreed responsibilities into enforceable documents: landlord works, delivery condition, modification rights, fit-out period, utility milestones, expansion, termination and reinstatement. Where feasible, connect major commitments to defined verification conditions. Make clear who bears the cost of delays or required modifications, and whether the lease term supports recovery of non-removable fit-out costs.

Record each issue as accepted, resolved, conditional or unacceptable. Assign an owner and deadline to conditional items. An unqualified statement that “due diligence is complete” is less useful than a short schedule identifying the remaining conditions before possession, fit-out and production.

CHAPTER 08

Entity, governance, funding and operating control

Plan the legal entity alongside the operating model. Identify the shareholder, contracting parties, asset owners, employers, technology licensors and trading entities. A simple group chart often reveals unanswered questions about who imports equipment, owns inventory, invoices customers and receives payments.

Company registration is one workstream within the project. Build a dependency list covering company documents, applicable investment information reporting, bank onboarding, tax administration, payroll, customs arrangements where relevant and internal authorization. Requirements depend on the structure and activities; obtain project-specific professional advice rather than use a generic registration checklist as proof of readiness.

Design controls before the first payment

Agree approval limits for procurement, bank payments, contracts and hiring. Define custody and use of company seals, access to bank systems, segregation of duties and escalation rules. Headquarters should be able to reconcile contracts, invoices, inventory and cash without depending on one person's informal records. These controls are particularly important during construction and fit-out, when spending precedes revenue and transaction volume rises quickly.

Fund the cash trough, not only the equipment order

Align equity, borrowing and intercompany arrangements with the payment schedule and obtain tax, legal and foreign-exchange advice as applicable. Check when funds must actually be available for deposits, equipment milestones, payroll and inventory. A financing plan that arrives after contractual payments are due cannot support the construction schedule, even if total funding is adequate.

CHAPTER 09

Feasibility study: connect evidence to an investment decision

A useful feasibility report is an argument that can be examined. It should explain why the project is commercially justified, technically workable, implementable at the selected site and financeable under realistic conditions. It must also show what could invalidate that conclusion. A report made from disconnected market, engineering and financial chapters can hide inconsistent assumptions.

Use one controlled assumption set

Link the sales plan to capacity, capacity to equipment, equipment to utilities and building needs, and the implementation schedule to cash flow. If production begins three months later, sales, inventory, staffing, financing and customer commitments must move consistently. Record each assumption's source, confidence, owner and next review date.

Develop at least a base case and a coherent downside case. A downside should reflect connected events: delayed customer qualification can reduce utilization, increase unit costs and extend the cash trough simultaneously. Testing each variable independently can understate the combined exposure.

The recommendation should distinguish “proceed,” “proceed subject to conditions,” “redesign” and “stop.” List the conditions explicitly: confirmed utility date, accepted product samples, signed customer program, financed cash requirement or an acceptable lease. This helps the board release funding in stages rather than approve every future commitment at once.

CHAPTER 10

Capital cost, operating economics and cash flow

Build CAPEX by work package: property access; civil or fit-out work; production equipment; utilities; environmental and safety facilities; IT; installation; commissioning; and professional services. Keep scope boundaries visible. A machine quotation may exclude freight, duties, foundations, power distribution, rigging, calibration or acceptance testing.

Build OPEX from operational drivers: material consumption and scrap, direct labor hours, indirect staffing, utility demand, maintenance, rent, quality costs, logistics and administration. Distinguish fixed and variable costs. Use the same output and yield assumptions as the technical model; otherwise the apparent unit margin can be overstated.

A transparent break-even example

Assume, only for illustration, annual fixed cash operating costs of RMB 12 million and contribution of RMB 40 per good unit after variable costs. Operating break-even is 300,000 units per year: 12,000,000 ÷ 40. If practical capacity is 500,000 units, that corresponds to 60% utilization. This simplified calculation excludes financing, tax, depreciation and recovery of initial investment, so it is not an investment-return calculation.

Now test a lower contribution of RMB 32. Break-even rises to 375,000 units, or 75% of the same capacity. The example shows why a modest margin change can materially alter the required loading. The project needs evidence that customer demand and process capability can support that loading.

Model working capital and timing explicitly

Track supplier deposits, production inventory, finished goods, customer receivables and payables month by month. Startup often requires inventory before revenue and payroll before stable yield. Model the most negative cumulative cash position and add a reasoned liquidity buffer. Do not equate accounting profit with available cash.

