Strategic Logistics Decisions and Network Economics

Introduction

Strategic logistics decisions determine the structure, capability and long-term economics of a supply chain. Unlike routine dispatching or short-term transport scheduling, strategic decisions commit substantial capital, alter operating risk and shape customer service for several years. Typical decisions include whether to outsource logistics, whether to insource a distribution activity, how much capacity to install, whether to automate a warehouse, and where to locate facilities.

These decisions cannot be evaluated by accounting cost alone. A sound analysis combines relevant costing, contribution and break-even analysis, discounted cash flow, real-options reasoning, scenario analysis and network optimisation. The central managerial question is not simply “Which option has the lowest unit cost?” but “Which option creates the greatest risk-adjusted economic value while satisfying service, resilience, regulatory and sustainability constraints?”

Learning Objectives

After studying this chapter, the reader should be able to:

1. distinguish relevant, sunk, committed, opportunity and avoidable costs;
2. evaluate outsourcing and insourcing using contribution and break-even methods;
3. apply NPV, IRR and equivalent annual cost to logistics investments;
4. explain real options in capacity, automation and network design;
5. formulate facility, transport and inventory trade-offs as network-economic decisions;
6. use scenario and sensitivity analysis to address uncertainty;
7. identify governance, ethical, resilience and implementation risks.

1. The Economic Logic of Strategic Logistics

A logistics network converts resources into service outcomes. Its economic performance depends on the interaction of five elements:

• Demand: volume, geography, variability and growth;
• Facility structure: plants, fulfilment centres, cross-docks and depots;
• Transport structure: modes, lanes, consolidation and routing;
• Inventory positioning: cycle stock, safety stock and postponement;
• Service promise: delivery speed, reliability, availability and flexibility.

A low-cost network may be strategically inferior if it creates excessive lead time, concentration risk or poor recovery after disruption. Conversely, a redundant network may be uneconomic if customers do not value the additional service. Network economics therefore treats service and resilience as economic outputs, not merely operational preferences.

Important cost concepts

The relevant-cost rule is: include future cash flows that differ between alternatives; exclude sunk costs; include opportunity costs and genuinely avoidable overheads. Allocated corporate overhead is not relevant unless the decision changes it.

For a decision with alternatives i and a planning horizon of T periods, the economic value of alternative i can be represented as:

Here, Iᵢ is initial investment, Bᵢₜ is incremental benefit or revenue in period t, Cᵢₜ is incremental cash operating cost, Sᵢ is terminal or salvage value, and r is the discount rate. The formula assumes cash flows are measured consistently, occur at known times and are incremental to the decision. Taxes, working capital, inflation and risk should be included when material.

2. Outsourcing and Insourcing Decisions

Outsourcing transfers some activities to a specialist provider; insourcing retains or restores internal control. The decision should compare the avoidable internal cost with the supplier’s total economic cost, not merely the quoted price.

For volume Q, outsourcing is financially preferable on a single-period basis when:

where p is supplier variable price per unit, Fₒ is outsourcing fixed or implementation cost, Rₒ is expected outsourcing risk cost, v is avoidable internal variable cost per unit, Fᵢ is avoidable internal fixed cost, Rᵢ is internal risk cost, and O is the opportunity cost of using internal capacity. All terms must refer to the same period and currency.

If risk and fixed-cost differences are temporarily ignored, the break-even volume is:

This expression applies only when v ≠ p and the denominator and numerator produce an economically meaningful positive volume. If the supplier price is higher than the avoidable internal variable cost, outsourcing may still be justified by capacity release, superior service, lower capital requirements or risk reduction.

Contribution analysis

Contribution per unit is selling price minus variable cost:

where P is revenue per unit and VC is variable cost per unit. Total contribution is Q × CM. A constrained logistics manager should allocate scarce capacity to the activity with the greatest contribution per unit of the binding constraint, not necessarily the greatest contribution per product.

The break-even volume with fixed cost F is:

The equation assumes a constant selling price, constant unit variable cost, one relevant activity and no capacity discontinuity. Multi-product operations require a stable sales mix or a weighted-average contribution margin.

