On-Time Delivery Rate, USPS

Metrics type: Key Metrics • Category: Shipping & Courier

At a glance

Share of USPS parcels delivered to the customer on or before the USPS-published service-standard date for the service class booked. The merchant-facing “did the parcel turn up when we said it would” number, computed across every shipment booked through USPS Web Tools or Tracking API in the period. Unlike FedEx Express, USPS Priority Mail does not carry a money-back guarantee (only Priority Mail Express does); a missed commit on Priority Mail is a customer-experience problem, not a refund opportunity.


What it counts	`COUNT(shipments WHERE deliveryDate <= expectedDeliveryDate) / COUNT(shipments WHERE deliveryDate IS NOT NULL)`. Each delivered shipment scores 0 or 1 against its USPS service-standard date. Shipments still in transit are excluded from both numerator and denominator until they receive a final scan.
Delivery success criterion	A shipment is “delivered” when USPS posts a Tracking event of “Delivered” (event code 01) with `deliveryDate` populated. POD signature is required only for Signature Confirmation services; standard parcels deliver on scan or carrier-attended drop.
On-time threshold	`deliveryDate` ≤ `expectedDeliveryDate`. USPS service standards are published as transit-day windows (Priority Mail: 1 to 3 days; First-Class Package: 1 to 5 days; Parcel Select: 2 to 9 days; Media Mail: 2 to 8 days). The card compares against the published end-of-window date.
Money-back guarantee	None for Priority Mail, First-Class Package, Parcel Select, or Media Mail. Only Priority Mail Express carries a guarantee. This is a structural difference from FedEx and a frequent merchant misunderstanding. The card flags lateness; there is typically no refund recourse.
Returns / RTO	Excluded. RTS shipments do not count in numerator or denominator. Tracked separately in Returned to Sender.
Service level scope	All USPS services pooled by default. Priority Mail, Priority Mail Express, First-Class Package, Parcel Select, Media Mail, Ground Advantage all contribute. Each shipment is judged against its own service-standard window. To isolate, see Shipments by Service.
Holiday/election surge	Critical context. USPS volume surges in November-December (holidays) and around US elections (mail-in ballots) typically degrade this card by 10 to 20 percentage points. Network capacity is finite, processing centres run on overtime, and rural routes are most affected. November-December readings are not directly comparable to the rest of the year; benchmark against the same period prior year.
Sunday/holiday delivery	USPS has limited Sunday delivery outside the Amazon partnership (USPS Sunday and holiday delivery is mostly for Amazon Prime parcels). For non-Amazon merchants, the service-standard window assumes Monday-Saturday delivery; the card respects this.
Tracking-data quality	USPS tracking has historically had less granularity than UPS or FedEx, with longer scan-event lags (4 to 24 hours for in-transit scans, sometimes 2 to 4 days for rural routes). Recent investments in IMpb (Intelligent Mail package barcode) have improved this; expect further improvement through 2026.
Time window	`30D vsP` (rolling 30 days, period-over-period comparison)
Alert trigger	`<95% (warn) / <90% (critical)`, sentiment thresholds at good=95, warn=90
Sentiment key	gauge with thresholds good=95, warn=90
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 small-goods merchant (sub-1lb supplements, jewellery, paperback books) shipping out of a Salt Lake City fulfilment centre, using USPS as primary carrier for parcels under 1 pound. Reading taken at 09:00 MT on 12 Mar 26 for the trailing 30 days (10 Feb 26 to 11 Mar 26).

