12. Logistics Costs, Performance Measurement and Analytics

Introduction and Context

Logistics performance cannot be managed effectively without understanding both cost and value. A distribution network may achieve high delivery reliability while consuming excessive transport, inventory, handling, and administrative resources. Conversely, an apparently low-cost operation may create stock-outs, customer dissatisfaction, expedited freight, product damage, or excessive working capital. The managerial task is therefore not simply to minimise logistics expenditure. It is to design and control a system that provides the required service at the lowest economically sustainable total cost.

This chapter examines the principal methods used to measure and improve logistics performance. It connects activity-based costing, landed cost, cost-to-serve, working-capital analysis, SCOR processes, key performance indicators, balanced scorecards, benchmarking, dashboards, and descriptive, predictive, and prescriptive analytics. The central theme is alignment: measures should reflect strategic priorities, costs should be assigned to the activities and customers that generate them, and analytical outputs should support decisions rather than merely describe historical performance.

Logistics measurement is particularly important because supply-chain decisions involve trade-offs. More inventory can improve availability but increases capital and obsolescence costs. Faster transport can reduce lead time but raise freight expenditure. Consolidating shipments may lower unit transport cost while reducing responsiveness. A sound measurement system makes these trade-offs visible and enables managers to evaluate total supply-chain consequences.

Learning Objectives

After studying this chapter, students should be able to:

1. Explain the role and structure of logistics costs in supply-chain decisions.

2. Apply activity-based costing, landed-cost analysis, and cost-to-serve analysis.

3. Calculate and interpret inventory, working-capital, service, productivity, and asset-utilisation measures.

4. Use SCOR processes and balanced scorecards to organise performance measurement.

5. Distinguish descriptive, predictive, and prescriptive analytics in logistics.

6. Design meaningful dashboards and evaluate benchmarking results.

7. Identify measurement risks, data-governance requirements, and ethical issues.

8. Recommend evidence-based improvements in logistics performance.

The Economics of Logistics Performance

Logistics cost includes the resources consumed in planning, purchasing support, transportation, warehousing, inventory holding, order processing, packaging, information management, returns, and related administration. Traditional accounting often reports these costs by department or legal entity. Such reporting is useful for financial control but may not reveal the cost of serving a particular product, order, channel, or customer.

A useful distinction is between direct and indirect logistics costs. Direct costs can be traced economically to a shipment or activity, such as carrier charges for a particular consignment. Indirect costs support several activities and require allocation, such as warehouse supervision, information systems, facility rent, or network planning. Allocation is not merely an accounting exercise: an inappropriate allocation may make a customer or product appear profitable when it consumes substantial logistics resources.

Logistics decisions should be assessed using total cost rather than isolated cost categories. Total logistics cost can be represented conceptually as:

Total logistics cost = transportation + warehousing + inventory carrying cost + order processing and information cost + packaging and handling + returns and quality-related cost + logistics administration.

The categories interact. A company that reduces warehouse locations may increase outbound distance. A firm that purchases larger lots may obtain a lower unit price but incur more inventory carrying and obsolescence cost. The relevant question is whether the total economic outcome improves while the required service level is maintained.

Cost behaviour also matters. Fixed costs, such as facility leases, do not vary immediately with volume. Variable costs, such as fuel-linked transport charges or handling labour, change with activity. Step costs arise when volume crosses a capacity threshold and additional vehicles, shifts, or facilities are required. Understanding cost behaviour supports decisions about outsourcing, capacity, shipment consolidation, and network design.

Activity-Based Costing in Logistics

Activity-based costing (ABC) assigns resource costs to activities and then assigns activity costs to cost objects through cost drivers. In logistics, cost objects may be customers, products, channels, orders, routes, or services. The method is valuable where overheads are substantial and different users consume logistics resources unevenly.

The ABC sequence is:

1. Identify significant activities, such as receiving, put-away, picking, packing, dispatch, delivery, invoicing, and returns processing.

