Logistics, Distribution and Network Design

Introduction and context

Logistics is the coordinated movement and storage of materials, products, information and related services from origin to consumption and, where necessary, back to recovery or disposal. Distribution is the customer-facing part of logistics concerned with order fulfillment, inventory positioning, transportation, delivery and service. Network design determines the location, capacity, ownership and roles of facilities and the flows connecting them.

For an entrepreneurial firm, logistics is not merely an operating expense. It shapes market reach, cash conversion, customer experience, resilience and environmental performance. A low-cost network may be unable to support rapid delivery; a highly responsive network may carry excess inventory, duplicated facilities and expensive transport. The managerial task is therefore to design an integrated system that balances cost, service, risk, flexibility and sustainability.

A useful systems view is:

Suppliers → plants or assemblers → regional distribution centres → customers

The reverse flow may include returns, repairs, reusable packaging, recycling and end-of-life recovery. Decisions at one stage affect every other stage. For example, decentralising inventory may reduce delivery time but increase safety stock and facility costs.

Learning objectives

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

1. Explain the strategic role and terminology of logistics and distribution.
2. Compare transportation modes, warehouse configurations and fulfillment models.
3. Formulate and interpret basic facility-location and network-cost models.
4. Design routing, packaging, cold-chain and last-mile solutions.
5. Evaluate omnichannel and reverse-logistics choices.
6. Balance service, cost, responsiveness, resilience and environmental impact.
7. Apply network-design principles to an entrepreneurial business case.

1. The logistics system and its design decisions

Logistics decisions operate at three levels:

The principal flows are physical, informational and financial. Physical flows move products; information flows communicate demand, inventory, orders, status and exceptions; financial flows include payments, freight charges, duties, credits and refund obligations. Weak information flow creates the bullwhip effect: small changes in downstream demand can produce progressively larger fluctuations upstream.

Important terminology includes:

• Inbound logistics: movement from suppliers to the focal organization.
• Outbound logistics: movement from the organization to customers or channels.
• Distribution centre (DC): a facility designed primarily to receive, sort, store temporarily and dispatch products.
• Fulfillment centre: an operation optimized for customer-order processing, often involving item-level picking and parcel shipping.
• Cross-docking: transferring goods from inbound to outbound vehicles with little or no storage.
• Transshipment: movement through an intermediate node without necessarily being stored for long periods.
• Milk run: a planned vehicle route serving several suppliers or customers.
• 3PL: a third-party logistics provider performing logistics activities for a client.
• 4PL or lead logistics provider: an integrator coordinating multiple logistics providers and capabilities.
• Service level: a defined probability or proportion of demand served under specified conditions.
• Order cycle time: elapsed time from order placement to customer receipt.

2. Network design as a strategic optimization problem

A network contains nodes—plants, suppliers, warehouses, stores, parcel hubs and customers—and arcs representing possible transportation links. A design must answer five questions:

1. How many facilities are needed?
2. Where should they be located?
3. What products and markets should each serve?
4. Which transport modes and providers should connect them?
5. What capacity and inventory should be positioned at each node?

The basic trade-off is shown in the following relationship:

Figure: More facilities can improve responsiveness while increasing facility and inventory costs and sometimes reducing line-haul cost.

Total logistics cost

A simplified annual network-cost model is:

where:

• TC = total annual logistics cost;
• F = facility operating and fixed costs;
• H = inventory holding cost;
• T = transportation cost;
• P = packaging and handling cost;
• R = cost of returns, damage and service failures;
• E = environmental or carbon-related cost, when monetized.

All terms must use the same currency and time period. The model is useful for comparing alternatives, but it is incomplete unless service, risk and capacity constraints are also evaluated.

A common annual holding-cost approximation is:

where i is the annual holding-rate fraction, V is unit value in currency per unit, and Ī is average inventory in units. It assumes inventory value and the holding rate are reasonably stable. If inventory has multiple products, the calculation should be performed by product or value class.