USPS Service	Shipments	Delivered on or before standard	On-Time Rate
Ground Advantage (sub-1lb)	8,420	7,830	93.0%
Priority Mail	4,210	3,873	92.0%
First-Class Package	2,180	2,049	94.0%
Priority Mail Express	360	351	97.5%
Media Mail	240	218	90.8%
All services (this card)	15,410	14,321	92.9%

The card reads 92.9% on the dial; sentiment shows amber (between warn=90 and good=95). Five things to notice:

The headline is structurally lower than FedEx for the same merchant. USPS service standards are wider (1 to 3 days for Priority, vs FedEx Ground’s zone-specific 1 to 5 days), but USPS hits its commits less reliably than FedEx Express; for sub-1lb, USPS is still the cheapest option by a wide margin and the trade-off is accepted. Compare against fedex.fed_otd_rate for the merchant’s mixed-carrier portfolio.
Priority Mail Express at 97.5% is the only service with a money-back guarantee. The 9 missed shipments are refund-claimable via PS Form 3533 at usps.com or in person at a Post Office; refund covers postage paid. Operations should run a weekly script to identify the missed Express tracking numbers and file claims; recovery typically averages $26 per shipment.
Q4 was much worse. This same merchant ran 78.4% on aggregate in their 5 Dec 25 reading, lost 14.5 points to election-and-holiday surge. December reading was 81.2%. The pattern repeats every November-December year-over-year; benchmark against the same period prior year, not against October.
Rural-route degradation drives most of Media Mail’s softness. Media Mail is the cheapest USPS service ( $3.49 to$ 4.13 for sub-2lb), but uses non-priority routing, often handed off to rural carriers running 1 to 3 day backups. Books and educational supplies that ship Media Mail typically miss windows in rural zones.
The 1,089 missed shipments are not all “USPS’s fault”. Address quality (apartment numbers, rural-route box numbers), recipient-not-home (carrier leaves notice), seasonal carrier-substitution coverage, weather embargoes in the Mountain West, and customer-requested redelivery all count against on-time. USPS Tracking API exception events (delivery attempted, delivery exception, undeliverable as addressed) help triage; the card pools them. Actionable subset is typically 50 to 70% of misses.

Note: this card does not separate Sunday-delivered Amazon parcels (which USPS handles for Amazon’s Sunday delivery network) because most non-Amazon merchants don’t use that service. If the merchant has an Amazon Prime SFP relationship, see Shipments by Service.

Sibling cards merchants should reference together

On-time delivery is a customer-facing outcome metric. Pair it with these to diagnose root cause:

Card	Why pair it with On-Time Delivery Rate	What the combination tells you
Priority Mail OTD	Priority-only subset.	Priority Mail is the merchant’s fastest-paid service tier (1 to 3 days). If the Priority subset slips below 95% while First-Class holds, the issue is in mid-tier USPS handling, not network-wide.
OTD by Origin-Destination ZIP Zone	Geography split.	USPS performance varies dramatically by zone; rural East to dense urban West is where most misses happen. Pair to identify whether the headline drop is geographic.
OTD by Route	Route-level granularity.	Splits by sectional centre facility (SCF) and 3-digit ZIP prefix.
Avg Transit (days)	Companion timing metric.	Rising transit days + falling on-time = the network is slower than the standard. USPS may be on its feet but the merchant’s customer expectation is now stale.
Late Shipments	The absolute count behind the percentage.	Useful when total volume swings; percentage can stay flat while late counts double during volume spikes.
SCAN-Form Acceptance Gap	Upstream input.	SCAN forms (manifest scan at acceptance) trigger the USPS service-standard clock. Missing SCAN forms cause “USPS Awaiting Item” status that can mis-score the on-time number.
Exception Rate	Lead indicator.	Exception scans (delivery attempted, delivery exception) precede misses by 12 to 48 hours.
Cross-connector: `fedex.fed_otd_rate`	Peer carrier. Different parcel populations entirely.	USPS is stronger sub-1lb and on Priority Mail flat-rate; FedEx is stronger 2lb+ and on Express. Most multi-carrier merchants run both.
Cross-connector: `shopify.unfulfilled_orders`	Upstream cause.	Shopify backlog predicts USPS on-time-rate dip 1 to 3 days later.
Cross-connector: `shopify.refund_rate`	Downstream impact.	Late deliveries drive refund and chargeback requests.