2. Accumulate resource costs associated with each activity.

3. Select a cost driver that reflects consumption, such as pallets received, order lines picked, kilometres travelled, deliveries made, or return units.

4. Calculate an activity rate: activity cost divided by practical activity capacity or expected activity volume.

5. Assign activity costs to products, customers, or orders according to actual driver usage.

For example, if a warehouse spends $120,000 on picking and processes 60,000 order lines, the indicative picking rate is $2 per order line. A customer order containing 15 lines would receive $30 of picking cost, before packing, transport, and other activities are considered. ABC becomes more informative when it distinguishes activities with different complexity, such as standard picking, fragile-item handling, temperature-controlled handling, and urgent fulfilment.

ABC has limitations. Establishing reliable activity data can be expensive, cost drivers may be imperfect, and managers may confuse accounting precision with economic truth. Rates should therefore be reviewed when processes, automation, volumes, or service policies change. ABC is most useful when it changes a decision—for example, by revealing that a low-margin customer generates unusually costly small orders and returns.

Landed Cost and Cost-to-Serve

Landed cost is the total cost of acquiring and making a product available at its destination. It commonly includes purchase price, international and domestic freight, insurance, duties and taxes where applicable, port and terminal charges, customs brokerage, documentation, handling, inspection, packaging changes, and other directly attributable costs. The exact definition should be agreed within the organisation because tax treatment and commercial terms differ.

A simplified landed-cost calculation is:

Landed cost per unit = purchase cost + freight and insurance + duties and border charges + handling and delivery costs, divided by received saleable units.

The calculation must account for losses, damage, shrinkage, quality failures, and exchange-rate effects when they materially affect the cost of saleable inventory. Incoterms and ownership transfer points are important because they determine which party bears specific transport, risk, and customs responsibilities.

Cost-to-serve extends the analysis beyond acquisition. It estimates the total cost incurred to serve a particular customer, channel, product family, or order profile. Relevant components may include order entry, account management, picking complexity, delivery frequency, special packaging, minimum-order exceptions, returns, claims, discounts associated with service, and payment-related administration. Revenue minus product cost and cost-to-serve provides a more realistic view of customer or channel profitability than gross margin alone.

Cost-to-serve analysis should not be used to penalise customers mechanically. It can support differentiated service design. A customer generating expensive urgent deliveries might accept delivery-day consolidation, minimum order quantities, a revised delivery window, or a service fee. The objective is to improve the fit between service requirements and economic value while preserving fairness and contractual obligations.

Working Capital and Asset Productivity

Logistics has a direct effect on working capital because inventory, receivables, and payables are tied to the operating cycle. Important measures include:

• Inventory turnover = cost of goods sold divided by average inventory.
• Days inventory outstanding (DIO) = average inventory divided by cost of goods sold, multiplied by the number of days in the period.
• Days sales outstanding (DSO) = average receivables divided by sales, multiplied by days in the period.
• Days payables outstanding (DPO) = average payables divided by purchases or cost of goods sold, multiplied by days in the period, subject to the organisation’s accounting convention.
• Cash conversion cycle = DIO + DSO − DPO.

These measures should be interpreted with care. A high inventory turnover may indicate efficiency, but it may also reflect under-stocking and lost sales. DPO may improve cash position but damage supplier relationships if payment practices become abusive. Working-capital improvement must therefore be assessed alongside service, resilience, quality, and supplier health.

Inventory performance can be decomposed into cycle stock, safety stock, pipeline inventory, seasonal inventory, and obsolete or slow-moving stock. Managers should examine the causes of inventory rather than rely on a single aggregate ratio. Forecast error, long replenishment lead time, high minimum order quantities, unreliable suppliers, and poor master data may each require different interventions.

SCOR and the Architecture of Measurement

The Supply Chain Operations Reference (SCOR) framework provides a process-oriented structure for analysing supply-chain performance. Its commonly used process categories include Plan, Source, Make, Deliver, and Return, with enabling activities supporting governance, data, relationships, compliance, and capability. SCOR encourages organisations to connect strategic objectives, process performance, and operational measures.