Facility-location logic

The centre-of-gravity method can provide an initial geographic estimate, not a final decision. For demand points i with coordinates (x_i,y_i) and weights w_i such as annual volume, the weighted location is:

It assumes transportation cost is approximately proportional to volume and distance, terrain and legal constraints are not dominant, and coordinates are meaningful in a local planar approximation. A real design must additionally examine taxes, labour, utilities, land, congestion, security, supplier access, disaster exposure, emissions and customer-time promises.

A more formal capacitated facility-location model uses binary variable y_j = 1 if facility j opens and x_ij = quantity shipped from facility j to customer i. A representative objective is:

subject to:

with x_ij ≥ 0 and y_j ∈ {0,1}. Here f_j is facility fixed cost, c_ij is variable shipment cost, d_i is demand and K_j is capacity. The first constraint requires demand satisfaction; the second prevents shipments from a closed facility and limits capacity. Service-time, product compatibility, emissions and disruption constraints may be added. This is a mixed-integer linear model when costs and constraints are linear; exact optimization can become computationally difficult as the network grows, so heuristics, scenario analysis and commercial solvers are often used.

Worked example: comparing two network options

An entrepreneur is comparing one national DC with two regional DCs. Annual fixed costs are 180,000 currency units for the national option and 2 × 125,000 for the regional option. Transport and handling costs are estimated at 420,000 and 300,000 respectively. Inventory holding is 90,000 and 155,000; returns and packaging are 40,000 and 50,000.

National total = 180,000 + 420,000 + 90,000 + 40,000 = 730,000.

Regional total = 250,000 + 300,000 + 155,000 + 50,000 = 755,000.

The national design is cheaper by 25,000 under these assumptions. However, if the regional design reduces lost sales by more than 25,000, or materially lowers risk and delivery time, it may still create greater total value. Cost comparison must therefore be accompanied by service and sensitivity analysis.

3. Warehousing and fulfillment design

Warehouses perform receiving, inspection, put-away, storage, replenishment, picking, packing, staging, dispatch and information capture. The design should follow product velocity, cube, value, fragility, expiry and order profile rather than habit.

Common configurations include:

• Centralized DC: efficient inventory pooling and line-haul, but longer delivery distances.
• Regional DC network: faster delivery and local market access, but duplicated stock and facilities.
• Cross-dock: appropriate for predictable, fast-moving flows with synchronized inbound and outbound schedules.
• Store fulfillment: uses retail inventory for online orders; it can shorten delivery but may disrupt shoppers and create inventory-accuracy problems.
• Micro-fulfillment: small, often automated facilities near demand; useful for dense urban markets, though capital and assortment constraints are significant.

Picking methods include discrete picking, batch picking, zone picking and wave picking. Batch picking combines similar orders; zone picking assigns workers to areas; wave picking releases work according to dispatch windows. Automation can improve throughput and accuracy, but only when data quality, process stability and volume justify the investment.

Warehouse productivity should be measured with metrics such as order accuracy, lines picked per labour hour, dock-to-stock time, inventory record accuracy, space utilization, damage rate and perfect-order percentage. High utilization is not always desirable: congestion and lack of staging space can lower throughput.

4. Transportation modes and routing

Mode selection should consider cost, speed, reliability, capacity, accessibility, product characteristics, emissions and risk.

Intermodal transportation combines modes, such as ocean–rail–road, to exploit the advantages of each. Incoterms allocate responsibilities, costs and risk between trading parties; they should be aligned with insurance, customs capability and actual control of the shipment.

Routing

A route plan must respect vehicle capacity, customer time windows, driver regulations, loading compatibility and depot constraints. The vehicle-routing problem is computationally difficult in its general form. Practical methods include savings heuristics, insertion heuristics, local search, mixed-integer optimization for smaller instances and dynamic routing supported by real-time data.

