8 Carrier Performance Metrics to Score Your Carriers

Score your carriers using eight metrics: on-time delivery, on-time pickup, first-attempt delivery success, transit-time consistency, loss rate, damage rate, tracking visibility, and cost per successful delivery. Together, these measures help logistics and ecommerce teams compare service reliability, customer experience, and total cost rather than choosing carriers on shipping rates alone.

A useful carrier scorecard also needs consistent definitions. Compare similar shipments, keep unresolved deliveries visible, and separate carrier-controlled failures from warehouse delays or customer address problems. Otherwise, a precise-looking score can reward the wrong carrier.

Set the comparison rules before calculating scores

The operations or logistics manager should own the scorecard, with input from finance, customer support, and fulfillment. Each team holds a different part of the record: dispatch timestamps, tracking events, invoices, claims, and customer complaints.

Start with shipment-level data. Ideally, each record connects the order, package, carrier, service level, destination, promised delivery deadline, physical handover, tracking history, and final outcome. Record costs and claims separately when they arrive later.

• Compare equivalent services. Separate economy from express, domestic from cross-border, and standard parcels from oversized shipments.
• Use consistent cohorts. A delivery-reliability cohort can contain shipments whose promised deadlines fall in the reporting period. A pickup cohort should contain pickups scheduled during that period.
• Distinguish label creation from handover. Creating a label does not, by itself, establish that a carrier physically received the parcel.
• Retain two views of failures. Show the customer-facing outcome and a second view based on documented responsibility. Do not quietly remove difficult shipments.
• Show sample size and data completeness. A percentage based on a handful of shipments should not carry the same confidence as a stable result across a larger comparable sample.

The formulas below are recommended operational definitions, not a universal industry standard. Agree on them with your team and, where relevant, your carriers before using the results in contract discussions.

1. On-time delivery rate

What it measures: Whether shipments reach customers within the delivery commitment being evaluated.

Formula: Shipments delivered by their agreed deadline ÷ shipments due for delivery in the cohort × 100.

Choose the deadline before shipping. For carrier evaluation, this may be the contracted service commitment. For the delivery experience, it may be the date shown to the customer. These are different promises, so track them separately when they differ.

A shipment that remains undelivered after its deadline is a missed commitment, not a reason to remove it from the denominator. Keep unknown delivery timestamps in a separate data-quality category rather than assuming they were on time.

How to act: Break missed commitments down by route, service, destination type, and handover day. If one service repeatedly misses deadlines on a particular lane, test a routing change there before moving all volume away from the carrier.

2. On-time pickup rate

What it measures: Whether the carrier collects ready shipments within the agreed pickup window.

Formula: Eligible scheduled pickups completed within the agreed window ÷ eligible scheduled pickups × 100.

Define the unit carefully. A scheduled collection of many parcels is normally one pickup event. If you also need parcel-level collection performance, report it separately rather than mixing events and parcels in one rate.

Use warehouse-ready timestamps, collection records, or other agreed handover evidence. The first tracking scan may occur after collection, so it is not always a reliable substitute for the actual pickup time.

How to act: Compare carrier arrival with warehouse readiness. Late packing is a fulfillment issue; a missed collection after documented readiness is a carrier issue. Repeated missed collections may justify changing pickup windows, escalation procedures, or backup arrangements.

3. First-attempt delivery success rate

What it measures: How often a shipment is successfully delivered on its first recorded delivery attempt.

Formula: Shipments delivered on the first attempt ÷ shipments with an identifiable first-attempt outcome × 100.

State how you classify parcel lockers, collection points, recipient redirection, and unsuccessful attempts. For example, delivery to a collection point may be successful fulfillment of the selected service, while a doorstep service redirected there after a failed attempt represents a different outcome.

Also report how many shipments lack usable attempt information. If one carrier supplies detailed attempt events and another supplies only a final delivery event, their apparent rates may not be directly comparable.

