Supply Chain Network Design and Global Operations

Introduction and context

Supply chain network design determines how an organisation structures the physical and organisational system through which products, information and money flow. It includes decisions about the number, location, capacity and ownership of plants, suppliers, distribution centres, warehouses, cross-docks, ports and final-market channels. Unlike daily routing or inventory replenishment, network design is a strategic activity: decisions may require large capital commitments and remain influential for 10–20 years.

Global operations add complexity because supply chains cross borders, currencies, legal systems, cultures, transport regimes and geopolitical environments. A network that minimises manufacturing cost may perform poorly when tariffs, long lead times, disruption risk, carbon emissions or customer-service requirements are included. The central managerial task is therefore not to minimise one visible cost, but to design a resilient and economically sustainable system that creates value for customers and stakeholders.

Learning objectives

By the end of this chapter, the learner should be able to:

1. Explain the strategic role of supply chain network design.
2. Evaluate facility location and network configuration alternatives.
3. Compare distribution channels, outsourcing and offshoring arrangements.
4. Calculate total landed cost and interpret break-even decisions.
5. Identify major trade, customs, tax, regulatory and documentation requirements.
6. Assess the effects of risk, resilience, sustainability and complexity on global operations.
7. Recommend a network design using quantitative and qualitative evidence.

1. The strategic architecture of a supply chain network

A supply chain network consists of nodes and links. Nodes include raw-material suppliers, contract manufacturers, plants, consolidation centres, distribution centres, retailers and customers. Links include transport lanes, information connections, financial transactions and contractual relationships. A network may be linear in a simple diagram, but real networks are usually multi-tier and many-to-many.

Network design decisions occur at three levels:

The design should align with the organisation’s competitive priorities. A responsive fashion retailer may value proximity and speed; a commodity producer may prioritise scale economies; a medical-device firm may require traceability and regulatory control. The same location is not optimal for every product or strategy.

Network configuration alternatives

Common configurations include:

• Centralised distribution: one or a few large facilities serve a broad market. This can reduce duplicated inventory and exploit scale, but increases delivery distance and concentration risk.
• Decentralised distribution: several regional facilities are positioned near customers. This improves responsiveness and may reduce outbound transport cost, but requires more inventory, management and fixed assets.
• Hub-and-spoke: a central hub consolidates flows to regional spokes. It supports transport efficiency, although hub failure can affect many markets.
• Direct shipment: suppliers or plants ship directly to customers. This can remove handling and intermediate inventory but may create small, expensive shipments.
• Cross-docking: inbound goods are rapidly sorted and transferred to outbound vehicles with little storage. It is effective when demand is predictable, shipment timing is coordinated and product handling is standardised.
• Postponement: generic products are produced or positioned centrally, while final configuration, labelling or packaging occurs near demand. Postponement reduces forecast risk and supports market customisation.

A useful distinction is between push activities based on forecasts and pull activities triggered by actual demand. The point at which control changes from push to pull is the customer-order decoupling point. Its location influences inventory exposure, lead time and customisation.

2. Facility location evaluation

Facility location is a multi-criteria decision. Important factors include customer proximity, supplier access, labour availability and skill, utilities, land and construction cost, transport infrastructure, taxes, incentives, political stability, regulatory conditions, disaster exposure, cybersecurity and environmental constraints.

Qualitative screening

A practical process begins by defining mandatory requirements, such as port access, power reliability, product certification or maximum delivery time. Candidate locations are then screened using a weighted scoring model. If location i receives a score s_ij for criterion j, and criterion weight w_j satisfies ∑w_j = 1, the weighted score is:

Here, S_i is the total score for location i, m is the number of criteria, and scores must use a consistent scale. The method is transparent but not objectively optimal: weights and scores reflect managerial judgement and may conceal uncertainty.

Centre-of-gravity analysis

For an initial distribution-centre estimate, the centre-of-gravity method uses customer coordinates weighted by demand:

x_i,y_i are coordinates of market i, and d_i is demand or shipment weight. The result is a geometric starting point, not a final site recommendation. It assumes transport cost is approximately proportional to distance and ignores roads, congestion, land availability, tariffs and service constraints.