A SCOR-oriented analysis asks several questions. How accurately are demand and supply balanced? How reliably are materials sourced? How efficiently are products made or configured? How consistently are orders delivered? How effectively are returns and reverse flows managed? Measures should be assigned to process owners and linked across functional boundaries.

SCOR is not a substitute for organisational strategy. It is a process language and diagnostic framework. A company competing on rapid customisation may emphasise order fulfilment cycle time and flexibility, while a low-cost commodity network may give greater weight to asset utilisation and cost efficiency. The same process framework can therefore support different strategic priorities.

Selecting KPIs That Drive Behaviour

A key performance indicator (KPI) is a measure selected because it provides meaningful evidence about an objective or critical process. A metric becomes strategically useful when it has a clear definition, owner, data source, reporting frequency, target, and decision rule.

Common logistics KPIs include:

• Perfect order rate: the proportion of orders delivered complete, on time, damage-free, and with accurate documentation.
• On-time-in-full (OTIF): orders meeting the agreed delivery-time and quantity requirements.
• Order cycle time: elapsed time from order receipt or confirmation to delivery, according to the defined boundary.
• Fill rate: the proportion of demand supplied from available stock within the agreed service conditions.
• Forecast accuracy: the degree to which forecasts correspond to actual demand, using an explicitly selected error measure.
• Freight cost per unit, order, weight, or shipment value.
• Warehouse productivity, such as lines picked per labour hour.
• Dock-to-stock time: time from receipt to inventory availability.
• Inventory accuracy: agreement between recorded and physically available stock.
• Damage, claims, return, and shrinkage rates.
• Capacity utilisation and asset availability.

Definitions matter. OTIF can be overstated if promised dates are repeatedly changed, while a low cost per shipment may reflect large batch sizes that harm customer responsiveness. KPIs should be balanced across cost, service, speed, quality, flexibility, sustainability, and asset utilisation. A measure that rewards one dimension without recognising its consequences may create dysfunctional behaviour.

Balanced Scorecards and Benchmarking

The balanced scorecard translates strategy into a set of related objectives and measures. In logistics, four perspectives can be adapted as follows:

1. Financial: total logistics cost, cost-to-serve, working capital, and return on logistics assets.

2. Customer: OTIF, perfect order rate, complaint resolution, availability, and perceived reliability.

3. Internal process: cycle time, inventory accuracy, picking productivity, forecast bias, and returns processing.

4. Learning and capability: employee competence, safety, digital adoption, process-improvement activity, and data quality.

A strategy map can show cause-and-effect assumptions. Better training and data quality may improve planning and warehouse execution; improved execution may raise service reliability; reliable service may support retention and profitability. These relationships should be tested rather than assumed.

Benchmarking compares performance with an internal unit, historical standard, peer organisation, industry reference, or best-in-class process. Internal benchmarking is often more reliable because definitions and data structures are comparable. External benchmarking may provide strategic insight but can be distorted by different product mixes, geography, labour costs, service promises, accounting rules, and capacity conditions. Benchmarking should therefore focus on understanding performance gaps and practices, not simply copying another organisation’s target.

Dashboards and the Analytics Hierarchy

A logistics dashboard is an interface that presents selected measures for monitoring and decision-making. Effective dashboards begin with decisions, not available data. A transport manager may need to know which routes are at risk today; a network executive may need to understand quarterly cost-to-serve and capacity trends. These users require different levels of aggregation, frequency, and visualisation.

A well-designed dashboard normally includes target and actual values, trend, variance, status, responsible owner, and the ability to drill down to causes. It should distinguish leading indicators, such as forecast bias or supplier confirmation reliability, from lagging indicators, such as monthly cost or customer complaints. Excessive visual detail can obscure priorities. Data definitions and timestamps should be visible, especially when measures combine information from different systems.

Descriptive analytics answers what happened and what is happening. It includes reporting, variance analysis, Pareto analysis, heat maps, and trend analysis. Diagnostic analysis, often treated as an extension of descriptive work, asks why a result occurred by examining drivers and exceptions.

