Cost-to-Serve, Customer Profitability and Pricing

Introduction

Revenue is not the same as profitability. Two customers with identical sales value may generate very different costs because one places frequent small orders, requires urgent delivery, returns products, demands special packaging, or operates in a difficult-to-reach location. Logistics costing therefore has to move beyond average cost per unit and identify the resources consumed by particular customers, products, channels, routes and orders.

Cost-to-serve (CTS) analysis provides this visibility. It links operational activities—order processing, warehousing, inventory holding, transport, delivery, returns and customer support—to the objects that cause them. When combined with contribution-margin analysis, customer lifetime value and pricing, CTS becomes a strategic instrument for service differentiation, contract negotiation and network design.

The central managerial question is not simply, “How much revenue does this customer generate?” It is: “After all relevant commercial and logistics costs, what economic value does this relationship create, and what service-price arrangement is appropriate?”

Learning Objectives

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

1. Explain the purpose, logic and limitations of cost-to-serve analysis.
2. Design CTS models by customer, product, channel, route and order.
3. Calculate contribution margin, customer profitability and customer lifetime value.
4. Distinguish cost allocation, cost causation and decision-relevant cost.
5. Use CTS evidence in differentiated service, contract and logistics-pricing decisions.
6. Evaluate governance, data, behavioural and ethical risks in customer profitability systems.

1. The Strategic Logic of Cost-to-Serve

Traditional financial accounting usually reports costs by function or cost centre. This is appropriate for statutory reporting and control, but it can conceal the economics of individual relationships. A warehouse may appear efficient in aggregate while serving a small group of customers whose special handling and fragmented orders consume disproportionate resources.

CTS assigns resource consumption to a cost object. A cost object may be:

• a customer or customer segment;
• a product or stock-keeping unit (SKU);
• a sales channel, such as retail, wholesale or e-commerce;
• a route, lane or delivery zone;
• an order, shipment or delivery stop; or
• a contract, project or service package.

The relevant distinction is between cost incidence and cost causation. Cost incidence asks where a cost is reported; cost causation asks which activity or decision generated it. A customer’s order may cause picking, packing, documentation, delivery and return costs even though those costs are recorded in several departments.

CTS is strategically important because it supports four linked decisions:

1. Service design: Which service levels should be offered to which customers?
2. Commercial design: Which terms, minimum order quantities and delivery conditions are economically sustainable?
3. Operational design: Which customers, products and routes should be consolidated, automated or redesigned?
4. Pricing: What price or surcharge is required to recover the cost of differentiated service?

A CTS result is not automatically a price recommendation. It is evidence for managerial judgement. Prices also depend on willingness to pay, competition, strategic importance, capacity constraints, switching costs and contractual commitments.

2. Building a Cost-to-Serve Model

A credible model follows the physical and information flows of the business rather than beginning with an arbitrary percentage allocation.

2.1 Define scope and cost objects

The analyst first specifies the decision period, geography, service scope and level of detail. A model for annual contract negotiation may use customer-month data; a model for route redesign may require order-stop data. Scope should include all material activities affected by the decision, while avoiding unnecessary complexity.

Typical cost objects and analytical questions are shown below.

2.2 Classify costs

Costs should be separated according to their decision relevance and behaviour:

• Direct costs can be traced economically to a cost object, such as a dedicated delivery or customer-specific packaging.
• Indirect costs support several objects and require a defensible allocation or attribution method.
• Variable costs change with activity over the relevant range, such as carrier charges per shipment.
• Fixed costs do not change with short-run volume, such as a committed warehouse lease.
• Avoidable costs would disappear if the relationship or activity were removed.
• Unavoidable costs remain under the decision being considered.

The distinction between avoidable and unavoidable cost is essential. A customer may appear unprofitable after allocation of a warehouse lease, but discontinuing that customer may not reduce the lease cost. For short-run acceptance decisions, the relevant measure may be contribution after avoidable fulfilment costs; for long-run contract design, capacity and fixed costs may become avoidable.

2.3 Activity-based costing logic

Activity-based costing (ABC) is frequently used in CTS because it connects resources to activities and activities to cost objects. The basic sequence is:

Resources → activities → cost pools → cost drivers → customers/products/orders.

For activity pool j, the activity rate is:

The rate has units such as currency per order, currency per pallet handled or currency per delivery stop. The cost assigned to a cost object i is:

Here, j indexes activities, i indexes cost objects, and Driver quantity_ij is the amount of activity j consumed by object i. The formula assumes that the selected driver is a reasonable proxy for resource consumption and that the rate is applied within the relevant capacity range.