Worked example: outsource or retain

A firm currently processes 100,000 orders annually. The avoidable internal variable cost is $4.80 per order and avoidable annual fixed cost is $180,000. A provider quotes $5.20 per order plus $60,000 annual contract administration cost. Outsourcing releases internal space that can generate $90,000 annual contribution elsewhere. Expected transition and governance risk is estimated at $35,000 annually.

Internal relevant cost is:

$4.80 × 100,000 + $180,000 = $660,000.

Outsourcing net economic cost is:

$5.20 × 100,000 + $60,000 + $35,000 − $90,000 = $525,000.

Outsourcing therefore improves annual economic contribution by $135,000, before tax and financing effects. If the released capacity cannot actually be used, the $90,000 opportunity benefit must be removed; the apparent advantage then falls to $45,000. This illustrates why capacity validation is essential.

A qualitative review must also assess data ownership, labour effects, supplier financial health, service-level credits, cybersecurity, continuity arrangements, switching costs and contract termination rights. Outsourcing is not risk transfer by default; it often changes the type and visibility of risk.

3. Capacity, Automation and Investment Appraisal

Capacity decisions involve indivisibilities. A facility or automated system may be cheap per unit at high utilisation but uneconomic when demand is low or volatile. Managers should distinguish installed capacity, practical capacity and utilised capacity.

Net present value

Net present value is:

where I₀ is initial investment, CFₜ is net after-tax cash flow in period t, Sₜ is terminal value, r is the risk-adjusted discount rate and T is project life. Accept an independent project when NPV is positive; among mutually exclusive projects, select the highest positive NPV subject to strategic constraints. NPV assumes the discount rate appropriately reflects systematic risk and that projected cash flows are internally consistent.

Internal rate of return

IRR is the discount rate that makes NPV equal to zero:

IRR can be useful for communication, but it can mislead when cash flows change sign more than once, projects differ in scale or reinvestment assumptions are unrealistic. NPV is generally the stronger decision criterion.

Worked example: warehouse automation

An automated storage and retrieval system requires $2,000,000 initially. It generates net cash savings of $650,000 in each of four years and has a $200,000 terminal value. At a 10% discount rate:

PV of savings = $650,000 × [(1 − 1/1.10⁴)/0.10] = approximately $2,058,500.

PV of terminal value = $200,000/1.10⁴ = approximately $136,600.

NPV = −$2,000,000 + $2,058,500 + $136,600 = approximately $195,100.

The project is financially attractive under the stated assumptions. However, the result should be tested for volume, labour, maintenance, energy, software integration, downtime and residual-value uncertainty. A positive NPV does not remove implementation risk.

When alternatives have unequal lives, equivalent annual cost or equivalent annual value may improve comparison. For a present value PV, life n and discount rate r:

This converts a capitalised cost into an equivalent constant annual cost, assuming repeated replacement or a comparable decision horizon.

4. Real Options and Strategic Flexibility

Traditional NPV often treats investment as now-or-never. Logistics investments commonly contain managerial choices: defer construction, expand after demand is proven, abandon an underperforming site, switch suppliers or install modular automation. These choices have option value.

A real option is the value of flexibility in a physical or operating decision. An investment may have a lower static NPV but greater option value if it allows staged commitment and protection against downside demand.

Examples include:

• leasing temporary warehouse space before building;
• installing modular conveyors rather than a fully integrated system;
• using a third-party logistics provider while demand is uncertain;
• reserving land for a future facility;
• designing a network with transferable inventory and transport capacity.

A simple decision-tree expected value is:

where pₛ is the probability of state s, Vₛ is the state-specific value and C₀ is current commitment. For multi-period decisions, state values should be discounted. Expected value alone is insufficient when management can observe information and exercise options conditionally; the value of flexibility is the difference between the flexible strategy and the best fixed strategy.

Real-options reasoning is especially important when uncertainty is high, investment is partly irreversible, information will improve over time and decisions can be staged. It should not be used to justify every delay: postponement may lose service, market share or scale economies.