A basic routing objective is:

where z_ij is 1 if an arc from node i to node j is used and c_ij is its cost. This objective is incomplete unless constraints ensure that every customer is visited appropriately, routes begin and end at a depot, capacity is respected and subtours are prevented. In practice, minimizing distance alone can produce poor routes if delivery windows, service times or reliability are ignored.

5. Packaging and material-handling decisions

Packaging protects products, enables unitization, communicates handling requirements and affects transport density. Primary packaging contacts the product; secondary packaging groups units; tertiary packaging supports handling and transport.

Effective packaging design balances:

• protection against shock, vibration, compression, moisture and temperature;
• cube efficiency and palletization;
• opening and repacking convenience;
• tamper evidence and product safety;
• recyclability, reusability and material reduction;
• regulatory labelling and accessibility.

Dimensional weight is important in parcel networks. A carrier may charge on the greater of actual mass and dimensional mass. A generic dimensional-mass expression is:

where L, W and H are package dimensions in the carrier’s specified units, D is the carrier divisor and m_d is dimensional mass in the corresponding mass unit. The divisor and rounding rules vary by carrier and country; therefore, the contract must be checked before using the formula. Reducing empty space can lower both damage and freight cost, but over-compression can increase damage.

6. Cold chains and regulated distribution

A cold chain preserves required temperature conditions from production to final use. It includes qualified packaging, preconditioning, refrigerated storage, vehicles, monitoring, handling procedures and contingency capacity. The relevant requirement is not simply “keep it cold”; each product has an acceptable temperature range, exposure limit and sometimes humidity or light specification.

A robust cold-chain program includes:

1. product-specific temperature mapping and risk assessment;
2. validated packaging and vehicle performance;
3. calibrated sensors and continuous or event-based monitoring;
4. documented handover and excursion procedures;
5. qualified carriers and trained personnel;
6. backup power, alternate routes and emergency storage;
7. traceability by batch or lot where required.

Temperature excursions should be quarantined and assessed by authorized technical personnel. Automatically releasing or discarding every excursion can be wasteful; automatically accepting every excursion can compromise safety. In pharmaceuticals, food and biologics, legal standards and product-specific validation override general logistics preferences.

7. Omnichannel distribution and last-mile delivery

Multichannel distribution uses several channels; omnichannel distribution integrates those channels so that customers experience consistent assortment, price, inventory visibility, service and returns. Typical models include ship-from-DC, ship-from-store, click-and-collect, locker collection, marketplace fulfillment and direct-to-consumer shipping.

Omnichannel complexity arises because one inventory pool may serve stores, websites, marketplaces and wholesale customers with different promises. Accurate available-to-promise inventory is essential. A useful allocation principle is to reserve inventory for priority orders while considering margin, promised date, substitution and cancellation risk. The lowest-cost shipment is not necessarily the best choice if it causes a missed promise.

Last-mile delivery is often the most expensive and least predictable segment because stops are dispersed and customers may be absent. Solutions include delivery windows, lockers, pickup points, parcel consolidation, route density, electric vehicles, cargo bikes, crowdsourced delivery and customer incentives for flexible timing. The appropriate solution depends on urban form, product size, security, labour rules and service expectations.

A service-cost dashboard should combine:

• on-time-in-full delivery;
• order cycle time and promise accuracy;
• first-attempt delivery success;
• cost per order and cost per stop;
• damage and complaint rate;
• emissions per shipment;
• customer effort and satisfaction.

8. Reverse logistics and circular flows

Reverse logistics manages movement from the customer or user back to a seller, manufacturer, repair centre, recycler or disposal point. Drivers include defects, consumer returns, warranty claims, reusable assets, recalls, recycling obligations and product take-back schemes.

The reverse process normally involves authorization, collection, consolidation, inspection, disposition and financial settlement. Possible dispositions are resale as new, resale as refurbished, repair, harvesting components, recycling or controlled disposal. The economically best disposition depends on recovery value, processing cost, time, legal obligations and data security.