Using practical capacity rather than actual volume can expose unused capacity. If a warehouse incurs £120,000 of receiving and put-away cost and has practical capacity of 6,000 pallets, the rate is £20 per pallet. If only 5,000 pallets are processed, allocating the entire £120,000 to those pallets would produce £24 per pallet and obscure the £20,000 cost of unused capacity.

2.4 Selecting cost drivers

A driver should satisfy three tests:

1. Causal validity: activity consumption changes predictably with the driver.
2. Measurability: reliable data can be collected at reasonable cost.
3. Managerial usefulness: the driver reveals an action that can be changed.

Examples include order lines for picking, pallets for handling, occupied pallet-days for storage, kilometres and stops for delivery, and return units for reverse logistics. Distance alone is often an inadequate transport driver because a short route with many stops may consume more labour than a long route with one drop.

2.5 Avoiding double counting

A model must distinguish costs already embedded in a carrier invoice from internal costs added separately. For example, if a third-party carrier charge includes fuel and delivery labour, these should not be added again unless they are internal incremental costs. Similarly, inventory carrying cost should not be counted twice through both a full warehouse rate and a separate storage charge.

3. Core Profitability Measures

3.1 Contribution margin

Contribution margin measures the amount remaining after costs that vary with the decision or volume under consideration:

Contribution margin ratio is:

Net revenue should deduct discounts, rebates, credits and returns. The cost boundary must be stated explicitly; “contribution” without a defined boundary is ambiguous.

3.2 Customer profitability

A practical customer profitability measure is:

where i denotes customer i. Customer-specific commercial costs may include commissions, account management, demonstrations, claims administration and marketing support. The model may report profit before or after allocated corporate overhead, but the boundary must be consistent.

A customer profitability ratio is:

This ratio supports comparison, but absolute profit is also important. A high-margin small account may contribute less total profit than a moderate-margin strategic account.

3.3 Cost-to-serve waterfall

A waterfall makes the movement from sales to profit visible:

Gross sales → discounts and rebates → net revenue → product cost → logistics CTS → commercial service cost → customer profit.

The waterfall should distinguish controllable causes from accounting allocations. A negative result can arise from excessive discounts, costly service, low order density, poor product mix, returns or a combination of these factors.

3.4 Customer lifetime value

Customer lifetime value (CLV) estimates the present value of expected future customer contribution. A simple discounted model is:

Here, t is the future period, T is the planning horizon, Revenue_t is expected revenue, the cost terms are expected period costs, p_t is the probability the customer remains active in period t, r is the discount rate, and Acquisition cost is the initial acquisition expenditure. Values must be expressed in the same currency and time units.

A steady-state simplified form can be used only when periodic contribution, retention probability and discount rate are stable and the horizon is effectively indefinite:

where m is periodic contribution and p is periodic retention probability. This expression assumes independent period retention and constant expected contribution; it is unsuitable where customer behaviour, prices or service costs change materially.

CLV is not a licence to discriminate arbitrarily. It is a planning estimate subject to uncertainty, data bias and assumptions about retention.

4. Worked Example: Customer-Level CTS

An illustrative distributor supplies Customer A during one month. The following information is available:

Net revenue is £50,000 − £3,000 = £47,000. Total CTS and customer-specific service cost is £720 + £1,500 + £1,800 + £900 + £480 + £700 = £6,100. Customer profit is therefore:

£47,000 − £30,000 − £6,100 = £10,900.

The customer profitability ratio is:

£10,900 ÷ £47,000 × 100 = 23.2% approximately.

Suppose the customer requests twice-weekly emergency delivery, adding 20 stops at £45 and 200 picking lines at £2.50, while generating no additional revenue. Incremental cost is £900 + £500 = £1,400. Profit falls to £9,500. Management could instead offer consolidated weekly delivery, retain the original service, or quote an emergency-delivery surcharge of at least £1,400 if the customer requires cost recovery. Whether the surcharge should be higher depends on capacity scarcity, willingness to pay and strategic policy.

The example illustrates why revenue ranking can be misleading: the service pattern, not only the sales value, determines economic value.

5. Profitability by Product, Channel, Route and Order

Product and SKU profitability

Product-level analysis should include handling characteristics, storage duration, damage, expiry, special packaging and return probability. A low-margin product may still be attractive if it moves in full pallets with high order density; a high-price product may be unattractive if it requires individual picking and controlled storage.