How to act: Review failure reasons before assigning responsibility. Address validation, access instructions, and recipient communications may be merchant-controlled improvements. Repeated access failures on a particular route may warrant a carrier service review.

4. Transit-time consistency

What it measures: How predictable delivery time is after physical carrier handover.

Calculation: Measure elapsed time from confirmed handover to delivery, then report the median and a tail measure such as the 90th percentile. A practical variability indicator is the gap between those two values.

The median describes the typical completed shipment. The 90th percentile shows the transit time within which approximately 90% of the measured completed shipments arrived. A smaller gap indicates less variation within that comparable group, but does not automatically mean faster service.

Use consistent calendar-day or business-day rules, time zones, and cutoff treatment. Report overdue open shipments alongside completed-shipment statistics; otherwise, the slowest parcels disappear from the calculation until they finally arrive.

How to act: Use predictable transit performance when setting delivery promises. A carrier with a slightly faster median but a much longer delay tail may require a larger promise buffer than a more consistent alternative.

5. Shipment loss rate

What it measures: The share of handed-over shipments classified as lost under an agreed investigation process.

Formula: Shipments confirmed lost in a dispatch cohort ÷ shipments handed over in that cohort × 100.

Use a sufficiently mature cohort so investigations have time to conclude. Report unresolved investigations separately and update earlier cohorts when outcomes change. Do not automatically label every parcel with stale tracking as lost.

Record the evidence behind each classification. Carrier-confirmed loss, a merchant refund for nonreceipt, and a disputed delivered scan are not necessarily the same event. They can all create business costs, but should remain distinguishable.

How to act: Investigate patterns by handover location, route, service, and shipment value. Assess operational loss separately from claim recovery: reimbursement may reduce financial exposure without fixing the delivery failure.

6. Shipment damage rate

What it measures: How frequently delivered shipments have documented product or packaging damage under your reporting policy.

Formula: Delivered shipments with documented damage ÷ delivered shipments in a sufficiently mature cohort × 100.

Choose a consistent reporting window and specify whether packaging-only damage counts. Count a shipment once even if it generates multiple customer messages or several damaged items. Where item-level damage matters, maintain a separate measure.

A customer report establishes an issue worth investigating, but it does not automatically prove carrier responsibility. Review packaging specifications, photographs, product fragility, warehouse handling, and any inspection findings.

How to act: Compare like products using similar packaging. If damage clusters around one packaging format across carriers, improve packaging first. If comparable shipments show a persistent carrier-specific pattern, investigate handling and service suitability.

7. Tracking visibility rate

What it measures: Whether shipment tracking provides the milestones your team needs to manage delivery and answer customers.

Formula: Eligible shipments with the required milestones available by their defined deadlines ÷ eligible shipments × 100.

Define a milestone policy for each service. You might require evidence of carrier acceptance after handover, a delivery outcome after completion, and exception information when an exception occurs. Evaluate only milestones that are due at the time of measurement.

Track event timing as well as event presence. An event that arrives after customer support has already investigated the shipment has less operational value. Where possible, preserve both the event timestamp and the timestamp when your system received it.

For ecommerce operations, a shipment lookup tool such as 17TRACK can serve as a reference when reviewing individual tracking histories. A carrier scorecard still requires your own agreed milestone rules, shipment cohorts, and reconciliation with warehouse and order records.

How to act: Determine whether missing information originates with the carrier, a tracking integration, or your internal processing. Missing data is a visibility problem; it is not, on its own, proof of a physical delivery failure.

8. Cost per successful delivery

What it measures: The cost of obtaining a completed delivery, including the costs associated with unsuccessful shipments.

Formula: Defined delivery-related costs for a shipment cohort ÷ successfully delivered shipments in that cohort.

For a carrier-only view, include invoiced transportation charges and applicable surcharges, adjustments, and redelivery charges. For a fully loaded business view, you may also include attributable support work, replacement shipping, and unrecovered loss or damage costs.