Predictive analytics estimates what is likely to happen. Applications include demand forecasting, estimated arrival times, stock-out risk, late-delivery prediction, equipment failure prediction, and return-volume forecasting. Models may use historical transactions, lead times, weather, traffic, promotions, order characteristics, and supplier signals. Prediction is probabilistic; a forecast should be accompanied by uncertainty and assumptions.

Prescriptive analytics recommends what should be done. Examples include inventory-position optimisation, transport-mode selection, shipment consolidation, warehouse slotting, vehicle routing, and allocation of scarce stock. Optimisation models require an objective function, constraints, and data. A mathematically optimal answer is not necessarily implementable if it ignores labour rules, customer commitments, safety, or supplier relationships.

Artificial intelligence and machine learning can enhance pattern recognition and exception management, but they do not eliminate the need for process knowledge. Poor master data, biased historical decisions, changing market conditions, and unclear objectives can produce unreliable recommendations. Human review remains important for high-impact decisions.

Applied Case: NorthStar Medical Devices

Background

NorthStar Medical Devices is an illustrative manufacturer supplying hospitals through direct orders and regional distributors. It offers standard products and temperature-sensitive diagnostic kits. Management believed that the direct hospital channel was more profitable than the distributor channel because its average selling price was higher.

Situation and Problem

The company experienced rising freight expenditure, frequent emergency shipments, and increasing inventory. The finance department reported acceptable gross margins, but customer complaints concerned late deliveries and incomplete orders. Senior management needed to determine whether the problem was pricing, warehouse execution, transport policy, or channel design.

Analysis

NorthStar created an ABC model covering order entry, quality checks, temperature-controlled packing, picking, dispatch, delivery, invoice correction, and returns. It then calculated landed cost by product family and cost-to-serve by customer segment. The analysis showed that several small hospital accounts placed frequent low-line orders and required specialised packaging. Their revenue was attractive, but their cost-to-serve consumed most of the apparent margin. Some distributor orders were larger and more predictable, producing lower handling and transport cost per unit.

Working-capital analysis showed that inventory turnover had fallen because safety stocks were set using outdated lead-time assumptions. The dashboard also revealed that the reported OTIF rate excluded orders split across multiple deliveries. When the measure was redefined to reflect complete delivery, performance was materially lower. A predictive model identified orders at high risk of late delivery using carrier capacity, destination, order size, and dispatch timing.

Approach and Solution

NorthStar introduced delivery calendars for selected hospital accounts, consolidated compatible orders, and offered a transparent premium for urgent temperature-controlled shipments. It revised safety-stock parameters using updated demand and lead-time variability. Warehouse supervisors received a dashboard showing dock-to-stock time, incomplete orders, picking exceptions, and imminent service risks. The company also established a balanced scorecard so that freight savings could not be pursued at the expense of service or product integrity.

Outcome

The illustrative case assumes that NorthStar achieved a reduction in emergency shipments and improved inventory productivity after implementation, while maintaining critical hospital service commitments. More importantly, management changed its understanding of profitability: channel economics depended not only on price and product margin but also on the pattern of logistics activities consumed.

Learning Points

The case demonstrates five principles. First, gross margin is not customer profitability. Second, KPI definitions influence reported performance. Third, working-capital measures require service interpretation. Fourth, predictive alerts are useful when linked to operational action. Finally, cost reduction is sustainable only when it respects product quality, customer requirements, and clinical risk.

Risks, Governance, and Professional Judgement

Measurement systems can fail through poor data, inconsistent definitions, delayed reporting, inappropriate targets, or intentional manipulation. Common gaming behaviours include changing promised dates to improve OTIF, postponing problem transactions until the next reporting period, prioritising easy orders, or reducing inventory below a safe level to improve turnover.

Governance should establish a metric dictionary, ownership, data lineage, calculation rules, validation routines, and access controls. Changes to definitions should be documented and reported transparently. Personal data in delivery records must be protected, and supplier or customer information should be used only for legitimate purposes. Automated decisions should be explainable to affected stakeholders when they influence allocation, service access, employment, or commercial treatment.

