Logistics Costing, Performance Analytics and Decision-Making

Introduction

Logistics decisions convert resources into customer service. Transport, warehousing, inventory, order processing, packaging, information systems and returns all consume resources, but their costs are not always visible in a general ledger. A shipment may appear inexpensive when only the carrier invoice is considered, yet become costly after customs duties, port storage, handling, inventory financing, damage and administrative activity are included. Effective logistics management therefore requires more than reducing individual costs. It requires understanding the total cost of serving a customer at a defined service level.

This chapter presents practical methods for analysing logistics costs and performance. It covers activity-based costing (ABC), total landed cost, cost-to-serve, break-even analysis, logistics trade-offs and KPI dashboard design. The central managerial principle is that cost and service must be analysed together. A low-cost option that causes stockouts, late deliveries or excessive returns may destroy value; a premium option may be justified for strategically important products or customers.

Learning objectives

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

1. Distinguish direct, indirect, fixed, variable and relevant logistics costs.
2. Apply activity-based costing to logistics processes.
3. Calculate total landed cost and customer cost-to-serve.
4. Use contribution margin and break-even analysis in logistics decisions.
5. Analyse cost–service and inventory–transport trade-offs.
6. Calculate and interpret OTIF, fill rate, inventory turns, dwell time, utilisation, accuracy and logistics cost per unit.
7. Design a management dashboard and use data critically in operational decisions.

1. The logic of logistics costing

A logistics cost system should answer four questions:

• What resources are consumed?
• Which activities consume those resources?
• Which products, orders, customers or channels cause the activities?
• What service and financial outcomes result?

A useful cost classification is shown below.

The distinction between accounting cost and decision-relevant cost is essential. For example, depreciation may be included in product profitability reporting, but it is irrelevant to a short-term routing decision if the asset will be retained under either option. Conversely, overtime, congestion and lost sales may be absent from a simple ledger but highly relevant to the decision.

A cost object is the unit being analysed: a product, order, customer, route, warehouse, channel or shipment. Cost allocation is useful only when the allocation basis reflects causation. Allocating all warehouse cost by sales value can penalise high-value, low-touch products and subsidise low-value products requiring many picks and returns.

2. Activity-based costing in logistics

Activity-based costing assigns resource costs to activities and then assigns activity costs to cost objects using cost drivers. The method is especially useful where customers and products consume logistics activities in different proportions.

The basic relationship is:

The activity rate has units such as currency per pick, currency per pallet receipt or currency per delivery. The cost assigned to a cost object is:

The approach assumes that the selected driver is a reasonable representation of resource consumption and that the activity pool is sufficiently homogeneous. A warehouse may therefore use separate pools for receiving, put-away, picking, packing, dispatch and returns rather than one undifferentiated pool.

Worked example: ABC for customer orders

A distribution centre has the following monthly activity pools:

The activity rates are:

• Receiving: $18,000 ÷ 600 = $30 per receipt.
• Picking: $42,000 ÷ 21,000 = $2 per order line.
• Packing: $24,000 ÷ 12,000 = $2 per order.
• Dispatch: $16,000 ÷ 8,000 = $2 per shipment.

Customer A generates 8 receipts, 420 order lines, 180 orders and 150 shipments. Its assigned cost is:

If Customer A produces $3,200 contribution before distribution cost, its contribution after the analysed activities is $1,240. The result is more informative than dividing the centre’s total cost by sales revenue, because it identifies the operational reasons for the cost.

ABC should not be interpreted as mathematically exact. It is a decision-support model. Excessively detailed pools increase data-collection cost and may create false precision. Drivers should be reviewed when processes, technology, order profiles or service agreements change.

3. Total landed cost

Total landed cost estimates the complete cost of acquiring and making goods available at the intended destination. It is particularly important in sourcing, international purchasing and network design.

A general model is:

where:

• TLC = total landed cost;
• P = purchase price;
• F = international and domestic freight;
• I = insurance;
• D = duties and tariffs;
• C = customs brokerage and compliance charges;
• H = handling, port and terminal charges;
• S = storage and demurrage before availability;
• Q = quality inspection, testing or rework;
• R = other destination costs required to make the product saleable.

