Logistics Costing, Performance and Analytics

Introduction and Context

Logistics decisions determine how efficiently an organisation moves, stores and fulfils products. Transport, warehousing, inventory, order processing, packaging and information systems consume resources, but their costs are often distributed across different departments and accounting codes. As a result, a company may report a low transport cost while experiencing excessive inventory, stockouts, rework or customer-service failures.

Logistics costing and performance management provide a more complete view. Costing identifies the resources consumed by logistics activities and customers. Performance measurement evaluates whether those activities achieve operational and strategic objectives. Analytics converts transaction data into evidence for decisions such as route selection, warehouse design, inventory policy, outsourcing and service-level differentiation.

A sound logistics measurement system therefore connects three questions:

1. What resources are being consumed?
2. What operational results are being achieved?
3. What actions will improve total supply-chain value?

The central principle is to optimise total cost and customer value rather than minimise one isolated cost.

Learning Objectives

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

• classify and calculate major logistics costs;
• explain logistics performance indicators and balanced scorecards;
• apply activity-based costing to logistics activities;
• calculate break-even points and evaluate make-or-buy decisions;
• analyse cost–service trade-offs;
• interpret dashboards, benchmarks and basic logistics data;
• identify limitations, ethical issues and governance requirements in logistics analytics; and
• recommend practical performance improvements.

1. The Structure of Logistics Cost

Logistics cost is the monetary value of resources used to plan, procure, move, store, handle, process, monitor and deliver goods. A useful total-cost model is:

where:

• C_L = total logistics cost per period;
• C_T = transport cost;
• C_W = warehousing and material-handling cost;
• C_I = inventory carrying cost;
• C_O = order-processing and information cost;
• C_P = packaging cost attributable to logistics;
• C_Q = quality, error and service-failure cost; and
• C_R = reverse-logistics cost.

The model is an analytical framework, not a universal accounting standard. Organisations may classify packaging, quality or reverse logistics differently. The important requirement is consistent classification and avoidance of double counting.

1.1 Fixed, variable and semi-variable costs

• Fixed costs remain broadly unchanged within a relevant capacity range. Examples include warehouse rent, salaried supervision and leased equipment.
• Variable costs change with activity. Examples include fuel per kilometre, delivery labour per shipment and packaging materials per unit.
• Semi-variable costs contain both components, such as vehicle ownership plus mileage-related fuel and maintenance.
• Direct costs can be traced economically to a product, order, customer or route.
• Indirect costs support multiple activities and require allocation.

The relevant-cost concept is essential for decisions. A cost already incurred and unrecoverable is a sunk cost and should not influence a future make-or-buy decision. A cost that changes because of a decision is a relevant or avoidable cost.

1.2 Inventory carrying cost

Inventory creates service capability but ties up capital and generates storage, insurance, obsolescence and handling costs. A common annual estimate is:

where:

• C_ic = annual inventory-carrying cost;
• I_avg = average inventory quantity in units;
• v = value per unit in currency per unit; and
• r = annual carrying-rate proportion, expressed as a decimal.

The carrying rate may include cost of capital, insurance, storage, shrinkage and obsolescence. It must be defined carefully because warehouse operating expenses should not be added again if they are already included in r.

Example. Average inventory is 4,000 units, unit value is $25 and the carrying rate is 22% per year:

C_ic = 4,000 × 25 × 0.22 = $22,000 per year.

This is an estimate. It assumes average inventory is representative and that the rate applies uniformly to all units.

2. Activity-Based Costing in Logistics

Traditional overhead allocation often assigns warehouse or customer-service overhead using one broad basis, such as sales value or labour hours. This can distort decisions because customers and products consume activities differently. Activity-based costing (ABC) assigns resource costs first to activities and then to cost objects such as products, orders, customers or channels.