Returns policies influence demand and behaviour. A generous policy may increase conversion but can also increase bracketing, fraud and avoidable transport. Governance should include condition grading, serial-number controls, fraud screening, clear customer communication and privacy protection when returned devices contain personal data. Environmental claims must distinguish genuine reduction from shifting impacts to another stage.

9. Sustainability, resilience and governance

Logistics emissions arise from fuel and electricity, vehicle manufacture, refrigeration, facilities, packaging and failed or repeated deliveries. A practical reduction hierarchy is: avoid unnecessary movement, improve load and route utilization, shift to lower-emission modes, electrify suitable operations, use renewable energy and measure residual emissions transparently.

Carbon accounting requires a defined boundary, activity data and emission factors. A simplified transport estimate is:

where distance and mass use units consistent with the emission factor EF, and CO₂e is estimated carbon-dioxide equivalent. Real calculations may require vehicle type, load factor, fuel, empty running, refrigeration and well-to-wheel versus tank-to-wheel boundaries. Results should not be compared unless boundaries and methods are consistent.

Resilience is the ability to prepare for, absorb, recover from and adapt to disruption. Design measures include dual sourcing, geographic diversification, postponement, strategic safety stock, alternate carriers, flexible contracts, visibility, scenario planning and mutual-aid agreements. Resilience has a cost; therefore, firms should prioritize vulnerabilities with high probability, high consequence or limited recovery options.

Governance concerns include carrier safety, worker conditions, modern-slavery risk, privacy, dangerous goods, customs compliance, competition law, sanctions, product traceability and algorithmic fairness. AI may support demand forecasting, ETA prediction, slotting, image-based damage inspection and route optimization. It should not be treated as an autonomous substitute for governance: biased data, unstable patterns, opaque decisions and cybersecurity attacks can create operational harm. Human override, audit trails, access controls and performance monitoring are necessary.

Applied case: GreenBasket’s regional fulfillment decision

Background

GreenBasket, an illustrative online grocery venture, serves 18,000 weekly orders from one metropolitan warehouse. Its 24-hour promise is attractive, but delivery cost is rising, chilled-product spoilage is 2.8%, and first-attempt delivery success is 86%. Management is considering a second micro-fulfillment site near the city centre.

Problem

The proposed site would reduce average delivery distance and support chilled capacity, but it would duplicate inventory, require new staff and increase energy consumption. The decision cannot be made using delivery distance alone.

Analysis

The team segmented orders by postcode, basket size, temperature class and promised window. It found that 42% of orders originated in a dense central zone, where delivery density was high but congestion caused delays. It modelled three alternatives: retain the current warehouse, open a full second site, or use a small chilled cross-dock plus pickup lockers. Scenarios included 10% demand growth, a fuel-price increase, a refrigeration outage and a carrier strike.

The evaluation criteria were total cost, on-time-in-full performance, spoilage, emissions, labour feasibility and recovery time after disruption. The analysis showed that a full second site achieved the fastest delivery but had the highest inventory duplication. The chilled cross-dock with lockers produced slightly less speed improvement but substantially lowered failed deliveries and spoilage. It also created a fallback point during disruption.

Approach and solution

GreenBasket selected a phased design: lease a small centrally located chilled cross-dock, retain ambient inventory at the original warehouse, introduce pickup lockers near transit stations, and use dynamic routing only within approved delivery windows. It established temperature monitoring, batch-level traceability and a contingency carrier contract. The pilot was limited to high-density postcodes and measured results against a pre-pilot baseline.

Outcome

In the illustrative pilot, first-attempt success increased to 93%, chilled spoilage fell to 1.4%, and average delivery distance fell by 18%. Operating cost per order initially increased by 3% because of the cross-dock lease and monitoring system, but total cost per successful order fell after failed-delivery costs and spoilage were included. The business postponed a full second site until demand justified the additional fixed cost.

