At a glance
Absolute count of USPS shipments that delivered after the published service-standard date in the period. Companion to On-Time Delivery Rate, which is the percentage. The count is what the operations team works through one row at a time, filing missing-mail searches, sending customer-service apologies, and triaging refund requests; the percentage is what the executive watches.
| What it counts | COUNT(shipments WHERE deliveryDate > expectedDeliveryDate AND deliveryDate IS NOT NULL) over the trailing 7 days. Each delivered-late shipment counts once. In-transit shipments past their commit are not counted until they receive a final scan. |
| Why count and rate matter together | A 95% on-time rate on 1,000 shipments = 50 late. The same rate on 10,000 shipments = 500 late. The customer-service team works through count, not percentage. A flat rate during 2x volume growth doubles the operational load. |
| Delivery success criterion | A shipment qualifies only after it actually delivers (Code 01). Pre-delivery commitments missed in transit are tracked in Exception Rate. |
| Returns / RTO | Excluded. RTS shipments do not count as late deliveries. |
| Service level scope | All USPS services pooled. Priority Mail, Priority Mail Express, First-Class Package, Ground Advantage, Parcel Select, Media Mail. To split, use Shipments by Service. |
| Customer-impact ladder | Each late count is a potential customer-service interaction: a “where is my order?” email, a refund request, a chargeback. Industry rule of thumb is 8 to 15% of late deliveries trigger an explicit customer touchpoint. The rest are tolerated silently but contribute to NPS drift. |
| Holiday/election surge effect | Counts spike sharply November-December and around mid-term/general elections. A merchant running 50 late shipments / week in October may see 200 to 350 / week in late November and 100+ during the election surge. Plan customer-service staffing around the count, not the rate. |
| Time window | 7D (rolling 7 days, weekly operational view) |
| Alert trigger | >5% of total shipments for the period |
| Roles | owner, operations |
Calculation
Calculated automatically from your USPS data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A US DTC home-fragrance merchant shipping out of Salt Lake City, primarily USPS Priority Mail and Ground Advantage. Reading taken at 09:00 MT on 12 Mar 26 for the trailing 7 days (06 Mar 26 to 12 Mar 26).| Service | Shipments | Late count | Late % |
|---|---|---|---|
| Priority Mail | 1,420 | 92 | 6.5% |
| Ground Advantage | 1,810 | 124 | 6.9% |
| Priority Mail Express | 80 | 2 | 2.5% |
| Media Mail | 110 | 14 | 12.7% |
| All services (this card) | 3,420 | 232 | 6.8% |
- Customer-service workload tomorrow: 25 to 35 inbound contacts. Industry rule of thumb is 8 to 15% of late deliveries trigger explicit customer touchpoints. The 232 late means 18 to 35 emails, calls, or chats this week. Schedule the customer-service team around this number, not the rate.
- Two of the late Priority Mail Express are refund-claimable under USPS’s money-back guarantee. File via PS Form 3533 within 30 days. The other 230 late deliveries have no refund recourse from USPS itself.
- Media Mail at 12.7% late is the worst-performing service. Books and educational supplies on Media Mail are routed through non-priority handling. The merchant should warn customers in checkout copy that Media Mail can take up to 8 days; the card is doing its job by surfacing the lane-level problem.
- Compare against the same week last year. This week’s 232 is up from 168 in week 11 of 2025 (+38%); the trend is climbing faster than volume (+22%). Investigation should focus on whether USPS has degraded or whether the merchant’s customer-mix shifted toward lower-tier services.
- The 232 are not all “USPS’s fault”. Address-quality issues (missing apt numbers, rural-route box errors) account for roughly 15 to 25% of late deliveries; weather embargoes for another 5 to 15%; recipient-not-home for 10 to 20%. Actionable subset is typically 50 to 70% of the count, the rest are exogenous.
actualDeliveryDate > expectedDeliveryDate for the period to get the row-by-row tracking-number list. Customer-service uses this list as their daily worklist.