The terms must be defined consistently. Taxes that are recoverable may be excluded from economic cost, while non-recoverable taxes should be included. Currency conversion should use an identified exchange rate and date. The model should also specify whether costs are per unit, shipment, container or purchase order.

Worked example: sourcing comparison

A buyer compares two suppliers for 1,000 units.

Supplier X total landed cost per unit is $54.90. Supplier Y is $56.35 per unit. Although Y has the lower purchase price, X is cheaper by $1.45 per unit, or $1,450 for the shipment. If Y offers materially shorter lead time, the decision should additionally include inventory carrying cost, stockout risk and the value of responsiveness.

A landed-cost model must not automatically include speculative costs. Expected disruption costs can be modelled separately through scenarios or probabilities rather than inserted as arbitrary charges.

4. Cost-to-serve and customer profitability

Cost-to-serve (CTS) measures the resources required to provide a defined service to a particular customer, segment, product or channel. It links commercial revenue with operational behaviour.

A simplified customer contribution model is:

Cost-to-serve may include order entry, credit administration, picking, special packaging, delivery, customer-specific labelling, returns, claims, expedited transport and account management. Discounts and rebates should be deducted from revenue where the purpose is customer profitability analysis.

A customer with high revenue may be unattractive if it places many small orders, requires urgent deliveries, returns a high proportion of goods and pays slowly. The appropriate response is not necessarily termination. Managers may redesign the service: minimum order quantities, delivery-day consolidation, revised packaging, electronic ordering or differentiated pricing.

A robust CTS analysis follows these steps:

1. Define the profitability unit and analysis period.
2. Map the customer order-to-cash process.
3. Identify activities and cost drivers.
4. Measure customer-specific driver consumption.
5. Assign costs using transparent rates.
6. Compare contribution across customer segments.
7. Test service redesign and pricing alternatives.

The analysis should separate controllable operational cost from commercial decisions and should avoid using profitability data to discriminate unfairly between customers without a legitimate service or contractual basis.

5. Break-even and trade-off analysis

Break-even analysis supports decisions such as outsourcing transport, opening a warehouse, purchasing automation or offering a premium delivery service. The standard model assumes a constant selling price, constant variable cost per unit and fixed cost within a relevant range.

The denominator is the contribution per unit. If fixed cost is $120,000, revenue per delivery is $18 and variable cost is $10, then:

At volumes below 15,000, the option does not cover its fixed cost under the assumptions. At volumes above it, each additional delivery contributes $8 before other fixed costs. The result is valid only within the capacity and price range assumed. Step-fixed costs, overtime, service penalties and multiple products can invalidate a simple one-product calculation.

For a decision between alternatives, use incremental analysis:

Include opportunity costs and benefits, but exclude costs unchanged by the decision. For example, if an internal fleet costs $200,000 fixed plus $4 per delivery and an external carrier costs $9 per delivery, outsourcing is cheaper below the volume at which the fixed internal cost is spread sufficiently. Set the costs equal:

Thus Q = 40,000 deliveries. At 30,000 deliveries, internal cost is $320,000 and external cost is $270,000, so outsourcing has a $50,000 cost advantage, assuming equivalent service and no relevant transition costs.

Trade-offs

Logistics systems contain interdependent trade-offs:

• Larger inventories can improve fill rate but increase capital, obsolescence and storage cost.
• Faster transport can reduce pipeline inventory and stockout risk but increase freight cost.
• Fewer warehouses may reduce fixed facility cost but increase delivery distance and response time.
• Full truckloads may reduce unit transport cost but increase cycle stock and handling risk.
• Automation may reduce variable labour cost but require capital, maintenance and specialised skills.

A total-cost model should include service consequences. A useful decision table compares alternatives across cost, speed, reliability, capacity, risk and strategic fit rather than selecting the lowest visible price.