Learning points

The case demonstrates that network design should evaluate successful fulfillment rather than nominal shipment cost; segment products and customers; treat cold-chain controls as a design requirement; test alternatives through pilots; and value resilience and failure reduction alongside direct transport savings.

Implementation framework and best practices

A disciplined network-design project can follow these stages:

1. Define customer promises, products, geography, legal constraints and planning horizon.
2. Clean demand, cost, lead-time, capacity and facility-location data.
3. Map current flows and identify service failures, bottlenecks and hidden costs.
4. Develop alternative network structures rather than optimizing only the current design.
5. Model cost, capacity, service, emissions and disruption scenarios.
6. Validate assumptions with operators, customers, carriers, finance and regulators.
7. Pilot changes where uncertainty is high.
8. Implement with clear ownership, system integration and change management.
9. Monitor leading and lagging metrics and revisit the design as demand changes.

Best practice is not to maximize one metric. A balanced scorecard should show cost-to-serve, inventory, capacity utilization, delivery reliability, customer outcomes, emissions, safety and resilience. Management should also distinguish controllable performance from external conditions and avoid incentives that create local optimization—for example, maximizing vehicle utilization at the expense of promised delivery time.

Challenges and limitations

Network models simplify reality. Demand may be seasonal, correlated and affected by promotions; costs may be nonlinear; facility capacity may be uncertain; and political, climate or labour conditions may change. Historical data can encode past stockouts and therefore understate true demand. Optimized plans may fail if master data, inventory accuracy or execution discipline is weak. Environmental comparisons can be misleading when system boundaries differ. Finally, mathematical precision does not remove managerial judgment: stakeholder acceptance, ethics, legal compliance and strategic flexibility remain essential.

Chapter Summary

Logistics and distribution design align facilities, inventory, transportation, information and customer promises. Strategic network choices balance facility, inventory, transport, packaging, returns and environmental costs against service and risk. Facility-location models, weighted geographic methods and capacitated optimization provide useful analytical foundations, but their assumptions must be tested.

Warehousing and fulfillment should reflect product velocity, order structure and service requirements. Mode and route decisions must consider time, reliability, capacity, emissions and regulatory constraints. Packaging affects protection, dimensional freight and sustainability. Cold chains require validated conditions, monitoring and excursion governance. Omnichannel distribution integrates channels, while last-mile design focuses on density, delivery success and customer convenience. Reverse logistics creates value through recovery, repair and responsible disposition.

Resilient and sustainable networks combine scenario analysis, diversification, visibility, efficient loads, lower-emission modes and credible measurement. Technology and AI can improve decisions but require human oversight, data governance and cybersecurity. The strongest entrepreneurial approach is phased, evidence-based and evaluated on total successful fulfillment rather than isolated logistics cost.

Key Terms

Logistics: Planning and control of product, information and related flows.

Distribution: Outbound movement and fulfillment of products to customers or channels.

Network design: Determination of facility structure, flows, capacities and service assignments.

Distribution centre: Facility receiving, sorting, storing and dispatching products.

Fulfillment centre: Facility optimized for processing customer orders.

Cross-docking: Transfer from inbound to outbound flow with minimal storage.

3PL: External provider performing logistics activities.

Service level: Defined probability or proportion of demand served under specified conditions.

Total logistics cost: Combined facility, inventory, transport, handling, return and related costs.

Facility location: Selection of geographic sites for logistics activities.

Intermodal transport: Movement using two or more coordinated transport modes.

Vehicle-routing problem: Design of routes subject to visits, capacity and operational constraints.

Dimensional weight: Freight charge basis calculated from package volume.

Cold chain: Temperature-controlled supply and distribution system.

Omnichannel distribution: Integrated management of multiple customer channels.

Last mile: Final movement from a local node to the customer.

Reverse logistics: Movement and management of returns, recovery and end-of-life products.