Channel profitability

Channels have different cost architectures. E-commerce may avoid some store-delivery costs but incur parcel handling, payment fees, home delivery and failed-delivery costs. Wholesale may have fewer orders but larger shipments. Channel comparison should use comparable customer and product mixes rather than attributing all differences to channel performance.

Route and lane profitability

Route economics depend on vehicle capacity, distance, travel time, stops, waiting, tolls, empty running and backhaul. A route may be profitable on a marginal basis while failing to cover long-run fleet and depot costs. Conversely, a route with low accounting profit may preserve strategic network coverage or enable profitable backhaul.

A route contribution measure can be stated as:

For a long-run decision, include avoidable fixed fleet, depot and supervisory costs. Do not allocate all network overhead mechanically when evaluating one additional load.

Order economics

Order-level cost is especially important where small, frequent orders are common. A simplified order cost may be:

Expected return cost can be estimated as:

This is an expected-value calculation, not a prediction that every order will be returned. It is useful when decisions are repeated across many orders.

6. Pricing and Service Differentiation

CTS should inform, but not dictate, pricing. A sound pricing process has four steps:

1. Establish the cost floor relevant to the time horizon.
2. Estimate customer value and willingness to pay.
3. Assess competitive alternatives and capacity constraints.
4. Design a transparent price structure aligned with service consumption.

Possible pricing mechanisms include:

• minimum order values;
• delivery-zone or remote-area surcharges;
• order-line or small-order charges;
• fuel or indexed transport adjustments;
• peak-period premiums;
• return or redelivery fees;
• service-level tiers;
• pallet, case, parcel or stop-based pricing; and
• gain-sharing arrangements based on consolidation or forecast accuracy.

Differentiated service is preferable to uniform service when customers value speed and reliability differently and when the firm can measure the cost of providing those levels. Service menus should specify cut-off times, lead times, delivery windows, fill rates, minimum quantities, claims rules and surcharge conditions.

Contract pricing

A logistics contract should distinguish base charges from pass-through or indexed elements. Common structures include a fixed management fee plus activity charges, a cost-plus arrangement with agreed efficiency incentives, or a bundled price with performance standards. Contracts should define:

• the cost and activity baseline;
• volume bands and capacity reservations;
• fuel, wage, toll and inflation adjustment rules;
• data ownership and audit rights;
• service-level credits and liability limits;
• change-control procedures; and
• termination and transition obligations.

A cost-plus contract can reduce supplier risk but may weaken cost-reduction incentives. A fixed-price contract creates stronger efficiency incentives but requires accurate forecasting and risk premiums. Open-book gain sharing can combine transparency with improvement incentives if measurement rules are robust.

7. Case Study: Northstar Industrial Distribution

The following is a fictional illustrative case.

Background

Northstar supplies maintenance components to three customer groups: national manufacturers, regional wholesalers and online buyers. All customers were offered next-day delivery without a surcharge. Sales reports ranked the national manufacturers as the most attractive customers.

Problem

Operating profit had declined despite stable revenue. Management suspected that emergency orders, fragmented online orders and remote deliveries were consuming warehouse and transport capacity. However, the accounting system reported cost by department rather than customer.

Analysis

The company built an ABC-based CTS model with pools for order entry, picking, packing, storage, transport stops, returns and account management. It used practical capacity for warehouse activities and separate route data for delivery. Results showed:

• national manufacturers generated high revenue and positive profit, but emergency orders reduced their margin;
• regional wholesalers were moderately profitable because orders were consolidated;
• online buyers produced attractive gross margins but negative fully loaded profit in remote zones because of parcel and redelivery costs;
• several low-volume accounts were profitable only when account-management time was excluded; and
• a product family with high unit margins created disproportionate returns and repacking work.

Approach and solution

Northstar introduced three service tiers: consolidated economy delivery, standard next-day delivery and premium emergency delivery. It established minimum order values for remote delivery, a transparent emergency surcharge, and a return-handling charge where returns were avoidable and contractually permitted. Sales incentives were changed from gross sales to contribution after CTS. Customers were offered discounts for electronic ordering, forecast sharing and delivery-window flexibility.

The company also redesigned two routes, combined low-density stops and reviewed the packaging of the high-return product family. Importantly, it did not terminate customers solely because of allocated corporate overhead. Instead, it separated short-run contribution from long-run account economics and negotiated improvement plans.

Outcome

After implementation, management reported improved order density, fewer emergency shipments and better visibility in contract negotiations. The most significant benefit was not a single surcharge; it was the alignment of service promises, operational behaviour and commercial incentives. The case demonstrates that profitability analysis is most valuable when it changes decisions rather than merely ranking customers.