Sibling cards merchants should reference together
| Card | Why pair it with Late Shipments | What the combination tells you |
|---|---|---|
| On-Time Delivery Rate | The percentage behind the count. | The percentage smooths volume swings; the count is the workload. Always read together. |
| Shipments Total | Volume context. | Late count rising in line with volume = stable rate. Late count rising faster = degrading rate. |
| Exception Rate | Lead indicator. | Exception scans (delivery attempted, exception, addressee unknown) precede late deliveries by 1 to 3 days. |
| Returned to Sender | Disposition. | Some late shipments eventually become RTS if address-correction fails or recipient-not-home retries exhaust. |
| Priority Mail OTD | Subset for the merchant’s most popular service. | If Priority Mail has 80% of the late count, focus there. |
| Open Claims | Refund pipeline. | Late Priority Mail Express shipments flow into claims; the count in this card is the upstream feed. |
| Cross-connector: customer-service ticket-volume | Direct correlation. | Customer-service contact rate correlates 0.7 to 0.9 with this count’s weekly trend at 1 to 3 day lag. |
Cross-connector: shopify.refund_rate | Downstream impact. | Late count surge predicts a refund-rate climb 7 to 14 days later. |
Reconciling against the vendor’s own dashboard
Where to look in USPS’s own dashboard: USPS Business Customer Gateway → PostalOne! → Reports → Service Performance Measurement → Late Mail by Volume. Filter to Last 7 Days, All Services, Shipper-level. PostalOne! is only available to PostalOne!-onboarded shippers; small shippers have no equivalent vendor view. Why our number may legitimately differ from USPS’s portal:| Reason | Direction | Why |
|---|---|---|
| Acceptance-scan timing | Either | The clock starts at acceptance scan. Counter-drop shipments may receive the acceptance scan a day after physical drop-off, deferring the late count by one day. |
| Service-standard window vs target performance | Either | USPS reports against target dates internally; the card scores against the customer-facing service standard. |
| Force-majeure exclusions | USPS may show lower count | USPS sometimes excludes shipments to ZIPs flagged as suspended (declared peak surge, named storms). The card includes everything. |
| Tracking-scan lag | Ours lower for “today” | USPS scan events lag by 4 to 24 hours typically; rural routes can lag 2 to 4 days. Today’s count may understate. |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
fedex.fed_late_shipments_count | Peer carrier, independent populations. | Different parcel-weight distribution and service-class mix. |
shopify.refund_rate | Downstream effect. | Late count predicts refund-rate at 7 to 14 day lag. |
Known limitations / merchant FAQs
The count keeps climbing every week. What’s driving it? Three usual culprits. (1) Volume growth, the count climbs naturally with shipment volume even at constant rate. Compare against the rate. (2) Customer-mix shift toward rural ZIPs, where USPS performance is structurally weaker. (3) Seasonality, especially November-December and election cycles. Open On-Time Delivery Rate to see whether the rate is also climbing; if rate is flat and count is up, growth is the explanation. Can I claim a refund for late USPS shipments? Only Priority Mail Express qualifies for USPS’s money-back guarantee. File via PS Form 3533 within 30 days of mailing for a refund of postage paid (typically 54). Priority Mail, Ground Advantage, First-Class Package, Media Mail, Parcel Select have no refund recourse. The card lists all late deliveries; only the Priority Mail Express subset is claimable. What’s the fastest way to get the tracking-number list? Drill into Shipments Total and filteractualDeliveryDate > expectedDeliveryDate for the period. Export to CSV; customer-service uses this as their daily worklist for proactive outreach.
Should I email the customer about a late shipment before they email me?
Often yes. Proactive “your parcel is running late, here’s the tracking” emails improve NPS even when the delivery itself is late. Industry rule: send the email when the parcel is 24 hours past commit and not yet delivered. Reduces inbound contact volume by 30 to 50% and improves customer perception.
Why does the card disagree with my cust-service team’s count?
Two reasons. (1) The team often counts customer-reported late, the card counts USPS-data-confirmed late. Customer-reported is sometimes wrong (the parcel did arrive, but customer didn’t notice). (2) The team’s count is cumulative across days; the card resets weekly. Reconcile periodically to keep both views aligned.
Should I switch services to lower the late count?
Possibly. If Priority Mail has 60% of the late count and you can move heavier sub-1lb to Priority Mail (faster than Ground Advantage at marginal cost difference), the operations team’s workload drops. The trade-off is shipping cost; weigh against Avg Shipping Cost.
Is 5% late acceptable?
Industry typical for USPS is 4 to 8% late in steady-state, climbing to 12 to 20% in November-December and election surges. Below 5% is healthy; below 3% is excellent (rarely sustainable for sub-1lb USPS-heavy shippers); above 8% in steady-state warrants action.
What’s my cust-service cost per late shipment?
Industry rule of thumb: a “where is my order?” inquiry costs 12 to handle (agent time + system overhead). 232 late deliveries at 12% contact rate = ~28 contacts ≈ 336 / week in handling cost. Compounded across the year, late-shipment cust-service cost on a moderate USPS account is 20K. The case for shipping cost increases that prevent late shipments must beat this.
Why is Q4 such a big spike?
USPS volume can rise 60 to 90% in November-December (holidays + Cyber Week + Black Friday). Network capacity rises maybe 15 to 25% with seasonal staffing. The gap is absorbed as later deliveries; expect 3x to 5x the late-shipment count of October. Plan customer-service staffing for it.