Last-Mile Delivery Cost: How to Calculate Cost per Successful Delivery
To calculate last-mile cost per successful delivery, divide the total last-mile costs attributable to a defined shipment group by the number of shipments in that group successfully delivered. Include the costs of unsuccessful attempts and redeliveries in the numerator, but count each successfully delivered shipment only once. For a merchant or delivery operator, this measures what it actually costs to complete a delivery—not simply to dispatch a parcel or make an attempt.
The formula is straightforward. The harder work is defining success, matching costs to the right shipments, and separating expenses you control from charges already included in a carrier invoice. This guide provides a practical management-accounting method, a worked example, and a repeatable reporting process. It is not a prescribed financial-reporting standard or an industry benchmark.
The formula and the measurement boundary
Cost per successful delivery = Total attributable last-mile delivery costs ÷ Unique successfully delivered shipments
Before calculating anything, document three definitions:
- Where last mile begins: For example, when a shipment enters the local delivery depot or is handed to the final delivery carrier. Choose a boundary that matches the costs your organization can identify.
- What counts as successful: A shipment with an accepted delivery outcome under your service promise. For home delivery, that might be a validated delivered event. For a collection service, it might require customer collection rather than arrival at the pickup location.
- What unit you count: Usually a shipment or parcel, not an order, stop, delivery attempt, or tracking event.
A delivery stop can serve several parcels, and one order can generate several shipments. These units are not interchangeable. If your business needs an order-level measure, calculate it separately and explain how split shipments are handled.
Use the same definition across periods. Otherwise, an apparent improvement may come from changing the denominator rather than improving delivery operations.
Which costs belong in the calculation?
Start with your operating model. An own-fleet operator needs to assemble delivery labor, vehicle, and operating expenses. A merchant buying delivery services should start with carrier invoices and add only the relevant costs it bears separately.
| Cost category | Own-fleet operation | Merchant using a carrier |
|---|---|---|
| Delivery execution | Driver wages, employer labor costs, and attributable contractor payments | Invoiced delivery charges within the chosen last-mile boundary |
| Vehicles and travel | Allocated lease or depreciation expense, maintenance, insurance, fuel or electricity, tolls, and parking | Separately billed surcharges or accessorial charges, where applicable |
| Dispatch and local facilities | Attributable dispatch labor, routing tools, and local depot costs | Merchant-paid delivery coordination costs not already included elsewhere |
| Delivery exceptions | Additional handling, address correction, and delivery-specific support costs | Separately charged reattempts, corrections, and attributable internal support |
| Credits and adjustments | Adjustments to costs already included in the calculation | Applicable invoice credits, refunds, and billing corrections |
Do not count the same expense twice. If driver pay includes time spent on redeliveries, that time is already in labor expense. You may classify part of it as exception labor, but you should not add it again. Likewise, a merchant should not estimate a carrier's fuel and driver costs on top of an invoice that already covers those services.
Keep the core delivery metric distinct from broader losses
Product replacement costs, customer compensation, lost merchandise, and reverse transportation can matter to profitability. They do not all need to sit inside the core last-mile delivery metric.
A useful approach is to report two clearly labeled measures:
- Core delivery cost per success: Delivery execution, attributable operating overhead, and delivery-exception handling within your boundary.
- Expanded delivery-related cost per success: The core measure plus explicitly defined items such as return-to-sender transport or delivery-related customer compensation.
Do not compare the expanded measure with another operation's narrow carrier-charge measure. The labels may sound similar while the underlying expenses differ substantially.
Allocate shared expenses using an operational driver
For fully loaded reporting, allocate shared costs using a documented basis: driver hours for dispatch support, vehicle hours or distance for vehicle expenses, and handling activity for depot work. A uniform per-parcel allocation is simpler, but it can hide differences between dense routes and remote deliveries.
Distinguish fixed and variable costs when interpreting the result. OpenStax's managerial-accounting treatment explains that fixed costs remain constant in total within a relevant activity range, while variable costs change with activity. Consequently, average delivery cost can fall as existing capacity handles more shipments, even without a lower rate for labor or fuel.