Postponement: Delaying final configuration or commitment until demand is clearer.

Resilience: Capacity to withstand, recover from and adapt to disruption.

Perfect order: Order delivered completely, accurately, on time and without damage.

Review Questions

Short Answer Questions

1. Distinguish strategic, tactical and operational logistics decisions.
2. What is cross-docking, and when is it appropriate?
3. State two advantages and two limitations of a regional distribution-centre network.
4. Why can dimensional weight increase parcel cost?
5. Identify four controls required in a cold chain.

Long Answer Questions

1. Explain how facility, inventory, transport and service costs interact in network design.
2. Compare road, rail, ocean, air and parcel transportation for an entrepreneurial firm.
3. Discuss how omnichannel distribution changes inventory and fulfillment decisions.
4. Explain the role of reverse logistics in a circular supply chain.
5. Evaluate the benefits and risks of using AI for routing and logistics forecasting.

Analytical / Application Questions

1. A firm must choose between one DC and two regional DCs. Design a total-cost and service analysis, identify required data, and explain how you would test demand-growth and disruption scenarios.
2. A pharmaceutical distributor experiences repeated temperature excursions. Diagnose possible process and network causes and propose a corrective cold-chain program.
3. An e-commerce company has high failed-delivery rates in a dense city. Recommend a last-mile redesign that balances cost, customer convenience, emissions and resilience.

Multiple Choice Questions with Answers

Question 1

Which decision is primarily strategic in logistics network design?

A. Selecting a delivery sequence for today

B. Choosing the location of regional distribution centres

C. Releasing a picking wave

D. Printing a dispatch label

Correct Answer: B. Choosing the location of regional distribution centres

Question 2

In the simplified total logistics cost model, which item represents the cost of facilities?

A. F

B. H

C. T

D. R

Correct Answer: A. F

Question 3

The centre-of-gravity method weights customer coordinates primarily by which factor?

A. Product colour

B. Annual demand or shipment volume

C. Driver seniority

D. Packaging material type

Correct Answer: B. Annual demand or shipment volume

Question 4

Which transportation mode generally offers the fastest movement but the highest unit cost?

A. Ocean

B. Rail

C. Air

D. Pipeline

Correct Answer: C. Air

Question 5

What is the principal purpose of cross-docking?

A. To maximize long-term storage

B. To transfer inbound goods to outbound flows with minimal storage

C. To increase inventory duplication

D. To inspect every product at the customer site

Correct Answer: B. To transfer inbound goods to outbound flows with minimal storage

Question 6

A package measures 40 cm × 30 cm × 20 cm and the carrier divisor is 5,000 cm³/kg. Its dimensional mass is:

A. 2.4 kg

B. 4.8 kg

C. 12 kg

D. 24 kg

Correct Answer: B. 4.8 kg

Question 7

Which practice is most directly associated with cold-chain governance?

A. Ignoring short temperature excursions

B. Using uncalibrated sensors

C. Validating packaging and monitoring temperature conditions

D. Removing batch traceability

Correct Answer: C. Validating packaging and monitoring temperature conditions

Question 8

What distinguishes omnichannel distribution from merely multichannel distribution?

A. It uses only physical stores

B. It integrates channels around a consistent customer experience

C. It prohibits online returns

D. It eliminates inventory records

Correct Answer: B. It integrates channels around a consistent customer experience

Question 9

Which is a reverse-logistics activity?

A. Delivering raw materials to a plant

B. Routing a new order to a customer

C. Inspecting and refurbishing a returned product

D. Forecasting next year's sales

Correct Answer: C. Inspecting and refurbishing a returned product

Question 10

Why should a routing model include time windows and vehicle capacity?

A. Distance is never relevant

B. They ensure routes are operationally feasible

C. They eliminate the need for customer data

D. They guarantee the lowest possible emissions

Correct Answer: B. They ensure routes are operationally feasible