Learning points

1. Gross margin is insufficient where logistics intensity varies.
2. Service tiers make cost differences commercially manageable.
3. Practical capacity reveals unused resources without overstating activity rates.
4. Customer exit should follow an avoidable-cost and strategic-capacity assessment.
5. Sales compensation can undermine CTS unless it rewards profitable revenue.

8. Data, Technology and Governance

A CTS model normally integrates enterprise resource planning, warehouse-management, transport-management, customer relationship and finance systems. Useful data include order timestamps, lines, weights, cubes, stops, kilometres, labour minutes, storage days, returns, claims and service failures.

Analytics and AI can assist with driver discovery, anomaly detection, shipment consolidation, demand forecasting and scenario simulation. However, algorithmic recommendations require human review. A model trained on historical service patterns may reproduce biased credit limits or systematically penalise remote customers without considering public-service obligations or strategic coverage.

Governance controls should include:

• documented cost definitions and allocation rules;
• version control for rates and assumptions;
• reconciliation to the general ledger;
• data-quality checks and exception reports;
• sensitivity and scenario analysis;
• access controls for commercially sensitive customer data; and
• periodic review of whether drivers remain causal.

Pricing and segmentation must comply with competition law, contract law, consumer-protection rules and applicable data-protection requirements. Charges should be disclosed clearly, applied consistently, and justified by service or cost differences rather than protected characteristics. Ethical practice also requires that essential access, small suppliers and vulnerable customers are not disadvantaged through opaque automated pricing.

9. Limitations and Critical Evaluation

CTS models are powerful but not objective facts. Their results depend on:

• the boundary between product cost and logistics cost;
• the quality and granularity of operational data;
• the choice of cost drivers;
• treatment of shared and unused capacity;
• assumptions about returns and service failures;
• allocation of account-management and corporate costs; and
• the time horizon of the decision.

Activity-based models can become expensive and bureaucratic. Excessive precision may create false confidence. A driver that is statistically correlated with cost may not be causally controllable. Customer profitability can also be dynamic: a currently unprofitable account may be strategically valuable because it provides scale, learning, network density or future growth.

Managers should therefore use ranges and scenarios rather than a single point estimate. Sensitivity analysis may vary order frequency, retention, fuel cost, labour rates, return probability and delivery density. The most decision-relevant question is often: “Which assumptions would change the recommendation?”

Best-Practice Decision Process

A robust managerial process is:

1. Define the decision and time horizon.
2. Map the physical and information flows.
3. Identify cost objects and material activities.
4. Select causal, measurable drivers.
5. Reconcile the model to financial accounts.
6. Separate avoidable, unavoidable, fixed and variable costs.
7. Calculate profitability at customer, product, channel, route and order levels.
8. Test assumptions and unused-capacity treatment.
9. Design service and price alternatives.
10. Pilot, monitor behavioural effects and revise the model.

Chapter Summary

Cost-to-serve analysis explains how customers, products, channels, routes and orders consume logistics resources. Its foundation is cost causation: activities and drivers should reflect how work is actually generated. Activity-based rates, contribution margin, customer profitability and customer lifetime value provide complementary views, but each depends on explicit cost boundaries and assumptions.

CTS supports service differentiation, minimum-order policies, route redesign, contract structures and transparent surcharges. It should distinguish short-run incremental decisions from long-run capacity and network decisions. Pricing must also consider customer value, competition, risk and strategic importance; cost is a floor or reference point, not the whole price.

Effective implementation requires integrated operational data, practical-capacity treatment, reconciliation, sensitivity analysis and governance. Ethical and regulatory safeguards are necessary when customer segmentation and algorithmic pricing affect access or commercial terms. The ultimate purpose of CTS is not to produce a sophisticated report, but to align service promises, operational resources and profitable customer relationships.

Key Terms

Cost-to-serve: The cost of activities required to serve a specified cost object.

Cost object: The customer, product, order, route or other unit whose economics are analysed.

Cost driver: A measurable factor that explains activity consumption.

Activity-based costing: A method assigning resource costs through activities and their drivers.

Practical capacity: The usable capacity available after allowing for normal interruptions.

Cost causation: The relationship between an activity and the cost it generates.

Contribution margin: Revenue remaining after specified variable or decision-relevant costs.

Customer profitability: Net customer revenue less product, service and relevant commercial costs.

Customer lifetime value: Present value of expected future customer contribution.