Ethically, logistics managers must not pursue low cost through unsafe labour practices, deceptive reporting, discriminatory service, or unreasonable pressure on carriers and suppliers. Sustainability measures should avoid selective accounting. Emissions reporting, for example, should state boundaries, assumptions, and whether estimates include transport, warehousing, packaging, and returns.

Implementation Guidance and Emerging Directions

A practical measurement programme can proceed through the following sequence:

1. Clarify strategic priorities and service promises.

2. Map end-to-end processes and decision points.

3. Define cost objects, activities, KPIs, and calculation boundaries.

4. Assess data quality and establish governance.

5. Build a small number of decision-oriented dashboards.

6. Pilot measures in one product, site, or customer segment.

7. Test for unintended behavioural effects.

8. Integrate descriptive, predictive, and prescriptive capabilities progressively.

9. Review targets as strategy, technology, and market conditions change.

Emerging practice is moving toward event-driven visibility, control towers, digital twins, real-time estimated-arrival information, automated exception management, and integrated carbon-cost analysis. These developments can improve responsiveness, but visibility is not the same as performance. Organisations still need disciplined processes, accountable decisions, and economically meaningful measures. The most valuable analytics are those that connect a signal to an action, an owner, and an outcome.

Chapter Summary

Logistics performance management combines cost economics, operational measurement, and analytics. Activity-based costing assigns resource costs according to activities and drivers, while landed cost captures the complete cost of making goods available at destination. Cost-to-serve reveals the economic consequences of differing customer and channel requirements. Working-capital measures show how inventory and operating-cycle decisions affect cash, but must be interpreted with service and resilience indicators.

SCOR provides a process language for connecting planning, sourcing, making, delivery, returns, and enabling capabilities. KPIs should be clearly defined, balanced, owned, and linked to decisions. Balanced scorecards connect financial, customer, process, and capability outcomes. Benchmarking is useful when differences in context are recognised. Dashboards should present relevant trends, variances, exceptions, and causes rather than excessive data.

Descriptive analytics explains what happened, predictive analytics estimates what may happen, and prescriptive analytics recommends actions under defined objectives and constraints. Effective application requires data governance, ethical judgement, human oversight, and continuous testing for unintended consequences. The ultimate purpose of measurement is not measurement itself, but better total-cost, service, asset, and strategic decisions.

Key Terms

Activity-based costing: A costing method that assigns resource costs to activities and then to cost objects through cost drivers.

Landed cost: The total cost of acquiring and making a product available at its destination.

Cost-to-serve: The total logistics and related service cost incurred for a customer, product, channel, or order profile.

Cost driver: A measurable factor that causes or represents consumption of an activity.

Working capital: Short-term operating investment represented principally by inventory, receivables, and payables.

Cash conversion cycle: DIO plus DSO minus DPO, indicating the time cash is tied up in operations.

Inventory turnover: Cost of goods sold divided by average inventory.

SCOR: A process-oriented framework for analysing supply-chain activities and performance.

KPI: A selected measure used to evaluate progress toward an important objective or process result.

OTIF: The proportion of orders delivered on time and in the required quantity.

Perfect order: An order delivered complete, on time, damage-free, and with accurate documentation.

Balanced scorecard: A strategic measurement system combining financial, customer, internal-process, and learning perspectives.

Benchmarking: Comparing performance or practices with a defined internal, external, historical, or best-practice reference.

Dashboard: A visual interface presenting selected measures, trends, variances, and exceptions for decision-making.

Descriptive analytics: Analysis that reports and investigates what has happened or is happening.

Predictive analytics: Analysis that estimates future outcomes or probabilities.

Prescriptive analytics: Analysis that recommends actions under objectives and constraints.

Data lineage: The documented origin, transformation, movement, and use of data.

Leading indicator: A measure that provides an early signal of a future outcome.