5. Network Design Economics

Network design determines flows between suppliers, facilities, customers and disposal or return points. A simplified capacitated network model minimises total cost:

subject to:

Here, Fⱼ is the fixed cost of opening facility j, yⱼ is a binary open/closed variable, cᵢⱼ is unit transport and handling cost from facility j to customer i, xᵢⱼ is shipped quantity, dᵢ is customer demand, Kⱼ is facility capacity and Iᵢⱼ represents inventory-related requirements where modelled. The formulation assumes known demand, linear costs and divisible flows. Real networks may require service-time, route, labour, emissions, risk and multi-period constraints.

The fundamental trade-off is that more facilities usually reduce outbound distance and response time but increase facility fixed cost, duplicated safety stock and managerial complexity. Fewer facilities may create scale economies but increase transport exposure and concentration risk.

Inventory and transport trade-offs

For independent replenishment, the classical economic order quantity is:

where D is annual demand in units, S is ordering or setup cost per order and H is annual holding cost per unit. The model assumes constant demand, known lead time, no quantity discounts and instantaneous replenishment. It is not a complete network-design model, but it clarifies the ordering-cost versus holding-cost trade-off.

Safety stock is often estimated as:

where z is the desired service-factor quantile and σᴸ is standard deviation of demand during lead time. If demand and lead-time uncertainty are independent, with daily demand standard deviation σ_d, mean daily demand d̄, mean lead time L̄ and lead-time standard deviation σ_Lt, then:

The formula assumes independence and approximately stable distributions. Correlation requires an additional covariance term. Safety stock should be valued against stockout cost, working capital, obsolescence and service commitments rather than treated as free protection.

6. Scenario, Sensitivity and Risk Analysis

A strategic model should expose what drives its conclusion. Sensitivity analysis changes one input at a time; scenario analysis changes a coherent set of assumptions; simulation represents repeated combinations of uncertain inputs.

Useful scenarios include base demand, high growth, recession, fuel-price shock, labour shortage, supplier failure, port disruption and regulatory change. For each scenario, evaluate NPV, service level, capacity utilisation, inventory, emissions and recovery time.

Expected monetary value is:

where probabilities must be explicit and collectively exhaustive. A risk-adjusted analysis should also report downside measures, such as the worst-case loss or percentile outcome, because two alternatives with equal EMV may have very different exposure.

A useful managerial dashboard includes:

7. Illustrative Case: Northstar Distribution Network

This is a fictional illustrative case. Northstar, a consumer-electronics distributor, serves 1,200 retailers from one national distribution centre. Annual demand is 500,000 cartons. Management is considering either retaining the central facility, opening a regional facility, or outsourcing fulfilment to a specialist provider.

Problem

The central facility has low average transport cost through consolidation, but average delivery time is four days in remote regions. Demand is growing unevenly, service penalties are increasing, and the existing site will reach 90% practical capacity within three years.

Analysis

The finance team first separated relevant costs from allocations. The regional option required $4 million capital, increased facility fixed cost by $1.1 million annually, reduced outbound transport by $1.5 million and reduced safety stock through shorter lead times by $0.3 million. It also increased duplicated systems and management cost by $0.2 million. Expected annual service contribution from improved availability was estimated at $0.6 million.

The annual operating benefit before capital recovery was:

$1.5m + $0.3m + $0.6m − $1.1m − $0.2m = $1.1m.

At a 10% discount rate, a four-year annuity of $1.1 million has a present value of approximately $3.49 million. With no terminal value, the regional option has an approximate NPV of:

−$4.00m + $3.49m = −$0.51m.

The central option therefore remains financially preferable in the base case. However, a high-growth scenario increases transport savings and service contribution to $1.7 million annually. The four-year present value becomes approximately $5.39 million, producing NPV of $1.39 million. A staged option—leasing a regional cross-dock for two years with an expansion right—has lower immediate commitment and preserves information value.

Approach and solution

Northstar selected the staged option. It negotiated a short-term cross-dock lease, introduced postponement for selected product families and reserved land near a major demand cluster. A quarterly trigger was established: invest in a permanent regional facility if demand exceeded 650,000 cartons, service penalties exceeded $0.5 million annually, or central utilisation exceeded 95% for two consecutive quarters.

Outcome and learning points

The decision avoided premature capital expenditure while creating a credible expansion path. The case demonstrates that the lowest base-case NPV is not always the best strategic choice when demand uncertainty and reversibility matter. It also shows why service benefits must be monetised carefully and why decision triggers should be defined before managers become committed to a project.