The main steps are:

1. identify significant logistics activities;
2. accumulate resource costs in activity cost pools;
3. select a cost driver that explains activity consumption;
4. calculate an activity cost rate;
5. assign costs to cost objects; and
6. compare cost and service results.

The basic rate is:

where R_a is the rate for activity a, C_a is the activity cost pool, and D_a is the practical quantity of the cost driver.

The cost assigned to a cost object is:

where C_ao is the activity cost assigned to object o, and d_ao is the object’s driver consumption.

The driver should have a credible causal relationship with resource consumption. Examples include order lines for picking, pallet movements for handling, kilometres for transport, number of deliveries for dispatch administration and number of returns for reverse logistics.

Worked ABC example

A warehouse incurs $120,000 for order picking and expects 30,000 order lines. The picking rate is:

R_pick = 120,000 ÷ 30,000 = $4 per order line.

Customer A generates 180 order lines, so its picking cost is:

C_A,pick = 180 × 4 = $720.

If the same customer also creates 12 special handling events at $15 each and 6 deliveries at $90 each, its identified activity cost is:

720 + (12 × 15) + (6 × 90) = $1,440.

ABC can reveal that a customer with high sales revenue is unprofitable because it places many small orders, requires special handling and demands frequent deliveries. The result should not automatically be to terminate the customer. Possible responses include minimum order quantities, delivery-day consolidation, revised pricing or redesigned packaging.

ABC has limitations. It requires reliable activity data, may be costly to maintain and can create false precision when driver relationships are weak. It is most valuable when cost-to-serve differences are material and management can act on the information.

3. Logistics Performance Measurement

A KPI is a quantified measure used to monitor an objective. A useful KPI has a clear definition, formula, data owner, frequency, target, time period and corrective-action process. Performance measurement should combine financial and non-financial indicators.

3.1 Important logistics KPIs

For a ratio, the numerator and denominator must refer to the same population and period. For example, an on-time rate based on shipment date should not be compared with a denominator based on invoice date.

Perfect order rate is especially demanding because one order must satisfy multiple conditions. If 940 of 1,000 orders are on time, 970 are complete, 990 are damage-free and 980 have correct documentation, it is incorrect to average these percentages to claim perfect-order performance. The organisation must identify the orders satisfying all conditions simultaneously.

3.2 Leading and lagging indicators

Lagging indicators describe results, such as logistics cost, complaints and late deliveries. Leading indicators help explain or predict results, such as forecast bias, preventive-maintenance completion, training compliance, cycle-count completion and dock congestion. A balanced system uses both. Measuring only cost can encourage behaviour that damages service; measuring only service can hide waste.

3.3 Balanced performance framework

A practical logistics scorecard may contain four perspectives:

• Financial: cost per shipment, cost-to-serve, working capital and claims cost.
• Customer: on-time delivery, perfect order rate, complaints and response time.
• Process: pick accuracy, order cycle time, inventory accuracy and vehicle utilisation.
• Capability and sustainability: employee skills, system availability, safety, energy use and emissions.

Targets must reflect strategy. A premium medical-supply distributor may rationally accept higher transport cost to achieve high reliability. A low-price commodity distributor may prioritise consolidation and asset utilisation.

4. Cost–Service Analysis

Logistics service is multidimensional. Customers may value availability, speed, reliability, flexibility, information and damage-free delivery. Higher service usually requires more inventory, capacity, technology or expedited transport, but the relationship is not always linear.

A simple decision model compares incremental cost with incremental contribution or customer value:

where ∆V is the change in net value, ∆B is incremental benefit, and ∆C is incremental logistics cost. A service improvement is financially attractive when ∆V > 0, subject to strategic, legal and operational constraints.

Benefits may include retained sales, fewer penalties, lower returns, improved customer lifetime value or reduced disruption risk. They should not be counted twice. For example, recovered sales revenue and the margin earned on those sales are not interchangeable measures.