Reconciling against the vendor’s own dashboard

Where to look in USPS’s own dashboard: USPS Business Customer Gateway → PostalOne! → Reports → Service Performance Measurement (SPM). SPM exposes USPS’s own service-standard performance reporting, but only for shippers using PostalOne! manifest acceptance. For non-PostalOne! shippers, USPS only exposes per-shipment tracking via USPS Tracking; aggregate reporting is not provided directly to small shippers. The card is the operational view across all volume. Why our number may legitimately differ from USPS’s portal:

Reason	Direction	Why
Service-standard window vs target performance	Either	USPS publishes service-standards (the promise to the customer) and target-performance (the network’s internal goal, e.g. 95% of First-Class within 2 days). The card scores against service-standards (the customer-facing promise). USPS’s internal SPM scores against target-performance dates which can differ.
Acceptance scan timing	Either	The clock starts when USPS scans the parcel as accepted (manifest scan or counter scan). Merchants using PostalOne! get an electronic acceptance scan; those dropping at retail counters get a delayed scan, which can move the on-time clock by 1 day.
Election and holiday surge exclusions	USPS may show higher	USPS’s internal performance reporting sometimes excludes “force majeure” days (named storms, declared peak surge, post-election surge). The card includes everything; the gap during November-December and around mid-term/general elections can be 5 to 15 percentage points.
Rural-route scan lag	Ours lower for “today”	Rural-route carriers may scan delivery 12 to 48 hours after actual delivery, especially in winter. Today’s number may understate; T-3 days fully reconcile.
Address-correction holds	Either	USPS Code 04 (delivery exception, address-corrected and forwarded) can re-clock the commit. The card scores against the original; USPS’s portal sometimes scores against the corrected commit.
Sunday and holiday delivery (Amazon)	Either	USPS’s Amazon Sunday parcels are scored on a separate Sunday-delivery network; non-Amazon Sunday-delivered parcels are rare. The card pools all; USPS’s reports often segregate.

Cross-connector reconciliation:

Card	Expected relationship	What causes legitimate divergence
`fedex.fed_otd_rate`	Peer carrier. Independent populations.	Different parcel-weight distribution; FedEx beats USPS for 2lb+ and Express, USPS beats FedEx for sub-1lb.
`easypost.eas_otd_rate`	If EasyPost is the booking layer in front of USPS, EasyPost’s USPS subset is a subset of this card.	EasyPost may rate-shop some USPS volume to FedEx or UPS based on dimensional pricing. Filter EasyPost by `carrier=USPS` to compare.
Customer NPS / refund-rate surveys	Downstream sentiment.	Late deliveries drive lower NPS and higher refund rate at 7 to 14 day lag.