Keep those views separate. Do not compare one carrier's freight-only cost with another carrier's fully loaded cost. Define how credits and claim recoveries are allocated, and avoid counting the same loss twice.

Wait for the cohort to mature, or label results provisional while invoices and delivery outcomes remain open. Define success consistently, including how you treat returns to sender and recipient-selected collection services.

How to act: Compare cost within similar weight, destination, and service groups. A lower label price is not a saving if failed deliveries and recoverable service problems make the completed delivery more expensive.

Turn the metrics into a practical carrier scorecard

Show each raw metric, shipment count, target, score, and previous-period result. A composite score is useful for summaries, but it should never replace the underlying measures.

These weights are an editorial starting point, not an industry benchmark. A fragile-goods merchant may assign more weight to damage; a time-sensitive business may prioritize delivery reliability.

Convert each metric to a common scale using predetermined thresholds. For a higher-is-better measure, one option is: score = 100 × (actual − unacceptable threshold) ÷ (target − unacceptable threshold), capped between 0 and 100. Reverse the direction for lower-is-better measures. Then multiply each score by its weight and add the results.

For example, with an illustrative on-time delivery target of 98% and an unacceptable threshold of 90%, an actual result of 94% scores 50. These numbers demonstrate the method; they are not recommended service standards.

Do not assign a zero when data is missing. Mark the metric unavailable and withhold the overall score unless a predefined missing-data policy makes the result genuinely comparable. Add minimum service requirements so a low cost cannot compensate for an unacceptable loss or damage outcome.

Use the scorecard to make specific changes

1. Validate the largest gaps. Check shipment records before treating an apparent decline as a real operational change.
2. Find the affected segment. Separate routes, service levels, parcel profiles, and seasonal conditions.
3. Assign an owner. State whether the next action belongs to the carrier, fulfillment, customer support, finance, or the tracking integration team.
4. Agree on a correction. Specify the process change, success measure, and review date.
5. Test before reallocating broadly. Move a controlled portion of comparable volume and evaluate the same measures.

The strongest carrier decision is often route-specific rather than company-wide. Retain carriers where they meet your requirements, correct avoidable process failures, and shift volume where a comparable alternative delivers a better combination of reliability and cost.

FAQ

What is the most important carrier performance metric?

On-time delivery is usually the best starting point because it measures whether the delivery promise was met. It should be considered alongside loss, damage, and cost: a carrier can meet many deadlines while still creating unacceptable problems on the remaining shipments.

How often should carrier performance be reviewed?

Monitor active exceptions frequently enough to intervene, and use a regular review cycle for carrier comparisons. Weekly operational reviews and monthly scorecards are a practical starting approach, not a universal requirement. Lower-volume operations may need longer comparison windows, while loss, damage, and cost measures need time to mature.

Should weather delays and incorrect addresses be excluded?

Keep them in the customer-outcome view, then show documented exclusions in a separate responsibility-adjusted view. Apply the same rules to every carrier. Contractual exclusions and the customer's actual delivery experience answer different questions.

Can tracking data alone produce a complete scorecard?

No. Tracking supports delivery timing, attempt analysis, and visibility checks, but pickup readiness, invoiced costs, damage evidence, and confirmed loss investigations require additional records. A tracking status should not automatically determine financial or contractual responsibility.

How many shipments are needed to compare carriers fairly?

There is no single minimum that fits every metric. Rare events such as loss need more observations than common delivery outcomes. Show counts, use comparable shipment groups, and avoid declaring a winner when a few shipments or unresolved cases could materially change the result.

Sources

The scorecard formulas, weighting example, and operating recommendations above are editorial methodology rather than carrier-published benchmarks. The official tracking references below provide context for shipment lookup and tracking information; they do not establish universal scoring thresholds.

• 17TRACK: Shipment tracking — reference for individual shipment lookup.
• USPS: USPS Tracking — The Basics — official background on tracking information and delivery-status records.