Fixed and variable cost analysis

For a candidate facility, annual cost may be represented as:

TC_i is annual total cost; F_i fixed facility cost; v_i variable processing cost per unit; Q_i annual volume; T_i transport cost; I_i inventory-related cost; and R_i quantified risk or expected disruption cost. All terms must be measured over the same period and in the same currency. If transport, inventory or risk is not estimated reliably, the model should present scenarios rather than false precision.

For two alternatives with fixed costs F_A,F_B and variable costs v_A,v_B, the break-even volume, where other costs are equal, is:

This result applies only when v_A ≠ v_B and costs are linear over the relevant range. Capacity limits, step costs and quantity discounts require a more advanced model.

3. Distribution channels and inventory positioning

Distribution channels describe how ownership and physical possession move from producer to user. Options include direct sales, distributors, wholesalers, retailers, marketplaces, third-party logistics providers and omnichannel combinations.

Direct-to-customer channels provide data and control but require investment in fulfilment, customer service and returns. Intermediaries provide market access, local knowledge and aggregation, but reduce margin and may limit visibility. Omnichannel design must address channel conflict, common inventory visibility, delivery promises, returns and consistent pricing.

The choice of channel should consider:

• order size and demand density;
• required delivery speed;
• product value, perishability and risk of damage;
• customer willingness to accept delivery charges;
• reverse-logistics requirements;
• channel partner capability; and
• information-sharing quality.

Inventory positioning involves deciding where stock should be held and in what form. Risk pooling can reduce safety stock when demand from several markets is aggregated. If independent demand standard deviations are σ_i, a simplified pooled standard deviation is:

This formula assumes independent demand. With correlation, the general variance is:

where ρ_ij is the correlation between markets i and j. Positive correlation reduces the benefit of pooling; negative correlation increases it.

4. Outsourcing, offshoring and governance

Outsourcing transfers an activity to an external organisation. Offshoring moves an activity to another country, whether internally owned or externally supplied. They are not synonyms: an organisation can offshore production to its own subsidiary or outsource domestically.

The make-or-buy decision should evaluate more than quoted price. Relevant criteria include total cost, capacity flexibility, intellectual property, quality, technology access, supplier financial health, labour practices, business continuity, data security and switching difficulty.

A transaction-cost perspective asks whether market contracting or internal ownership better manages coordination, information asymmetry and opportunism. The resource-based view asks whether an activity depends on a capability that is valuable, rare, difficult to imitate or strategically important. Activities that create distinctive customer value or protect critical know-how are usually candidates for stronger internal control.

A sound outsourcing governance model includes:

1. clear specifications and measurable service-level agreements;
2. quality, delivery and compliance metrics;
3. audit and access rights;
4. change-control and escalation procedures;
5. business-continuity and disaster-recovery obligations;
6. data, intellectual-property and confidentiality protection;
7. ethical sourcing and worker-safety requirements; and
8. exit, transition and knowledge-transfer provisions.

Single sourcing may produce volume discounts and closer collaboration, but dual or multiple sourcing can reduce dependency. Supplier segmentation should distinguish strategic, bottleneck, leverage and routine categories, while recognising that a low-spend item can still be operationally critical.

5. Total landed cost and global trade requirements

Purchase price is only one component of international sourcing economics. Total landed cost (TLC) is the cost of making a product available at the intended destination:

Where:

The formula is a management framework rather than a universal legal definition. Some organisations include inventory carrying cost, financing, carbon cost or warranty cost; therefore the scope must be stated. Each term must be expressed per unit or for the same shipment and currency.

A sourcing comparison should also account for currency. If a foreign price is P_f in foreign currency and the exchange rate is e domestic currency units per foreign currency unit, domestic price is:

Exchange-rate scenarios should be used because the rate at quotation, shipment and payment dates may differ. Hedging can reduce volatility but introduces financial cost and contractual obligations.

International shipments depend on accurate classification, valuation and origin information. Managers must verify the applicable tariff classification, customs valuation method, preferential-origin rules, import licences, export controls, sanctions, product standards, labelling, packaging, documentation and record-retention requirements. Incoterms® allocate specific delivery obligations, costs and risks between buyer and seller, but they do not by themselves determine ownership, payment terms, tariff classification or regulatory compliance. Professional practice requires using the current official rules and obtaining specialist advice for complex transactions.

6. Global supply chain complexity and risk

Complexity arises from the number of nodes, products, suppliers, countries, hand-offs, systems, regulations and interdependencies. More complexity may increase market coverage and flexibility, but it also increases coordination effort, data requirements and failure points.

