Can a network mix both structures?

Most mature manufacturers do. A common pattern concentrates the capital-intensive or technically demanding stages at one or two sites and distributes the final stages — assembly, filling, packing, configuration — close to markets. That keeps scale where scale matters while shortening the last leg and satisfying local requirements. The design decision is which stage forms the boundary, and it usually lands where the product becomes bulky, market-specific, or subject to local approval.

Does a second plant genuinely provide resilience?

Only if it can actually make the affected products, which requires deliberate investment rather than geography alone. That means qualified equipment, approved processes, trained people and, in regulated sectors, formal approval at both sites. Manufacturers who assume a sister plant will step in often discover during an incident that tooling is site-specific, the process was never qualified there, or the customer approval names one location. Test the assumption before you need it, ideally by running some volume at the alternative site periodically.

How do local content requirements affect the decision?

They can override the cost comparison entirely. Where public procurement, tariff preferences or sector rules require production or substantial transformation within a market, importing from a single efficient plant may simply exclude you from that market regardless of price. Because such requirements differ by country and by sector and change over time, they need checking against the specific markets in the plan rather than assumed, and the conclusion belongs in the network design rather than in a later commercial review.

One large plant or several regional ones: shaping the network

A manufacturer with customers in several regions has to decide whether production concentrates in one place or sits near the people buying it. Concentration collects scale, deep technical capability and a single process to control. Spreading it shortens delivery, satisfies market rules that concentration cannot, and removes the single point at which everything stops. The right structure follows from what your product costs to ship and how much capability each site would actually need.

Comparison criteria

Criteria are stated explicitly and neither option is declared a winner: which one fits depends on the constraint that binds hardest in your operation.

Criterion	Centralised: one plant supplying every market	Distributed: several plants located close to demand
Scale in conversion cost	Volume concentrates onto one set of equipment, so fixed costs divide across everything the business makes and large machines can be justified.	Volume splits, so each site runs smaller equipment at lower utilisation and the same total output costs more to convert.
Cost of reaching the customer	Every unit travels the full distance, and for bulky or low-value goods that freight can exceed what conversion saved.	Short final journeys, with the network cost moving upstream into moving raw material or components to several places instead.
Responsiveness to a regional order	Set by transit, and an urgent request either waits or moves by a premium mode that erodes the product's margin.	Local production can respond within days, which matters where customers hold little stock or expect scheduled call-off.
Technical and management depth required	One team of specialists, one maintenance organisation, one quality function, all deep enough to solve difficult problems properly.	Each site needs its own layer, and thin capability at a small plant is a recurring cause of quality drift and slow problem resolution.
Concentration of disruption	A fire, flood, strike or utility failure at one location interrupts supply to every customer at once, with no internal alternative.	An outage is contained to one region, and sister plants can sometimes cover, provided processes were kept genuinely transferable.
Market access and local requirements	Products cross borders, so tariffs, market-specific approvals and local content expectations all apply to imports.	Production inside a market can satisfy local content rules and approvals that no import arrangement will satisfy.
Consistency of the process	One process, one set of equipment, one specification. Any change is implemented once and applies to everything shipped.	Multiple processes that drift apart unless deliberately controlled, so a change must be replicated and requalified at every site.
Inventory in the system	Finished stock has to be positioned forward in each market to cover transit, or customers accept longer lead times.	Less in transit and less forward stock, but each site holds its own raw material and spares, duplicating the same items.

Choose Centralised: one plant supplying every market when

The process depends on equipment too costly to duplicate at sensible utilisation
The product carries high value relative to its bulk, so shipping it is a minor part of its cost
Process control rests on scarce specialists who cannot be recruited several times over
Approval is tied to a named site and every additional location would need approving separately

Choose Distributed: several plants located close to demand when

Freight or duty on the finished item is large compared with the cost of converting it
Customers expect short lead times or delivery commitments that transit cannot meet
Local content rules or market-specific approvals require production inside the market
A stoppage at a single site would leave every customer without supply simultaneously

The product's shipping economics settle most of this argument

Two numbers decide the shape of a network more than any strategy discussion: what it costs to convert a unit and what it costs to move it to a customer. Where conversion dominates, concentration has the advantage, because scale attacks the larger figure and the smaller one hardly registers. Where movement dominates, as with bulky, heavy, fragile or temperature-controlled products, the arithmetic flips and local production stops looking like a luxury. Packaging, beverages, insulation, precast concrete and similar goods rarely travel far for exactly this reason, while precision instruments and pharmaceutical actives are supplied globally from very few sites. Run that comparison for your own product before debating the organisational merits of either structure.