Worked example: dispatched parcels are not completed deliveries
The following figures are hypothetical and expressed in USD. They illustrate the method, not typical market pricing.
A delivery operation tracks a cohort of 1,000 shipments. Of these, 800 succeed on the first attempt. Another 100 succeed after one additional attempt. The remaining 100 do not achieve a successful delivery within the reporting window. The operation therefore records 1,100 attempts and 900 unique successful shipments.
| Cost item | Illustrative amount |
|---|---|
| Delivery labor, including reattempt time | $6,000 |
| Allocated vehicle costs, excluding fuel | $1,800 |
| Fuel, tolls, and parking | $900 |
| Allocated dispatch and delivery software | $600 |
| Additional exception handling, excluding costs above | $900 |
| Total attributable costs | $10,800 |
- Cost per dispatched shipment: $10,800 ÷ 1,000 = $10.80.
- Cost per delivery attempt: $10,800 ÷ 1,100 = approximately $9.82.
- Cost per successful delivery: $10,800 ÷ 900 = $12.00.
- First-attempt success rate: 800 ÷ 1,000 = 80%.
- Delivery completion rate within the window: 900 ÷ 1,000 = 90%.
All three cost measures are valid, but they answer different questions. Cost per attempt describes activity. Cost per dispatched shipment describes the cost of accepting work. Cost per successful delivery connects total spending to completed outcomes.
The costs of the 100 unsuccessful shipments remain in the numerator. Removing them would hide part of the expense required to produce the 900 successful deliveries.
Build a calculation your team can repeat
1. Choose a shipment cohort and observation window
For diagnostic analysis, group shipments by a consistent starting event, such as first dispatch during a particular week. Follow that cohort through a defined observation window and attach its attributable costs.
A calendar-month ratio can also support budgeting, but it needs caution: this month's costs may include attempts on shipments dispatched last month, while some shipments dispatched this month remain open. Label a period-based ratio as such rather than presenting it as a fully matched cohort result.
2. Establish the shipment identifier
Create a stable shipment-level record that links dispatch data, tracking events, carrier charges, and exception handling. Deduplicate repeated delivery scans and repeated invoice lines. Where tracking numbers change, preserve the relationship between the original shipment and its subsequent identifiers.
A replacement shipment is a separate physical movement. Define whether it belongs to a new delivery cohort or an expanded order-level analysis; do not silently combine the two approaches.
3. Normalize delivery outcomes
Map operational statuses into a small set of reporting outcomes, such as delivered, awaiting collection, unsuccessful and closed, or still open. Keep the underlying carrier status so disputed mappings can be reviewed.
A delivered scan is evidence of an operational status, not an automatic resolution of every customer claim. Establish how confirmed misdeliveries, corrected scans, and proof-of-delivery disputes affect the final classification.
4. Reconcile costs and credits
Match carrier invoices to shipments where possible. For own-fleet work, reconcile payroll or contractor payments, vehicle expenses, and shared-cost allocations to the relevant accounting records. Use a consistent currency-conversion policy when costs arrive in different currencies.
Separate actual charges from estimates or accruals. If late invoices are material, issue a provisional result and then a finalized version rather than allowing missing costs to look like an efficiency improvement.
5. Report open shipments explicitly
A shipment still in transit has not yet failed. Report the open-shipment count alongside the ratio and state the observation cutoff. For comparisons, use cohorts with similar maturity or apply the same fixed observation window.
If a cohort has no successful deliveries, the ratio is undefined—not zero. Display the total cost and unsuccessful or open shipment counts instead of forcing a numeric result.
6. Publish supporting measures
At minimum, show shipment count, successful deliveries, total attempts, open shipments, first-attempt success rate, and total attributable cost. These make the headline number auditable and help explain why it changed.
Use the result to find the right operational action
A blended average is a starting point, not a diagnosis. Break the calculation down by characteristics that could change the operating decision: delivery zone, service level, parcel size, residential versus commercial destination, carrier, or route type.