Net revenue: Sales after discounts, rebates, credits and returns.

Avoidable cost: Cost that would disappear as a consequence of a decision.

Unavoidable cost: Cost that remains despite the decision under review.

Service tier: A defined package of service attributes offered at specified terms.

Order economics: The revenue and resource consumption associated with an individual order.

Route profitability: Economic contribution of a transport route or lane.

Expected return cost: Return probability multiplied by the cost if a return occurs.

Open-book contract: A contract in which relevant cost and activity information is shared for pricing or governance.

Unused capacity: Available resource capacity not consumed by current activity.

Review Questions

Short Answer Questions

1. What is the difference between cost incidence and cost causation?
2. Why is practical capacity useful when calculating activity rates?
3. Define customer profitability and identify two costs it may include.
4. Why should net revenue be used instead of gross sales in profitability analysis?
5. State two examples of differentiated logistics pricing.

Long Answer Questions

1. Explain how an activity-based cost-to-serve model can be designed for customer profitability analysis.
2. Critically evaluate the use of customer lifetime value in logistics account management.
3. Discuss how CTS evidence can influence logistics contract structures and service-level design.
4. Compare short-run contribution analysis with long-run customer profitability analysis.
5. Examine the data, ethical and governance risks of algorithmic customer segmentation and pricing.

Analytical / Application Questions

1. A customer generates £80,000 of net revenue, has £52,000 of product cost and consumes £9,500 of CTS. Calculate customer profit and the customer profitability ratio. Explain whether the account should automatically be terminated if allocated corporate overhead makes it appear unprofitable.
2. Design a cost-to-serve model for an e-commerce channel with home delivery, failed deliveries and returns. Specify at least five cost pools and an appropriate driver for each.
3. A customer requests 30 additional urgent delivery stops at £42 each and 300 additional picking lines at £2.20 each. Calculate the incremental cost and recommend two possible pricing or service responses, stating the assumptions required.

Multiple Choice Questions with Answers

Question 1

Which principle is most important when selecting a cost driver?

A. It must be identical for every activity

B. It should causally explain resource consumption

C. It should maximise allocated overhead

D. It should be based only on revenue

Correct Answer: B. It should causally explain resource consumption

Question 2

A cost pool is £90,000 and practical capacity is 4,500 activity units. What is the activity rate?

A. £2 per unit

B. £20 per unit

C. £40 per unit

D. £405 per unit

Correct Answer: B. £20 per unit

Question 3

Net revenue is £47,000, product cost is £30,000 and CTS is £6,100. What is customer profit?

A. £10,900

B. £17,000

C. £23,100

D. £36,100

Correct Answer: A. £10,900

Question 4

Which cost is most likely to be an order-level cost?

A. Corporate head-office rent

B. Annual warehouse lease

C. Order documentation processing

D. Long-term fleet depreciation

Correct Answer: C. Order documentation processing

Question 5

Why can a high-revenue customer be less profitable than a lower-revenue customer?

A. Revenue never affects profit

B. The high-revenue customer may consume disproportionately costly services

C. Lower-revenue customers always have higher margins

D. Profitability excludes logistics costs

Correct Answer: B. The high-revenue customer may consume disproportionately costly services

Question 6

In the simplified CLV model, what does the retention probability represent?

A. The probability that the customer remains active in a future period

B. The probability that a shipment is delivered on time

C. The proportion of warehouse capacity used

D. The probability that a product is returned

Correct Answer: A. The probability that the customer remains active in a future period

Question 7

Which pricing mechanism directly discourages very small orders?

A. Minimum order value

B. Corporate overhead allocation

C. Annual financial audit

D. Fixed warehouse lease

Correct Answer: A. Minimum order value

Question 8

Why should avoidable and unavoidable costs be distinguished?

A. To ensure every cost is allocated equally

B. To identify which costs a decision can actually remove

C. To eliminate all fixed costs from analysis

D. To replace operational data with financial data

Correct Answer: B. To identify which costs a decision can actually remove

Question 9

A route has revenue of £12,000, variable transport cost of £7,000 and route-specific service cost of £1,500. What is route contribution?

A. £3,500

B. £5,000

C. £6,500

D. £20,500

Correct Answer: A. £3,500

Question 10

Which is an important governance control for a CTS model?

A. Changing drivers whenever results are inconvenient

B. Avoiding reconciliation with the general ledger

C. Documenting definitions, assumptions and allocation rules

D. Using only historical sales revenue as a driver

Correct Answer: C. Documenting definitions, assumptions and allocation rules