8. Governance, Ethics and Contemporary Practice

Strategic logistics decisions affect employees, communities, suppliers and the environment. Governance should address:

• transparent assumptions and audit trails in financial models;
• fair treatment, consultation and reskilling during automation or outsourcing;
• supplier labour standards, safety and modern-slavery controls;
• privacy and cybersecurity in connected logistics systems;
• emissions, congestion, land use and environmental justice;
• conflict-of-interest controls in facility selection and procurement.

Artificial intelligence can improve demand forecasts, routing and scenario generation, but it does not replace managerial accountability. Models may reproduce historical bias, conceal uncertain assumptions or optimise cost at the expense of safety and service equity. Human review, explainability, data-quality controls, access governance and stress testing are essential.

Contemporary networks increasingly use digital twins, event-driven visibility, autonomous handling, robotics-as-a-service and distributed fulfilment. These technologies increase flexibility but also create integration dependency, cyber risk, skills requirements and new forms of vendor lock-in. Technology should therefore be evaluated as a capability and option, not merely as equipment.

9. Best-Practice Decision Process

A disciplined strategic logistics appraisal should:

1. define the decision, time horizon, service promise and constraints;
2. map the current network and identify true capacity bottlenecks;
3. separate relevant, sunk, allocated and opportunity costs;
4. quantify demand, lead-time, service and disruption assumptions;
5. develop feasible alternatives, including staged and reversible designs;
6. evaluate contribution, break-even, NPV and, where useful, IRR;
7. test scenarios, sensitivities, correlations and downside exposure;
8. include inventory, transport, facility, labour, tax and working-capital effects;
9. assess resilience, sustainability, ethics, governance and implementation;
10. establish decision rights, triggers, metrics and post-investment review.

The final recommendation should state not only the preferred option, but also the assumptions under which it is preferred, the conditions that would reverse the decision and the controls that will detect those conditions.

Challenges and Limitations

Quantitative logistics models can create false precision. Demand forecasts may be biased, probabilities may be subjective, service benefits may be difficult to monetise and linear costs may fail at peak congestion. NPV can understate flexibility; real-options analysis can overstate it when exercise is operationally difficult. Network optimisation may ignore labour relations, political constraints or behavioural responses. Finally, a locally optimal facility design may damage total supply-chain performance through bullwhip effects, supplier instability or customer dissatisfaction. Models should therefore support structured judgement, not replace it.

Chapter Summary

Strategic logistics decisions require an integrated economic view of cost, service, capacity, risk and flexibility. Relevant-cost analysis excludes sunk costs and includes avoidable and opportunity costs. Outsourcing and insourcing can be tested using contribution and break-even logic, but supplier risk and released-capacity benefits must be validated. NPV is the primary investment criterion, while IRR and equivalent annual cost provide complementary perspectives. Real-options analysis captures the value of staging, waiting, expansion and abandonment under uncertainty. Network design balances facility fixed cost, transport, inventory, service and resilience subject to capacity and flow constraints. Scenario and sensitivity analysis reveal the assumptions driving results. Effective decisions combine rigorous financial modelling with governance, ethical scrutiny, sustainability assessment and clear implementation triggers.

Key Terms

Relevant cost: A future cost that differs between alternatives.

Sunk cost: A past, unrecoverable cost that should not affect a new decision.

Opportunity cost: Contribution sacrificed by choosing one use of a resource over another.

Avoidable cost: A cost eliminated when an activity is discontinued.

Contribution margin: Revenue less variable cost.

Break-even volume: Activity level at which contribution covers relevant fixed cost.

Capacity constraint: A limit on throughput, labour, space, equipment or another resource.

Net present value: Discounted value of incremental cash inflows less investment.

Internal rate of return: Discount rate that makes project NPV equal to zero.

Equivalent annual cost: Constant annual cost equivalent to a present value.

Real option: Value arising from managerial flexibility in an uncertain investment.

Scenario analysis: Evaluation of outcomes under coherent alternative assumptions.

Sensitivity analysis: Examination of the effect of changing selected inputs.

Network design: Strategic configuration of facilities, flows, transport and inventory.