A useful marginal approach compares the additional cost of service improvement with the additional gross margin or avoided loss. Service segmentation is often superior to one universal service policy. High-value or time-sensitive customers may receive priority, while standard customers receive consolidated service.

5. Break-Even Analysis

Break-even analysis identifies the activity level at which total revenue equals total cost. Under a simple linear model:

where:

• Q_BE = break-even quantity in units or shipments;
• F = fixed cost per period;
• p = revenue or price per unit; and
• v = variable cost per unit.

The contribution margin per unit is p - v. The model assumes constant price and variable cost, fixed cost within the relevant range, one product or a constant sales mix, and equal production and sales volumes.

Example. A delivery operation has fixed monthly cost of $48,000, revenue of $18 per shipment and variable cost of $10 per shipment:

Q_BE = 48,000 ÷ (18 - 10) = 6,000 shipments per month.

At 7,500 shipments, operating profit before other costs is:

(7,500 × 18) - [48,000 + (7,500 × 10)] = $12,000.

Break-even results are sensitive to assumptions. Overtime, congestion, tiered carrier pricing and capacity limits can make the cost curve non-linear. Managers should use scenario analysis rather than treating one break-even number as a forecast.

6. Make-or-Buy Decisions

A make-or-buy decision compares the relevant costs and strategic consequences of internal provision with outsourcing. The basic quantitative comparison is:

where F_m is avoidable internal fixed cost, v_m is internal variable cost per unit, Q is required volume, p_b is supplier price per unit and C_sw is switching, monitoring or contract-management cost.

If some internal fixed costs remain after outsourcing, they are not avoidable and should be excluded from the short-run comparison. Qualitative factors include control, resilience, confidentiality, labour capability, supplier dependency, regulatory compliance, service flexibility and reputational risk.

Example. Internal fulfilment has avoidable fixed cost of $30,000 per month and variable cost of $4.50 per order. A provider charges $6.20 per order plus $5,000 monthly contract-management cost. At 10,000 orders:

C_make = 30,000 + (4.50 × 10,000) = $75,000.

C_buy = 5,000 + (6.20 × 10,000) = $67,000.

Outsourcing has a short-run quantitative advantage of $8,000 per month. However, the decision is not complete until service-level agreements, data access, transition costs, capacity, contingency arrangements and supplier financial stability are assessed.

The indifference volume, when both options cost the same, is obtained by equating the cost functions:

This expression applies only when the denominator is non-zero and the compared cost structures remain valid over the relevant volume range. In the example, Q^* = (30,000 - 5,000) ÷ (6.20 - 4.50) ≈ 14,706 orders. Below this volume, buying is cheaper under the stated assumptions; above it, making is cheaper.

7. Logistics Dashboards and Data Analytics

A dashboard is a decision-support display that combines selected measures, trends, thresholds and explanations. It should not be a collection of attractive charts without decision logic.

7.1 Data pipeline

A reliable logistics analytics process includes:

1. Define the decision: for example, reduce late deliveries or improve warehouse productivity.
2. Specify measures: define numerator, denominator, unit, time window and exclusions.
3. Collect data: obtain transactions from transport-management, warehouse-management, enterprise-resource-planning and customer systems.
4. Validate data: check missing values, duplicate records, impossible timestamps and inconsistent identifiers.
5. Transform data: standardise units, join master data and calculate KPIs.
6. Analyse: examine trend, variance, segmentation, correlations and root causes.
7. Act and review: assign owners, implement actions and test whether performance changes.

A KPI data dictionary should state the measure name, business definition, formula, source system, refresh frequency, owner, target, tolerance and known limitations.

7.2 Descriptive, diagnostic and predictive analytics

• Descriptive analytics reports what happened, such as weekly delivery performance.
• Diagnostic analytics investigates why it happened, such as separating carrier delay, stock shortage and picking delay.
• Predictive analytics estimates what may happen, such as expected demand or late-delivery risk.
• Prescriptive analytics evaluates possible actions, such as alternative routes or inventory policies.