Risk should be assessed through a structured register covering probability, impact, detectability, time to recovery and dependencies. Common risks include:

• geopolitical conflict, sanctions and sudden trade restrictions;
• port congestion, transport interruption and infrastructure failure;
• natural hazards and climate change;
• supplier insolvency or quality failure;
• cyberattack and counterfeit data;
• labour disputes and human-rights violations;
• demand shocks and forecast error; and
• currency, interest-rate and commodity-price volatility.

Resilience is not simply holding more inventory. It may involve dual sourcing, regional capacity, strategic buffers, modular products, alternate transport routes, supplier development, visibility platforms and tested recovery plans. A useful performance view balances cost, service, resilience and sustainability rather than optimising cost alone.

Carbon assessment should include transport mode, distance, load factor, production energy, packaging and inventory effects. Nearshoring may reduce transport emissions but increase production energy or material use; therefore life-cycle analysis is preferable to a single-distance comparison.

7. Quantitative network design models

A basic capacitated facility-location model illustrates the structure of a network decision. Let y_j equal 1 if facility j is opened and 0 otherwise; x_ij be units shipped from facility j to customer i; F_j be fixed opening cost; c_ij be unit supply and transport cost; K_j be facility capacity; and D_i be customer demand.

Subject to:

The first constraint satisfies demand; the second prevents shipments from a closed facility and respects capacity. This model assumes known demand, cost, capacity and transportation relationships. Real models may add multiple products, inventory, service-time constraints, tariffs, emissions, disruption scenarios and facility-opening limits. Optimisation results are decision support, not automatic decisions: poor data or omitted constraints produce elegant but unsuitable answers.

8. Worked example: landed cost and location choice

An organisation compares two supply options for 10,000 units. Option A is domestic: purchase cost £42 per unit, inland freight £3, inspection £1 and risk allowance £0.50. Option B is overseas: purchase cost £30, origin charges £1.50, ocean freight £4, insurance £0.40, duty £2.80, destination and inland charges £3.20, compliance £0.60 and risk allowance £2.50.

Domestic landed cost is:

£42 + £3 + £1 + £0.50 = £46.50 per unit.

Overseas landed cost is:

£30 + £1.50 + £4 + £0.40 + £2.80 + £3.20 + £0.60 + £2.50 = £44.00 per unit.

The apparent purchase-price saving is £12, but the complete saving is only £2.50 per unit, or £25,000 for 10,000 units. If an additional delay-related cost of £3 per overseas unit is probable, overseas effective cost becomes £47 and the domestic option is economically preferable. The result demonstrates why sourcing decisions should use scenarios and expected risk cost rather than supplier price alone.

9. Illustrative case: Northstar Appliances

Background. Northstar Appliances, a fictional mid-sized manufacturer, sells small kitchen equipment in Europe, East Africa and the Gulf region. Its single plant in East Asia offers low conversion cost, but average replenishment lead time is 11 weeks. Ocean disruption and customs delays have caused stockouts, while finished goods are customised with market-specific labels.

Problem. Management must choose between retaining one global distribution centre, adding regional facilities, or using a postponement strategy. It must also decide whether to continue sole sourcing electronic controls from one supplier.

Analysis. The team mapped customer demand, lead times, customs requirements, return flows and disruption dependencies. Total landed cost showed that the East Asian plant remained cheapest for base production, but urgent airfreight and lost-sales exposure erased much of the advantage. Demand correlation between Europe and the Gulf was moderate, suggesting partial risk pooling. The electronic-control supplier was low cost but represented a bottleneck component with a 16-week recovery time.

Approach and solution. Northstar retained global production, established a regional postponement centre near the main European market, and used generic packaging until final labelling. It contracted a qualified secondary control supplier and required both suppliers to maintain tested continuity plans. A customs broker was integrated into the product-launch process, and shipment visibility was connected to the enterprise planning system. Management evaluated the design using landed cost, service level, working capital, emissions and disruption scenarios.

Outcome. The redesigned network reduced average customer lead time and lowered emergency freight. It increased regional fixed cost, but the scenario analysis indicated improved total value and lower concentration risk. The case does not claim a universal solution: a market with low volume or unreliable regional infrastructure might favour centralisation.