6. Logistics performance analytics

A KPI is useful when it has a clear definition, owner, data source, frequency, target and corrective-action rule. Metrics should be balanced: cost-only systems encourage service deterioration, while service-only systems can conceal waste.

Core KPI definitions

For inventory turns, average inventory is often calculated as opening inventory plus closing inventory divided by two, although a periodic average is preferable where inventory is seasonal. The numerator should be cost of goods sold, not sales revenue, because both numerator and denominator are measured at cost.

OTIF must specify the promised date, shipment-level or order-level measurement, tolerance window and treatment of customer-caused delays. A shipment delivered on time but short is not OTIF. Fill rate may be measured by units, order lines or orders; these measures are not interchangeable.

Dwell time is a duration and must use consistent timestamps and time zones. Utilisation above 100% often indicates that nominal capacity is understated, temporary overflow is being used or the denominator is incorrectly defined. High utilisation can also reduce resilience and increase congestion.

7. Dashboard design and analytical governance

A dashboard should support decisions, not merely display numbers. A practical structure contains four layers:

1. Outcome layer: OTIF, customer complaints, revenue at risk and logistics cost.
2. Process layer: picking accuracy, dwell time, cycle time and dock productivity.
3. Resource layer: labour utilisation, vehicle fill, warehouse capacity and inventory.
4. Diagnostic layer: site, route, product, customer, shift and exception details.

Figure: A dashboard links outcomes to process, resource and diagnostic measures.

Dashboard governance requires a metric dictionary. Each KPI should state its numerator, denominator, exclusions, source system, refresh time, owner and target. Data quality checks should identify missing timestamps, duplicate orders, changed master data and inconsistent units. A red indicator is not an explanation; it is a prompt for investigation.

Useful analytical methods include trend analysis, Pareto analysis, segmentation, control charts and root-cause analysis. Correlation does not prove causation. For example, a relationship between high warehouse utilisation and late orders may reflect a third factor such as seasonal demand. Analysts should compare like-for-like periods and control for product mix, promotions and exceptional events.

AI and advanced analytics can forecast demand, detect anomalous dwell times, optimise routes and classify customer complaints. However, models require reliable historical data, monitoring for drift and human review. Automated recommendations should not override safety, legal requirements, contractual commitments or explainability obligations. Personal data in customer, driver or employee records must be minimised, protected and used for legitimate purposes.

8. Applied case: Northstar Home Distribution

Background

Northstar Home Distribution is an illustrative case. It imports kitchen equipment and supplies retailers from one regional distribution centre. Management reports strong sales growth but declining operating margin. Retailers complain about partial deliveries, while the warehouse reports rising overtime.

Problem

The company’s existing report allocates warehouse cost by sales value. Large retailers appear profitable, even though some place many small orders and request special labels. Management must decide whether to redesign service rules and whether a second carrier is justified.

Analysis

An ABC review identifies monthly costs and drivers as follows:

Retailer R generates 900 orders, 5,400 lines, 600 labels, 70 returns and 1,100 shipments per month. Its assigned activity cost is:

The analysis shows that orders, labelling and shipments—not merely sales value—drive cost. Northstar also measures 91% OTIF, 94% fill rate, 5.2 inventory turns, 14-hour average dock dwell, 78% warehouse utilisation and 97.1% inventory accuracy. The combination of low OTIF and high, though not extreme, utilisation suggests process instability rather than a simple lack of space.

A Pareto review finds that 20% of retail accounts create 64% of special labels and 58% of returns. Transport data show that urgent deliveries are concentrated on two delivery days.

Approach and solution

Northstar introduces consolidated delivery windows, a minimum order value for free delivery, electronic order templates and a charge for customer-specific labelling. It reserves premium carrier capacity for orders with a documented service requirement. The warehouse redesigns the dispatch cut-off and creates a daily exception review.

Targets are set at 96% OTIF, 98% inventory accuracy, less than 10 hours average dwell and logistics cost below $6.50 per shipped unit. The company does not set a target to maximise utilisation; it maintains a capacity buffer for promotions and disruption.