Correlation does not establish causation. A dashboard showing that late deliveries rise when volume rises does not prove volume is the cause; staffing, weather, carrier capacity or order mix may be confounding factors.

Basic descriptive statistics are often useful. For observations x₁, x₂, ..., x_n, the arithmetic mean is:

The mean is sensitive to extreme values. Median delivery time can better represent a skewed distribution in which a few severe delays occur. Percentiles are useful for service promises because the 95th percentile indicates a time at or below which approximately 95% of observations fall, subject to the data and estimation method used.

7.3 Benchmarking

Benchmarking compares performance with an internal unit, historical period, competitor, industry reference or best-in-class process. Valid comparisons require similar definitions, product characteristics, geography, service promise, demand pattern and accounting treatment. Benchmarking should stimulate learning, not encourage manipulation of denominators or unsafe cost cutting.

8. Applied Case: Northstar Home Appliances

This is an illustrative case. Northstar distributes household appliances through a regional warehouse. Management reports low transport cost per shipment, yet customer complaints and returns are increasing.

Background and problem

The company serves retail stores and online customers. Its dashboard contains transport cost, warehouse labour cost and total orders, but not cost-to-serve by customer. Several large customers place frequent small orders and request special delivery windows. Returns are coded as sales administration rather than logistics failures.

Analysis

A cross-functional team maps activities: receiving, storage, picking, special handling, dispatch, delivery and returns. ABC rates show that small orders consume disproportionate picking and delivery resources. The team also defines a perfect-order measure requiring on-time, complete, damage-free and correctly documented delivery. It discovers that the previous on-time KPI excluded orders awaiting stock replenishment, concealing an inventory-related service failure.

Monthly information indicates:

• fixed warehouse and supervision cost: $60,000;
• variable fulfilment cost: $5.00 per order;
• average revenue contribution: $14.00 per order; and
• proposed delivery-consolidation programme cost: $8,000 fixed plus $1.50 per order saved through route and handling efficiencies.

At 8,000 orders, the programme’s monthly cost is:

8,000 + (1.50 × 8,000) = $20,000.

If it prevents 1,600 failed or reworked deliveries at an avoidable cost of $14 each, the benefit is:

1,600 × 14 = $22,400.

The estimated net benefit is therefore $2,400 per month, before considering customer-retention benefits. This calculation assumes the avoided cost estimate is valid and that the programme does not reduce sales through unacceptable delivery delays.

Approach and solution

Northstar introduces delivery-day segmentation, minimum order quantities for selected accounts, a returns activity code and a dashboard with cost-to-serve, perfect-order rate, fill rate, inventory accuracy and returns rate. It pilots the changes with two customer groups. The pilot includes a control group so that changes can be compared rather than attributed automatically to the programme.

Outcome and learning points

The pilot identifies that a small increase in planned consolidation reduces failed deliveries without materially reducing customer satisfaction. The company keeps premium delivery for urgent orders but prices it separately. It also improves inventory-cycle counting for products associated with stock-related cancellations.

The case illustrates five lessons: total cost is more informative than one cost ratio; KPI definitions influence behaviour; ABC supports customer and service segmentation; pilots reduce implementation risk; and cost improvements must be checked against customer outcomes.

9. Technology, Governance and Ethical Issues

Modern logistics analytics may use barcode and RFID data, GPS, cloud platforms, optimisation software, machine learning and generative AI assistants. These tools can improve visibility and scenario analysis, but technology does not correct poor definitions or inaccurate master data.

AI-based forecasts and exception systems require human oversight. Managers should test model accuracy across products, locations and customer groups, monitor drift and document how recommendations are used. Automated decisions affecting employees, suppliers or customers should be explainable and reviewable.