Show incentive income separately and state its evidence status. Keep unconfirmed benefits outside the financing base case. Use a separate sensitivity for payment delays or failure to qualify. A project that is viable only if an uncertain grant arrives on schedule has a funding dependency that the board should see directly.

CHAPTER 11

Approvals and schedule: build a dependency map

Factory setup is a network of decisions rather than a universal numbered application sequence. The applicable procedures vary with industry, construction scope, process and location. Map project filing or approval questions, environmental assessment, energy-related review, building and fire matters, safety requirements and product licences as applicable. For each, verify the competent authority, legal basis, documents and dependencies.

Separate entity readiness, facility readiness and production readiness

An incorporated company may still lack a usable factory. A completed factory may still lack equipment acceptance or product qualification. A successful trial run may still be below target yield or unable to supply a customer's approved product. Give each milestone its own acceptance criteria instead of combining them into an optimistic “opening date.”

Estimate duration from actual work packages, authority feedback and supplier quotations. Identify activities that can overlap and those that cannot. In a leased-building project, utility upgrades or customer qualification may dominate the schedule; in a greenfield project, land and construction dependencies may do so. Publishing one generic completion time would obscure this difference.

Assign one accountable owner per dependency. Maintain a weekly log showing planned date, current forecast, blocker and decision required. A percentage-complete report is insufficient when a single unresolved condition controls the startup date.

CHAPTER 12

Procurement, commissioning and operational ramp-up

Procure equipment against a user requirement specification that connects product characteristics, capacity, quality, interfaces and acceptance tests. The comparison should explain scope differences, service support, spare parts, software access, training and warranty. Lowest purchase price can create a higher lifecycle cost if the machine fails acceptance or cannot be serviced promptly.

Define acceptance before ordering

Where appropriate, use factory acceptance testing before shipment and site acceptance testing after installation. Agree test materials, cycle time, quality limits, repeatability, documentation and responsibility for correcting defects. A demonstration with easy-to-process material does not necessarily prove performance on the client's production mix.

Control interfaces among machinery, power, ventilation, extraction, compressed air, cooling and IT. Assign responsibility for connection points and capacity. Record approved changes and their budget and schedule consequences. Many installation delays arise from gaps between otherwise complete supplier scopes.

Recruit and train against the ramp-up curve

Appoint plant leadership early enough to shape layout, hiring, maintenance and quality systems. Phase recruitment against installation and training rather than the original calendar alone. Plan supervisors, maintenance, quality and logistics as well as operators. Overseas management visits are useful but should not substitute for an accountable local operating team.

Track good output, yield, downtime, changeover, safety observations, delivery and cash consumption through ramp-up. Compare actual performance with the feasibility assumptions. If yield stays below plan, determine whether the cause is material, process, equipment, training or specification before adding another shift. Expanding a poorly controlled process can increase losses faster than output.

CHAPTER 13

The investment decision pack and first 30 days

The final decision pack should allow a reader who did not participate in site visits to understand the recommendation. Include the operating model, customer evidence, process and capacity model, location alternatives, policy screen, site findings, investment budget, cash requirement, schedule, risks and conditions. State which documents support each material conclusion.

A practical first-month investigation plan

1. Week 1: define the decision. Confirm product families, target markets, required control, timing and investment ceiling. Gather available drawings, forecasts and existing cost data.
2. Week 2: test assumptions. Interview commercial and technical owners, identify missing demand evidence, map process constraints and start the access review.
3. Week 3: investigate alternatives. Compare entry models and selected candidate regions. Issue a consistent preliminary requirement brief to relevant locations and providers.
4. Week 4: decide the next investigation. Summarize gaps, initial economics and potential fatal flaws. Agree whether to commission detailed feasibility and site verification, revise scope or pause.

This is an illustrative investigation cadence, not a promise that due diligence or approval can be completed in a month. Complex products, missing data and specialist studies can require additional time. The goal is to spend the next increment of money on the uncertainty that matters most.

What to assemble before an initial discussion

Prepare a concise product and process description; target customers and markets; expected annual volume; critical quality requirements; production equipment list if available; major utility and environmental characteristics; preferred regions and the reason for each; budget range; and target milestones. Mark unknowns explicitly. An incomplete but honest brief is more useful than precise-looking numbers without support.

Related working material: 160 investment questions, factory cost guide and workbook, Site Selection guide and industrial park profiles.