Lagging indicator: A measure that records an outcome after it has occurred.

Review Questions

Short Answer Questions

1. Why is total logistics cost more useful than analysing transport cost alone?

2. What is the role of a cost driver in activity-based costing?

3. Define cost-to-serve and distinguish it from gross margin.

4. What does the cash conversion cycle measure?

5. Differentiate descriptive, predictive, and prescriptive analytics.

Long Answer Questions

1. Explain how activity-based costing can be designed for a warehouse and discuss its limitations.

2. Discuss the role of SCOR in organising logistics performance measurement.

3. Explain how a balanced scorecard can prevent excessive focus on cost reduction.

4. Analyse the importance of KPI definitions, ownership, and governance.

5. Discuss how landed cost and cost-to-serve support channel and customer decisions.

Analytical / Application Questions

1. A customer generates high revenue but requires frequent small orders, special packaging, and urgent deliveries. Design a cost-to-serve analysis and recommend possible service alternatives.

2. A firm improves inventory turnover while stock-outs and customer complaints increase. Diagnose the likely measurement problem and propose a balanced KPI set.

3. Design a dashboard for a distribution centre manager that combines operational, financial, service, and predictive measures. Explain the action associated with each measure.

22. Multiple Choice Questions with Answers

Question 1

Which situation best demonstrates the value of activity-based costing?

A. Allocating all warehouse overhead equally across products

B. Assigning handling cost according to the number and complexity of activities consumed

C. Using sales revenue alone to distribute logistics cost

D. Recording only externally invoiced transport charges

Correct Answer: B. Assigning handling cost according to the number and complexity of activities consumed

Question 2

Which item is most directly associated with landed cost?

A. Corporate dividend payments

B. Employee training unrelated to logistics

C. Customs, freight, insurance, and destination handling

D. Future sales commission on the product

Correct Answer: C. Customs, freight, insurance, and destination handling

Question 3

A customer has high gross margin but low profitability after frequent urgent deliveries and returns. Which analysis reveals this problem most directly?

A. Cost-to-serve analysis

B. Inventory turnover analysis

C. Supplier defect analysis

D. Market-share analysis

Correct Answer: A. Cost-to-serve analysis

Question 4

If average inventory is $2 million and annual cost of goods sold is $12 million, inventory turnover is:

A. 2 times

B. 4 times

C. 6 times

D. 12 times

Correct Answer: C. 6 times

Question 5

Which practice is most likely to distort an OTIF measure?

A. Including complete and partial deliveries separately

B. Using an agreed customer promise date

C. Changing the promised date after a delay occurs

D. Recording quantity and date performance together

Correct Answer: C. Changing the promised date after a delay occurs

Question 6

Which balanced-scorecard perspective includes employee capability and data quality?

A. Financial

B. Customer

C. Internal process

D. Learning and capability

Correct Answer: D. Learning and capability

Question 7

A model estimates the probability that an order will arrive late. This is an example of:

A. Descriptive analytics

B. Predictive analytics

C. Prescriptive analytics

D. Activity-based costing

Correct Answer: B. Predictive analytics

Question 8

Which is the best example of prescriptive analytics?

A. Reporting last month’s freight cost

B. Displaying current warehouse utilisation

C. Estimating next week’s demand

D. Recommending shipment consolidation subject to delivery constraints

Correct Answer: D. Recommending shipment consolidation subject to delivery constraints

Question 9

Why should external benchmarking be interpreted cautiously?

A. External benchmarks never contain useful information

B. Different service, product, geography, and accounting conditions affect comparability

C. Internal performance cannot be measured

D. Benchmarking applies only to financial measures

Correct Answer: B. Different service, product, geography, and accounting conditions affect comparability

Question 10

Which governance practice most improves the credibility of a logistics dashboard?

A. Adding as many metrics as possible

B. Changing definitions whenever results are unfavourable

C. Maintaining a documented metric dictionary and data lineage

D. Allowing each department to calculate the same KPI differently

Correct Answer: C. Maintaining a documented metric dictionary and data lineage