Outcome

After three months, the illustrative results are 96.4% OTIF, 97.8% fill rate, 10.5-hour dwell, 98.4% inventory accuracy and a 9% reduction in logistics cost per unit. The case demonstrates that profitability improvement came from redesigning demand and service behaviour, not from indiscriminate budget cuts.

Learning points

• Revenue is not a substitute for customer profitability.
• Service policies influence operational cost.
• KPIs must be interpreted as a system.
• A capacity buffer can improve resilience even when it lowers apparent utilisation.
• Cost transparency supports constructive customer and internal negotiations.

9. Limitations, ethics and implementation risks

Cost models can create misleading certainty. Allocated costs depend on driver selection, data quality and scope. A driver may be statistically associated with cost without being causally responsible. Managers should test sensitivity: if a rate changes by 10%, does the decision change?

Performance targets can also produce dysfunctional behaviour. A warehouse may improve picking speed by sacrificing accuracy; a transport team may protect OTIF by using expensive emergency freight; a buyer may minimise purchase price while increasing landed cost. Balanced scorecards, audit trails and exception review reduce these risks.

Ethical governance includes transparent measurement rules, fair treatment of employees, privacy protection, non-discriminatory customer segmentation and accurate reporting. Safety incidents must never be hidden to protect KPI performance. Regulatory obligations may include customs compliance, tax documentation, dangerous-goods controls, driver-hour rules and financial record retention, depending on the jurisdiction and product.

Implementation is best conducted incrementally: define the decision, map the process, establish a baseline, pilot a small number of drivers and KPIs, validate results with operational staff, then scale. A sophisticated dashboard built on unreliable master data is less valuable than a simple, trusted report.

Best-practice checklist

• Define the cost object and decision before collecting data.
• Use practical capacity rather than theoretical maximum where appropriate.
• Separate fixed, variable, avoidable and opportunity costs.
• Reconcile operational data to financial totals.
• Use activity drivers that reflect causation.
• State KPI definitions and exclusions explicitly.
• Combine cost, service, quality, capacity and resilience measures.
• Investigate trends and exceptions rather than reacting to one number.
• Test scenarios and sensitivity before committing capital.
• Assign owners, review dates and corrective actions to every major KPI.

Chapter Summary

Logistics costing is a managerial system for understanding how resources are consumed and how service is created. Activity-based costing assigns resource costs through activities and cost drivers, producing more meaningful product, order and customer profitability information than broad allocations. Total landed cost extends purchase price to include freight, duties, handling, compliance and other costs required to make goods available. Cost-to-serve links operational effort to customer contribution.

Break-even analysis identifies the volume at which contribution covers fixed cost, but its assumptions and relevant capacity range must be respected. Logistics decisions require trade-off analysis because transport, inventory, warehousing, service and resilience are interdependent. KPIs such as OTIF, fill rate, inventory turns, dwell time, utilisation, accuracy and logistics cost per unit should be defined precisely and interpreted together. Effective dashboards connect outcomes to processes, resources and diagnostic detail. Finally, sound analytics depends on data governance, ethical use, sensitivity testing and human judgement.

Key Terms

Activity-based costing: Costing method that assigns resource costs through activities and cost drivers.

Cost object: Product, order, customer, route or other unit for which cost is measured.

Cost driver: Measurable factor that causes or explains activity cost.

Cost-to-serve: Cost of providing a defined logistics service to a customer or segment.

Total landed cost: Complete cost of acquiring and making goods available at destination.

Contribution margin: Revenue remaining after variable costs.

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

Relevant cost: Future cost that differs between alternatives.

Opportunity cost: Benefit forgone by selecting one alternative.

OTIF: On-time, in-full delivery performance.

Fill rate: Proportion of ordered quantity supplied immediately.

Inventory turns: Cost of goods sold divided by average inventory at cost.

Dwell time: Time an asset or shipment remains in a facility or process.

Capacity utilisation: Used practical capacity expressed as a proportion of available capacity.

Inventory accuracy: Agreement between recorded and physically verified stock.

