Cost-to-Serve | Why Two Identical Products Can Have Very Different Delivery Costs

Cost-to-serve is an analytical view of the resources consumed to fulfil a particular customer, order, product, route, channel or service promise.

In one line: the product may be identical, but the logistics work required to place it into different receivers’ hands can be completely different.

This is Article 24 in eduKateSG’s 100-article logistics authority build and completes Batch 06. The canonical parent is How Logistics Works. Articles 21–23 established the promise, composite quality and service target. Cost-to-serve asks what the chosen service actually consumes.

Same Product, Different Logistics Job

Imagine two customers buying the same carton of the same product at the same unit price.

Customer A orders one hundred cartons at a time, receives full pallets at a warehouse beside a major route, accepts deliveries throughout the day and rarely returns goods.

Customer B orders one carton at a time, requires next-day delivery to several high-rise locations, accepts only narrow time windows and generates frequent failed-delivery attempts and returns.

The product cost is the same. The logistics work is not.

Freight Cost Is Only One Part of Cost-to-Serve

Transport charges are visible, so they often dominate discussion. But logistics cost begins before the vehicle moves and can continue after delivery.

• Order-processing work.
• Picking labour or automation capacity.
• Packing materials and labour.
• Special handling.
• Staging and dock time.
• Transport and fuel.
• Carrier minimum charges.
• Delivery appointments.
• Waiting time.
• Customs or regulatory processing.
• Customer-service intervention.
• Premium recovery or expedites.
• Failed delivery attempts.
• Returns and reverse logistics.
• Damage, claims and rework.

Cost-to-serve tries to reconnect those activities to the demand that caused them.

Distance Is a Weak Proxy for Logistics Cost

A nearby receiver can be expensive if every delivery is tiny, urgent and difficult to access. A distant receiver can be efficient if orders consolidate into full loads on a stable route.

Distance matters. It is only one dimension among order size, density, frequency, handling, service level, network balance and receiver constraints.

Cost follows work, not kilometres alone.

Order Size Changes Unit Economics

One hundred cartons in one order and one hundred cartons in one hundred orders are physically the same quantity of product.

The second pattern may create one hundred order releases, more pick tasks, more labels, more packages, more parcels, more delivery stops and more chances for exception.

Small orders often consume more logistics activity per unit even when total product volume is unchanged.

Order Frequency Changes Network Rhythm

Frequent small orders can improve the receiver’s inventory position while increasing logistics handling and transport frequency.

Less frequent larger deliveries may reduce per-unit logistics cost while forcing the receiver to hold more stock.

This is a system trade rather than a universal answer. Logistics cost can be shifted into inventory cost, and inventory cost can be shifted into logistics frequency.

Service Level Changes Cost Structure

Service Levels in Logistics showed why not every shipment needs maximum service.

Faster service may require later warehouse cut-offs, premium carriers, dedicated capacity, air freight, more local inventory or smaller consolidation windows. Tighter reliability can require more buffer, redundancy or monitoring.

Those are real resources. Service promises have cost consequences even when the customer sees only a delivery-time label.

Reliability Can Reduce Hidden Cost

A cheap route with frequent late deliveries may create customer-service calls, manual tracking, emergency replacements and compensation.

A slightly more expensive but reliable route can lower total cost-to-serve if it prevents enough exception work.

This is why ASCM’s SCOR framework keeps Cost and Reliability as separate performance attributes. Optimising one while ignoring the other can damage the whole system.

Warehouse Profile Changes Cost-to-Serve

Some customers order full cases or pallets. Others require each-pick work.

A pallet movement can consolidate many units into one forklift task. Picking twelve different small items for a customised order requires more travel, identification and handling.

Two customers buying the same revenue value can therefore consume very different warehouse capacity.

Packing Profile Changes Cost-to-Serve

One receiver may accept standard cartons. Another may require special labels, repacking, retail-ready presentation, temperature control or security seals.

Those requirements add labour, materials, validation and sometimes additional cube or weight.

The selling product is unchanged. The delivered service package is not.

Delivery Density Changes Final-Mile Cost

Ten deliveries in one compact district can be far cheaper than ten deliveries scattered across a large area even when total distance from the depot appears similar.

Stop density influences driving between customers, parking, access, route duration and how many deliveries a vehicle can complete in one shift.

This is why final-mile economics depend heavily on the geography of receivers, not merely origin-to-destination distance.

Receiver Access Can Dominate the Last Kilometre

A commercial building with a dedicated loading bay, clear appointment and fast unloading is different from a dense site with restricted parking, security checks, lift queues and narrow receiving windows.

The driver may cover only a few extra kilometres while spending an additional hour at the stop.

Cost-to-serve should therefore include dwell and handling at the receiver, not only driving time.

Failed Delivery Attempts Multiply Cost

A failed first attempt can create another route stop, customer-service contact, depot handling and additional inventory dwell.

The revenue from the order does not increase, but the logistics work does.

Article 50 later in the authority spine will take failed delivery attempts in full; here the key point is economic: exception frequency belongs in cost-to-serve.