Keep enough shipments in each segment to avoid drawing conclusions from isolated exceptions. Also separate differences in service mix from differences in execution. A remote delivery service and a dense urban route should not be treated as equivalent work simply because both deliver parcels.
| Observed pattern | What to investigate | Possible action |
|---|---|---|
| Low cost per attempt but high cost per success | Repeated attempts, incorrect addresses, access problems, or low completion | Address the verified failure causes before optimizing attempt cost |
| High cost with strong completion | Route distance, service commitments, parcel characteristics, and unused capacity | Review route design, capacity allocation, or service pricing |
| Costs rise without a completion change | Invoice changes, labor rates, allocation methods, or a different shipment mix | Separate commercial and accounting changes from operational changes |
| More successful deliveries but a higher ratio | Whether incremental recovery costs exceed the benefit of additional completions | Evaluate the recovery policy using incremental costs and business value |
Shipment visibility can support outcome classification and exception analysis, but tracking events alone do not provide labor costs, carrier invoices, or allocation rules. Build the cost calculation from financial and operational records together; do not treat a tracking dashboard as a complete delivery-cost ledger.
Evaluate improvements on total cost, not just the delivery rate
Return to the hypothetical example. Suppose an intervention produces 50 additional successful deliveries for $300 of additional attributable cost. The revised result is $11,100 ÷ 950, or approximately $11.68 per successful delivery, compared with $12.00 previously.
The incremental cost is $300 ÷ 50, or $6 per additional success. That figure helps assess the intervention, but it does not replace the fully loaded average. Before adopting it, check that the comparison includes all additional costs and that service quality has not deteriorated elsewhere.
For an own-fleet operation, additional work may fit within existing capacity until another driver, vehicle, or shift is required. A short-term improvement should therefore be tested at the volume where the business expects to operate.
Common errors that make the number misleading
- Counting attempts as successes: A shipment delivered on its third attempt contributes one success, not three.
- Excluding failed-delivery expenses: These expenses are part of the cost of producing successful outcomes.
- Mixing orders, parcels, and stops: Choose one unit and retain it throughout the calculation.
- Comparing unequal cost boundaries: Invoice-only costs and fully loaded fleet costs need reconciliation before comparison.
- Changing allocation rules without disclosure: A revised overhead allocation can move the result without changing operations.
- Optimizing the ratio by avoiding difficult shipments: A lower average may reflect a different customer or destination mix rather than better execution.
The most useful reporting package combines a stable headline metric with a short explanation of changes in rates, shipment mix, completion, and capacity utilization. Assign finance responsibility for cost reconciliation and operations responsibility for outcome definitions and exception analysis.
FAQ
Should failed deliveries be included in the calculation?
Include their attributable delivery costs in the numerator. They do not enter the successful-delivery denominator unless they subsequently achieve the defined success outcome within the reporting window. Report shipments still open separately from confirmed unsuccessful outcomes.
Is a parcel delivered to a pickup point a successful delivery?
That depends on the service boundary you define. Arrival at a pickup point and collection by the recipient are different events. If the promise being measured is customer receipt, use confirmed collection. If measuring the carrier's completed handoff to a pickup location, label that narrower outcome explicitly.
Should customer shipping fees reduce delivery cost?
Keep customer-paid shipping revenue separate from the core cost metric. You can also calculate net delivery subsidy per success by subtracting attributable shipping revenue from delivery costs and dividing by successful deliveries. Label it separately so revenue changes do not appear to be operational cost improvements.
What is a good cost per successful delivery?
There is no universal target suitable for every geography, parcel profile, and service promise. Establish an internal baseline using a consistent method, compare similar shipment segments, and assess the result against order economics and service requirements. A low figure is not inherently better if it comes with unacceptable delivery outcomes.
How often should the metric be updated?
Use an operating cadence that matches your data availability. Weekly provisional reporting can highlight emerging issues, while a reconciled monthly view can support financial review. The important distinction is between a timely estimate and a finalized result with mature shipment outcomes and sufficiently complete costs.
Sources
- OpenStax, Principles of Accounting, Volume 2: Managerial Accounting — Identify and Apply Basic Cost Behavior Patterns. Reference for fixed, variable, and mixed cost behavior and the relevant activity range. The shipment-cohort method, operational checklist, and numerical examples in this article are illustrative management-analysis recommendations, not figures or delivery benchmarks from this source.