Learning points. Network design requires cross-functional analysis; postponement can separate scale production from local customisation; a low-price supplier may still be strategically risky; and resilience benefits should be evaluated alongside cost and service.

10. Technology, governance and contemporary practice

Digital twins, control towers, IoT tracking, electronic trade documents and advanced planning systems can improve visibility and scenario analysis. Artificial intelligence may support demand sensing, disruption prediction, document classification and route recommendations. However, AI outputs depend on representative data, stable definitions and appropriate human oversight. Models can reproduce bias, misclassify products or recommend actions that violate trade controls. Governance should include data ownership, access control, audit trails, explainability, cybersecurity and escalation to qualified professionals.

Contemporary network design increasingly considers regionalisation, nearshoring, circular flows, repair and remanufacturing, renewable energy, forced-labour due diligence and climate adaptation. The best practice is not automatic relocation. It is structured scenario planning that tests cost, service, compliance, carbon, recovery time and strategic flexibility under plausible futures.

Challenges and limitations

Network models can be distorted by inaccurate demand forecasts, omitted indirect costs, unstable exchange rates and overconfident probability estimates. Weighted scoring can hide subjective assumptions. Historical data may not represent unprecedented disruption. Local incentives may expire, while infrastructure and labour markets change. A design optimised for average conditions may fail under stress.

Managers should therefore use sensitivity analysis, alternative scenarios, independent validation, periodic review and explicit decision rights. They should document assumptions and distinguish facts, estimates and judgements. Ethical performance must be treated as a design requirement, not merely a public-relations consideration.

Best-practice decision process

1. Define customer promises, products, markets and planning horizon.
2. Map the multi-tier network and identify critical dependencies.
3. Establish mandatory legal, service, capacity and sustainability constraints.
4. Estimate total landed cost and working-capital implications.
5. Develop location, channel, sourcing and postponement alternatives.
6. Model capacity, demand, transport, tariffs and disruption scenarios.
7. Evaluate qualitative factors, including ethics and strategic capability.
8. Stress-test the preferred design and conduct sensitivity analysis.
9. Implement governance, performance measures and contingency plans.
10. Review the network as demand, technology, regulation and risk evolve.

Chapter Summary

Supply chain network design determines the location, capacity, ownership and connections of facilities and channels. Strategic choices must align with competitive priorities and balance cost, service, resilience, sustainability and control. Centralisation creates scale and pooling benefits, while decentralisation improves responsiveness but adds fixed cost and inventory. Facility evaluation can combine weighted scoring, centre-of-gravity analysis and cost modelling, provided assumptions are explicit.

Outsourcing and offshoring are different decisions and require governance of quality, continuity, intellectual property, labour standards and exit risk. Total landed cost extends beyond purchase price to include freight, insurance, duties, handling, compliance, risk and other defined costs. Global operations require accurate classification, valuation, origin, documentation and regulatory controls. Quantitative models can structure facility-location and allocation decisions, but results depend on data and omitted constraints. Resilient networks use visibility, postponement, multiple sources, regional capability and tested recovery plans. Technology and AI can improve analysis only when supported by sound data, cybersecurity and accountable human governance.

Key Terms

Centre of gravity: A demand-weighted geographic estimate for an initial facility location.

Channel conflict: Tension among distribution channels competing for customers, margin or inventory.

Cross-docking: Transfer of inbound goods to outbound vehicles with minimal storage.

Customer-order decoupling point: The stage where forecast-driven activity changes to customer-order-driven activity.

Distribution centre: A facility that receives, stores, consolidates and distributes products.

Incoterms®: Standardised trade rules allocating specified delivery responsibilities, costs and risks.

Landed cost: The cost of making a product available at a defined destination.

Nearshoring: Locating an activity in a nearby foreign country to reduce distance or complexity.

Network design: Strategic configuration of supply chain nodes, links, capacities and flows.

Offshoring: Relocating an activity to another country.

Outsourcing: Obtaining an activity or service from an external organisation.

Postponement: Delaying final product differentiation until demand or market information is available.

Risk pooling: Combining demand or inventory positions to reduce aggregate uncertainty.

Service level: A measure of the probability or extent to which customer requirements are met.

Supply chain resilience: The ability to prepare for, respond to and recover from disruption.

Total landed cost: A defined comprehensive cost including purchase, logistics, trade, compliance and risk elements.

Trade compliance: Conformance with customs, export-control, sanctions and related legal requirements.