Important governance practices include:

• access controls based on job responsibility;
• audit trails for changes to costs, orders and KPI definitions;
• validation of supplier and customer master data;
• retention and deletion rules for personal and location data;
• protection of commercially sensitive rates and volumes;
• compliance with applicable privacy, transport, labour, safety and financial-reporting requirements; and
• controls against KPI manipulation, such as excluding late orders or redefining the denominator.

Ethical measurement also requires attention to worker safety and fairness. Productivity targets should not encourage unsafe lifting, excessive driving hours or concealment of incidents. Sustainability metrics should state their boundaries, such as whether emissions include only transport fuel or also warehousing energy and outsourced operations.

10. Implementation Best Practices and Limitations

An effective logistics performance programme should:

1. begin with strategic and customer requirements;
2. use a small set of clearly defined KPIs;
3. assign each KPI an accountable owner;
4. combine cost, service, quality, safety and sustainability measures;
5. distinguish controllable from uncontrollable variation;
6. investigate root causes rather than blame individuals;
7. use ABC where cost-to-serve variation justifies the effort;
8. review targets when network design, product mix or service promises change; and
9. close the loop through documented corrective actions.

No model is perfect. Accounting allocations may not reflect economic causality. Average costs can hide peak-capacity effects. KPIs can create gaming and local optimisation. Forecasts are uncertain, and benchmarks may be incomparable. Managers should therefore present assumptions, confidence or sensitivity where appropriate and combine quantitative evidence with operational judgement.

Chapter Summary

Logistics costing measures the resources consumed by transport, warehousing, inventory, order processing, packaging, quality and returns. Total-cost thinking prevents local decisions from transferring cost elsewhere in the supply chain. Activity-based costing assigns resource costs through activities and causal drivers, helping organisations understand cost-to-serve by order, product or customer.

Performance management requires clearly defined financial, customer, process, capability, safety and sustainability KPIs. Balanced dashboards should combine leading and lagging indicators and support decisions rather than merely display data. Break-even analysis estimates the volume at which contribution covers fixed cost, while make-or-buy analysis compares avoidable internal costs with supplier costs and strategic risks. Cost–service analysis evaluates whether incremental service benefits exceed incremental cost.

Analytics depends on sound definitions, clean data, appropriate comparisons and responsible governance. Descriptive, diagnostic, predictive and prescriptive methods can improve logistics decisions, but correlation is not causation and technology cannot replace managerial judgement. Effective implementation links measures to owners, corrective action, customer value, safety and ethical conduct.

Key Terms

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

Benchmarking: Comparing performance with a reference organisation, process, period or standard.

Break-even point: The activity level at which total revenue equals total cost.

Carrying cost: Cost of holding inventory, including capital, storage, insurance and obsolescence components.

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

Cost-to-serve: The logistics cost required to serve a particular customer, product, order or channel.

Dashboard: A visual decision-support display of selected performance measures.

Fixed cost: A cost that remains broadly constant within a relevant capacity range.

Fill rate: The proportion of demand supplied immediately from available stock.

Indifference volume: The volume at which two alternatives have equal relevant cost.

Inventory accuracy: Agreement between recorded and physically available inventory.

KPI: A key performance indicator used to monitor an important objective.

Lagging indicator: A measure of an outcome that has already occurred.

Leading indicator: A measure that may signal or influence future performance.

Make-or-buy analysis: Comparison of internal provision and external procurement alternatives.

Perfect order: An order delivered on time, complete, damage-free and correctly documented.

Relevant cost: A future cost that changes as a result of a decision.

Service level: The agreed or achieved standard of logistics service provided to a customer.

Variable cost: A cost that changes with activity volume.

Review Questions

Short Answer Questions

1. What is meant by total logistics cost, and why is it preferable to examining transport cost alone?
2. Distinguish fixed, variable, direct and indirect logistics costs.
3. What is a cost driver in activity-based costing? Give two logistics examples.
4. Why should a logistics dashboard include both leading and lagging indicators?
5. State two limitations of benchmarking logistics performance.