Every additional site brings a layer that is easy to under-budget

Plans for regional plants typically cost the building and the equipment, then assume management can be stretched. What each site actually needs is its own maintenance capability, its own quality function with the authority to stop production, its own supply management, and enough engineering to solve a process problem without waiting for someone to fly in. Where those are underprovided, the symptoms are consistent: slow problem resolution, specifications drifting apart, and a central team permanently travelling. Any comparison of network structures should carry the full site overhead, not just the incremental production cost, and should be honest about whether the necessary people can be recruited in that location.

Transferring a process is harder than the drawings suggest

Distributing production assumes the process can be reproduced elsewhere, and the assumption is usually optimistic. Specifications, parameters and drawings transfer easily; what does not transfer is the accumulated adjustment knowledge, the local supplier whose material behaves a particular way, and the operator judgement built over years. The evidence for this shows up as a new site taking far longer to reach yield than its business case allowed. Plants that transfer well do so deliberately: process characterisation written down before the transfer, people exchanged in both directions for extended periods, a formal requalification, and acceptance that early output from the receiving site will need heavier verification.

Frequently asked questions

Can a network mix both structures?: Most mature manufacturers do. A common pattern concentrates the capital-intensive or technically demanding stages at one or two sites and distributes the final stages — assembly, filling, packing, configuration — close to markets. That keeps scale where scale matters while shortening the last leg and satisfying local requirements. The design decision is which stage forms the boundary, and it usually lands where the product becomes bulky, market-specific, or subject to local approval.
Does a second plant genuinely provide resilience?: Only if it can actually make the affected products, which requires deliberate investment rather than geography alone. That means qualified equipment, approved processes, trained people and, in regulated sectors, formal approval at both sites. Manufacturers who assume a sister plant will step in often discover during an incident that tooling is site-specific, the process was never qualified there, or the customer approval names one location. Test the assumption before you need it, ideally by running some volume at the alternative site periodically.
How do local content requirements affect the decision?: They can override the cost comparison entirely. Where public procurement, tariff preferences or sector rules require production or substantial transformation within a market, importing from a single efficient plant may simply exclude you from that market regardless of price. Because such requirements differ by country and by sector and change over time, they need checking against the specific markets in the plan rather than assumed, and the conclusion belongs in the network design rather than in a later commercial review.

Data limitations

Manufacturing figures are operator-supplied inputs, not market data. GeoBusinessIQ holds no factory costs, production volumes, yields, cycle times, tooling prices or capacity data and does not estimate them — every result reflects only the figures you enter.
No manufacturer, supplier, vendor or factory is recommended, rated or ranked anywhere in this cluster, and no directory of them is published. Selection material describes how to run your own assessment; the assessment itself remains yours.

Explore the graph

Across GeoBusinessIQ

Sources

United Nations Industrial Development Organization — UNIDO (accessed 2026-08-30)
Covers: Industrial development analysis, industrial statistics methodology, and manufacturing capability programmes across member states.
Does not cover: Company-level data, factory costs, supplier information, or real-time production statistics.
Why it matters: The United Nations agency for industrial development; used for structural framing of how manufacturing sectors develop, never for point figures.
Review cadence: annual
OECD — OECD — economic and tax statistics (accessed 2026-05-14; reviewed 2026-05-18)
Covers: Comparable corporate tax, statutory rate, and economic indicators across member and partner economies.
Does not cover: Effective tax rates, deductions and incentives, local surtaxes, and personal residency rules.
Why it matters: Used as a cross-country baseline to sanity-check rates against primary tax-authority figures.
Review cadence: Annual, plus on major statutory changes.

Educational and operational information only — not legal, engineering, safety, customs, tax, or financial advice. Requirements vary by jurisdiction, product, process, and contract; confirm with the relevant authority or a qualified professional before acting.

Last updated: 2026-08-30