Known limitations / merchant FAQs

Why is my USPS on-time rate lower than my FedEx rate? Is USPS bad? Different jobs. USPS handles a much higher proportion of rural and last-mile-difficult addresses than FedEx (which often hands such parcels off to USPS via SmartPost / Ground Economy). USPS also has wider service-standard windows, so the same calendar slip looks similar but the underlying network constraints are different. For most DTC merchants, USPS is the cheapest option for sub-1lb parcels by 30 to 60% versus FedEx Ground; a 2 to 5 ppt lower on-time rate is the cost of the cost saving. Does USPS Priority Mail have a money-back guarantee? No. This is the most common merchant misunderstanding. Only Priority Mail Express carries a guarantee; Priority Mail does not. A late Priority Mail parcel is a customer-experience problem with no refund recourse. If you need money-back-guarantee semantics, ship Priority Mail Express ($26+ per parcel) or use FedEx Express services. How does USPS handle the November-December holiday surge? USPS hires seasonal staff and adds vehicles, but volume scales faster than capacity. Expect 10 to 20 ppt drop in on-time rate from late November through Christmas. Plan for it: communicate longer-than-normal delivery windows in checkout copy (“delivery in 5 to 10 business days” instead of “1 to 3”), shift premium-tier orders to Priority Mail Express or FedEx Express in the last 5 days before Christmas, and set up a customer-service playbook for delay inquiries. The pattern repeats every year. What is USPS Ground Advantage and how does it differ from First-Class Package? USPS Ground Advantage launched in mid-2023 and merged the old First-Class Package Service, Parcel Select Ground, and Retail Ground into one product with a 2 to 5 day service standard. Most merchants now ship sub-1lb on Ground Advantage by default. The card includes Ground Advantage in the aggregate; if your merchant still has scripts referencing First-Class Package, retire them. My USPS shipment shows “Delivered” in tracking but the customer says they didn’t receive it. USPS marks “Delivered” on carrier scan, which sometimes precedes physical handoff (especially with leave-at-door deliveries or rural cluster boxes). Three steps: (1) ask the customer to wait 24 hours, the parcel is sometimes scanned as delivered to a neighbour or front office. (2) Ask the customer to check with neighbours and household members. (3) If still missing after 24 hours, file a Missing Mail Search at usps.com (free) and offer a replacement or refund. The card counts these as on-time delivered; the loss is operational, not a tracking metric problem. Why is rural delivery so much weaker than urban for USPS? Two structural reasons. (1) Coverage frequency. Urban routes are walked or driven daily (Monday-Saturday); rural routes are sometimes covered every-other-day during low-volume seasons. (2) Substitute carrier coverage. Rural carriers are often part-time and cover multiple routes; substitute carriers covering vacation or sickness can run 1 to 3 days behind on scan-and-deliver. Both factors compound during holidays. The aggregate hides this; pair with OTD by Origin-Destination ZIP Zone. How do I plan for elections? US mid-term (every 2 years) and general (every 4 years) elections cause USPS volume surges from political mail and mail-in ballots. The surge typically runs 4 weeks pre-election and 1 week post. Plan for a 5 to 12 ppt drop in on-time rate during the surge, especially in swing-state ZIPs. The 2024 general election cycle saw the worst USPS performance on record in November of any non-pandemic year. Should I switch from USPS to FedEx for sub-1lb? Run the math. USPS Ground Advantage at

4 to

6 per sub-1lb parcel is roughly 30 to 50% cheaper than FedEx SmartPost / Ground Economy at

7 to

9 per parcel for the same weight. The on-time gap is typically 3 to 5 percentage points (USPS lower); for low-AOV merchants, the cost saving wins. Switch only if (a) your customer is high-AOV and intolerant of the on-time gap, or (b) you can negotiate a better FedEx rate via volume. Why does my USPS rate look better than the public USPS performance dashboard? USPS’s public service-performance dashboard reports against target-performance dates (the network’s internal goal), not service-standard dates (the customer-facing promise). Service standards are wider; the card scores against them. The public dashboard often reads 3 to 8 percentage points lower than the card for the same period. Both are right; they answer different questions (“did the network hit its internal goal?” vs “did the customer get the parcel by the promised date?”). The card swings day-to-day, why? Volume. USPS tracking-data quality has higher latency than UPS or FedEx, especially for rural routes. Today’s reading often understates by 0.5 to 2 percentage points and reconciles within 2 to 3 days. Look at the rolling 30-day on the card, not a daily figure; the alert at <90% trips on persistent issues.

Tracked live in Vortex IQ Nerve Centre

On-Time Delivery Rate is one of hundreds of KPI pulses Vortex IQ tracks across USPS and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.

Get Started

The AI OS

At a glance

Calculation

Worked example

Sibling cards merchants should reference together

Reconciling against the vendor’s own dashboard

Known limitations / merchant FAQs

Tracked live in Vortex IQ Nerve Centre

​At a glance

​Calculation

​Worked example

​Sibling cards merchants should reference together

​Reconciling against the vendor’s own dashboard

​Known limitations / merchant FAQs

​Tracked live in Vortex IQ Nerve Centre

At a glance

Calculation

Worked example

Sibling cards merchants should reference together

Reconciling against the vendor’s own dashboard

Known limitations / merchant FAQs

Tracked live in Vortex IQ Nerve Centre