Returns Can Reverse the Economics

Some channels generate much higher return rates than others.

A return can require collection, inbound transport, receiving, inspection, restocking, refurbishment, repackaging, disposal or credit processing.

Looking only at outbound delivery cost can therefore understate the true logistics cost of serving a customer or channel.

Exceptions Are Expensive Because They Break the Standard Path

Standard flow benefits from scale and repetition. Exceptions often require manual work.

• Find the shipment.
• Investigate what failed.
• Call the carrier.
• Communicate with the receiver.
• Rebook or reroute.
• Approve premium freight.
• Repair inventory or documentation records.

A customer with many unusual requests or recurrent exceptions can therefore have a high cost-to-serve even if basic freight rates are low.

Cost Allocation Is the Difficult Part

Some costs attach directly to an order: a carrier invoice, special packaging or dedicated courier.

Other costs are shared: warehouse rent, software, management, automation, line-haul capacity and support teams.

Cost-to-serve analysis therefore needs allocation rules. Those rules should reflect actual resource consumption as closely as practical rather than simply dividing every shared cost by revenue or order count.

Activity Drivers Make Cost More Legible

Useful drivers may include:

• Orders processed.
• Order lines picked.
• Units or cases handled.
• Pallet movements.
• Packages produced.
• Vehicle stops.
• Kilometres or route hours.
• Delivery attempts.
• Returns processed.
• Exception cases.
• Storage days or cube.
• Special-service events.

No one driver explains all cost. The point is to map shared resources to the activities that consume them.

Revenue Can Hide an Unprofitable Service Pattern

A large customer may generate impressive sales and equally impressive logistics complexity.

If every order is urgent, customised, small, difficult to deliver and frequently returned, gross revenue alone can hide weak contribution after service cost.

Cost-to-serve does not answer the full profitability question by itself, but it makes the logistics part of that question visible.

Do Not Use Cost-to-Serve as an Excuse to Punish Difficult Customers

The purpose is diagnosis, not blame.

High cost-to-serve can reveal a better design opportunity: consolidated delivery days, improved order minimums, lockers, different packaging, route redesign, clearer appointments, service-tier changes or removal of recurrent process failure.

Sometimes the customer genuinely needs the expensive service and pays appropriately for it. Sometimes both sides are trapped in a pattern that can be redesigned.

Cheap Logistics Can Create Expensive Receiver Problems

A cost-to-serve model that includes only the seller’s direct logistics expense can encourage under-service.

If lower freight cost causes chronic lateness, the receiver may hold more stock or suffer disruption. Those costs sit outside the seller’s ledger but can damage the relationship or market position.

The immediate cost view should therefore be interpreted alongside service and value, not as a command to choose the cheapest possible route.

Cost-to-Serve at Three Zoom Levels

One order

Which activities and exceptions did this order consume?

One customer or channel

Which recurring service pattern makes this segment more or less expensive than its peers?

One network

Which shared capacity, route design or service promise creates cost across many customers, and where would redesign improve both economics and reliability?

A Singapore Lens

Singapore’s compact geography can tempt observers to assume local delivery cost should be uniformly low. In practice, dense high-rise access, delivery windows, parking constraints, security procedures, e-commerce parcel fragmentation and specialised cold-chain or industrial requirements can create very different service costs within short distances.

The island is small. The logistics work can still be large.

Hostile Test: “Customer B Costs More Because It Is Farther Away”

Maybe. Trace the activities first.

Does Customer B place smaller orders? Require more pick lines? Demand tighter service? Create more delivery attempts? Hold drivers longer at the site? Generate more returns? Need special packaging?

Distance may be the cause. It may merely be the easiest visible explanation.

Cost-to-Serve Audit

• What object is being analysed—order, customer, product, lane or channel?
• Which costs are direct?
• Which costs are shared?
• Which activity drivers allocate shared resources credibly?
• How many order lines and handling touches are created?
• What packaging and special-service work is required?
• How much transport time and stop time is consumed?
• How many failed attempts or exceptions occur?
• What reverse-logistics cost follows delivery?
• Which service-level promise creates premium cost?
• Which costs arise from preventable unreliability?
• Could a different delivery pattern improve both customer outcome and economics?
• Does the cost view remain paired with reliability, responsiveness and receiver value?

Evidence and Further Reading

ASCM’s current SCOR Digital Standard 2025 separates Reliability, Responsiveness, Agility and Cost as distinct performance attributes and identifies Total Supply Chain Management Costs as a strategic cost metric. That separation is important for cost-to-serve analysis: cost should be diagnosed alongside service performance rather than used as a substitute for it.

Return to the Logistics Hub

Cost-to-serve completes Batch 06’s measurement sequence: keep the promise, test the full order, set the right service level, then trace what that service consumes. Return to How Logistics Works | How the Right Thing Reaches the Right Place at the Right Time to reconnect measurement to warehouse execution, transport, handoffs and final receipt.

---

Final compression: two identical products can have different delivery economics because logistics cost follows the work required to keep each receiver’s promise. Cost-to-serve makes that hidden work visible so service can be redesigned instead of merely declared expensive.