Long Answer Questions

1. Explain the steps in implementing activity-based costing for a warehouse and discuss its benefits and limitations.
2. Discuss how a balanced logistics scorecard can prevent excessive focus on cost reduction.
3. Explain the assumptions of break-even analysis and why real logistics cost curves may be non-linear.
4. Describe a structured make-or-buy decision, including quantitative and qualitative considerations.
5. Explain how data governance and ethical controls influence the reliability of logistics analytics.

Analytical / Application Questions

1. A warehouse has fixed monthly cost of $36,000, revenue of $16 per order and variable cost of $10 per order. Calculate the break-even number of orders and the operating result at 7,000 orders. State the assumptions used.
2. An internal operation costs $22,000 avoidable fixed cost plus $4.80 per shipment. An external provider charges $6.10 per shipment plus $3,000 monthly administration cost. Determine the indifference volume and recommend how the decision should be evaluated beyond cost.
3. A customer has a high annual sales value but places many small orders, requests urgent deliveries and returns damaged products. Design an ABC cost-to-serve analysis and recommend at least three possible management actions while protecting customer service.

Multiple Choice Questions with Answers

Question 1

Which costing method assigns resource costs first to activities and then to products, orders or customers?

A. Activity-based costing

B. Absorption by sales value only

C. Cash-flow accounting

D. Standard delivery costing

Correct Answer: A. Activity-based costing

Question 2

A logistics operation has fixed cost of $48,000 per month, revenue of $18 per shipment and variable cost of $10 per shipment. What is the break-even volume?

A. 4,800 shipments

B. 6,000 shipments

C. 8,000 shipments

D. 10,800 shipments

Correct Answer: B. 6,000 shipments

Question 3

Which is a leading logistics indicator?

A. Last month’s claims cost

B. Completed preventive-maintenance activities

C. Annual logistics cost

D. Previous quarter’s return rate

Correct Answer: B. Completed preventive-maintenance activities

Question 4

Average inventory is 4,000 units, unit value is $25 and the annual carrying rate is 22%. What is the estimated annual inventory-carrying cost?

A. $5,500

B. $18,000

C. $22,000

D. $100,000

Correct Answer: C. $22,000

Question 5

Which cost should generally be excluded from a short-run make-or-buy comparison if it will remain after outsourcing?

A. Avoidable labour cost

B. Supplier purchase price

C. Variable internal handling cost

D. Unavoidable fixed cost

Correct Answer: D. Unavoidable fixed cost

Question 6

What condition is required for a perfect-order rate?

A. The order must have the highest sales value

B. The order must be delivered by the cheapest carrier

C. The order must satisfy all specified service conditions simultaneously

D. The order must contain only one product

Correct Answer: C. The order must satisfy all specified service conditions simultaneously

Question 7

A warehouse picking activity costs $120,000 and has 30,000 order lines as its driver. What is the activity rate?

A. $0.25 per order line

B. $2 per order line

C. $4 per order line

D. $25 per order line

Correct Answer: C. $4 per order line

Question 8

Which statement about correlation in logistics analytics is correct?

A. Correlation always proves causation

B. Correlation can indicate association but does not by itself prove causation

C. Correlation is only used for financial data

D. Correlation eliminates the need for data validation

Correct Answer: B. Correlation can indicate association but does not by itself prove causation

Question 9

At 8,000 orders, a consolidation programme costs $8,000 fixed plus $1.50 per order. If it avoids 1,600 failed deliveries costing $14 each, what is the estimated net benefit?

A. $400

B. $2,400

C. $14,400

D. $22,400

Correct Answer: B. $2,400

Question 10

Which practice best supports ethical logistics performance management?

A. Changing KPI definitions whenever targets are missed

B. Excluding delayed orders from the denominator

C. Using access controls, audit trails and documented KPI definitions

D. Prioritising productivity targets over worker safety

Correct Answer: C. Using access controls, audit trails and documented KPI definitions