Logistics cost per unit: Total logistics cost divided by units shipped.

Practical capacity: Achievable capacity after normal downtime and operating constraints.

Metric dictionary: Controlled record defining KPI meaning, source, owner and calculation.

Review Questions

Short Answer Questions

1. Distinguish direct cost from indirect cost in a warehouse.
2. What is a cost driver, and why is driver selection important in ABC?
3. Define total landed cost.
4. Why must inventory turns use cost of goods sold rather than sales revenue?
5. State two limitations of logistics dashboards.

Long Answer Questions

1. Explain the process of implementing activity-based costing in a distribution centre.
2. Discuss how cost-to-serve can change customer segmentation and service policy.
3. Explain the assumptions and limitations of break-even analysis in logistics.
4. Discuss the major transport, inventory and warehousing trade-offs faced by logistics managers.
5. Design a balanced logistics dashboard for a regional distribution centre and justify the KPIs selected.

Analytical / Application Questions

1. A warehouse has fixed cost of $150,000, revenue of $25 per shipment and variable cost of $15 per shipment. Calculate break-even volume and explain one assumption that could make the result unreliable.
2. A customer creates 300 orders, 1,200 order lines, 80 returns and 350 shipments. Activity rates are $4 per order, $1.50 per line, $18 per return and $5 per shipment. Calculate the customer’s assigned cost and identify the activity with the greatest cost.
3. A site reports 99% utilisation, 89% OTIF and 96% inventory accuracy. Analyse two possible causes and recommend three additional data points needed before approving warehouse expansion.

Multiple Choice Questions with Answers

Question 1

Which cost is most directly relevant to a short-term decision when it changes between alternatives?

A. Sunk cost

B. Relevant incremental cost

C. Historical depreciation

D. Allocated corporate rent

Correct Answer: B. Relevant incremental cost

Question 2

An activity costs $30,000 and has a practical driver volume of 6,000 orders. What is the activity rate?

A. $0.20 per order

B. $5 per order

C. $20 per order

D. $180 per order

Correct Answer: B. $5 per order

Question 3

Which item is normally included in total landed cost?

A. Unrelated advertising expense

B. Future shareholder dividends

C. Customs duty required to import the goods

D. A sunk cost from a previous warehouse

Correct Answer: C. Customs duty required to import the goods

Question 4

A company has fixed cost of $120,000, revenue per delivery of $18 and variable cost of $10. What is break-even volume?

A. 8,000 deliveries

B. 12,000 deliveries

C. 15,000 deliveries

D. 21,600 deliveries

Correct Answer: C. 15,000 deliveries

Question 5

If annual cost of goods sold is $2,000,000 and average inventory at cost is $500,000, inventory turns equal:

A. 0.25

B. 2

C. 4

D. 10

Correct Answer: C. 4

Question 6

OTIF measures whether an order was:

A. Cheap and rapidly picked

B. Delivered on time and in full

C. Stored at maximum utilisation

D. Returned without damage

Correct Answer: B. Delivered on time and in full

Question 7

Which denominator is most appropriate for logistics cost per unit?

A. Units shipped, consistently defined

B. Number of employees only

C. Total sales value regardless of volume

D. Number of suppliers

Correct Answer: A. Units shipped, consistently defined

Question 8

A customer requires many small orders and special labels. Which method best reveals the resulting operational cost?

A. Sales-value allocation alone

B. Activity-based costing

C. Ignoring indirect cost

D. Using purchase price only

Correct Answer: B. Activity-based costing

Question 9

Very high warehouse utilisation may be undesirable because it can:

A. Eliminate all inventory cost

B. Increase congestion and reduce resilience

C. Guarantee higher OTIF

D. Make capacity unlimited

Correct Answer: B. Increase congestion and reduce resilience

Question 10

Which dashboard practice is strongest?

A. Changing KPI definitions each month

B. Using targets without data owners

C. Defining numerator, denominator, source and owner

D. Reporting only cost measures

Correct Answer: C. Defining numerator, denominator, source and owner