Weighted scoring model: A multi-criteria method that combines criterion scores using assigned weights.

Review Questions

Short Answer Questions

1. Distinguish strategic, tactical and operational supply chain decisions.
2. What are two advantages and two disadvantages of a centralised distribution network?
3. Explain the difference between outsourcing and offshoring.
4. Why is purchase price insufficient for an international sourcing decision?
5. What is the purpose of postponement?

Long Answer Questions

1. Explain how facility location should be evaluated using both quantitative and qualitative criteria.
2. Discuss the relationship between distribution-channel design, inventory positioning and customer service.
3. Analyse the governance requirements of an outsourced global manufacturing relationship.
4. Explain the principal trade-compliance requirements that affect international supply chains.
5. Evaluate how resilience and sustainability should be incorporated into network design.

Analytical / Application Questions

1. A domestic supplier quotes £58 per unit. An overseas supplier quotes £39, with £2 origin charges, £5 freight, £1 insurance, £4 duty, £3 destination charges, £1 compliance and £2 expected risk cost. Calculate and compare total landed cost, then identify one non-cost factor that could change the decision.
2. A firm is considering one large distribution centre versus two regional centres. Develop a decision framework that includes fixed cost, transport, inventory pooling, service time, disruption risk and emissions.
3. A sole-source supplier provides a low-cost component but has a long recovery time after disruption. Recommend a sourcing and governance strategy, explaining its effects on cost, resilience and control.

Multiple Choice Questions with Answers

Question 1

Which decision is primarily strategic in supply chain network design?

A. Selecting tomorrow’s delivery sequence

B. Choosing the number and location of distribution centres

C. Releasing a daily purchase order

D. Assigning a dock door for one shipment

Correct Answer: B. Choosing the number and location of distribution centres

Question 2

What is the main limitation of a centre-of-gravity calculation?

A. It cannot use demand weights

B. It automatically includes customs duties

C. It ignores practical constraints such as roads and site availability

D. It requires binary optimisation variables

Correct Answer: C. It ignores practical constraints such as roads and site availability

Question 3

Which statement correctly distinguishes outsourcing from offshoring?

A. Outsourcing always involves another country

B. Offshoring always involves an external supplier

C. Outsourcing uses an external organisation, while offshoring moves activity abroad

D. They are two names for exactly the same decision

Correct Answer: C. Outsourcing uses an external organisation, while offshoring moves activity abroad

Question 4

A product has a purchase price of £40, freight of £4, duty of £3 and compliance cost of £1. What is its partial landed cost excluding other items?

A. £44

B. £47

C. £48

D. £49

Correct Answer: C. £48

Question 5

What is the principal purpose of postponement?

A. To increase early product customisation

B. To delay final differentiation until demand information is available

C. To eliminate all inventory

D. To replace customs documentation

Correct Answer: B. To delay final differentiation until demand information is available

Question 6

In the capacitated facility-location model, what does a binary variable for a facility normally indicate?

A. The facility’s transport distance

B. Whether the facility is opened

C. The customer’s demand forecast

D. The product’s tariff rate

Correct Answer: B. Whether the facility is opened

Question 7

Which condition increases the benefit of inventory risk pooling?

A. Highly positive demand correlation

B. Independent or negatively correlated demand

C. Identical demand peaks in every market

D. A single customer with fixed demand

Correct Answer: B. Independent or negatively correlated demand

Question 8

Which item is generally not determined solely by Incoterms® rules?

A. Certain delivery responsibilities

B. Allocation of specified transport risks

C. Tariff classification and customs valuation

D. Some cost responsibilities between buyer and seller

Correct Answer: C. Tariff classification and customs valuation

Question 9

Why might a low-price overseas source have a higher effective cost than a domestic source?

A. International freight is always free

B. Total landed cost includes logistics, duties, compliance and disruption exposure

C. Domestic suppliers cannot provide quality assurance

D. Purchase price is never relevant

Correct Answer: B. Total landed cost includes logistics, duties, compliance and disruption exposure

Question 10

Which governance practice is most appropriate for a strategically critical outsourced supplier?

A. Relying only on an informal verbal promise

B. Removing audit rights to build trust

C. Using measurable service levels, continuity obligations and an exit plan

D. Avoiding performance data

Correct Answer: C. Using measurable service levels, continuity obligations and an exit plan