Safety stock: Inventory held to protect against demand or lead-time uncertainty.

Postponement: Delaying final configuration or commitment until demand information improves.

Resilience: Ability to withstand and recover from disruption.

Stranded cost: Continuing cost remaining after an activity is transferred or closed.

Cost-to-serve: Total resource cost of serving a specified customer, product or channel.

Review Questions

Short Answer Questions

1. What distinguishes a relevant cost from a sunk cost?
2. Why may allocated corporate overhead be excluded from an outsourcing analysis?
3. State the principal decision rule for NPV.
4. What is the economic meaning of a real option?
5. Name two trade-offs created by adding a distribution facility.

Long Answer Questions

1. Explain how contribution and break-even analysis can support an insourcing or outsourcing decision. Discuss its limitations.
2. Compare NPV and IRR as methods for evaluating an automated warehouse investment.
3. Explain how facility location affects transport, inventory, service and resilience costs.
4. Discuss the role of staged investment and real-options reasoning in uncertain logistics environments.
5. Design a governance framework for a strategic logistics network redesign.

Analytical / Application Questions

1. A firm processes 80,000 units internally at an avoidable variable cost of $6.00 and avoidable fixed cost of $120,000. A supplier charges $6.80 per unit plus $30,000 fixed cost. Calculate the annual relevant cost of each alternative and recommend one, ignoring qualitative factors.
2. An investment costs $1,000,000 and produces $300,000 annual cash savings for five years with no terminal value. At 10%, calculate approximate NPV and interpret the result.
3. A network manager faces uncertain demand and can either build a permanent facility now or lease temporary capacity for two years before deciding. Identify the information, downside protection and opportunity costs that should be included in the comparison.

Multiple Choice Questions with Answers

Question 1

Which cost should normally be excluded from a strategic outsourcing comparison?

A. An avoidable internal labour cost

B. A supplier implementation fee

C. A sunk cost from a warehouse built five years ago

D. The opportunity cost of released capacity

Correct Answer: C. A sunk cost from a warehouse built five years ago

Question 2

If selling price is $50 and variable cost is $32, the contribution margin per unit is:

A. $18

B. $32

C. $50

D. $82

Correct Answer: A. $18

Question 3

A project with a positive NPV at the appropriate discount rate generally:

A. Destroys shareholder value

B. Creates value under the stated assumptions

C. Must have an IRR below zero

D. Has no implementation risk

Correct Answer: B. Creates value under the stated assumptions

Question 4

IRR is the rate that:

A. Maximises accounting profit

B. Equals annual fixed cost

C. Makes project NPV equal to zero

D. Equals the inflation rate

Correct Answer: C. Makes project NPV equal to zero

Question 5

Which situation most strongly supports real-options analysis?

A. A fully reversible purchase with certain demand

B. An irreversible investment under uncertain demand with staged choices

C. A sunk cost already paid

D. A transaction with no managerial flexibility

Correct Answer: B. An irreversible investment under uncertain demand with staged choices

Question 6

In a capacitated facility-location model, a binary variable commonly indicates whether a facility is:

A. Open or closed

B. Profitable or unprofitable each day

C. Full or empty every hour

D. Domestic or international

Correct Answer: A. Open or closed

Question 7

Adding more distribution centres commonly reduces outbound transport distance but increases:

A. Facility duplication and inventory complexity

B. Customer demand certainty

C. Supplier reliability automatically

D. Sunk costs already incurred

Correct Answer: A. Facility duplication and inventory complexity

Question 8

The EOQ model balances ordering cost against:

A. Holding cost

B. Corporate tax

C. Facility depreciation only

D. Customer credit risk

Correct Answer: A. Holding cost

Question 9

Sensitivity analysis generally changes:

A. Every assumption simultaneously in a coherent story

B. One selected input while other inputs are held constant

C. Only historical data

D. The accounting standard used

Correct Answer: B. One selected input while other inputs are held constant

Question 10

Which is an example of an opportunity cost in logistics?

A. A past equipment purchase

B. Revenue contribution lost because warehouse space is used for another activity

C. A non-refundable permit paid last year

D. A cost that never changes between alternatives

Correct Answer: B. Revenue contribution lost because warehouse space is used for another activity