How to Count USPS Business Days for Delivery Estimates

What to take away

The metric: USPS delivery business days

A USPS business day is a day the mail moves toward its published window. For most retail classes that means Monday through Saturday. Sunday sits out. So do the eleven postal holidays. Part 608, section 3.2 of the U.S. Domestic Mail Manual groups them into "Widely Observed" (New Year's Day, Memorial Day, Independence Day, Labor Day, Martin Luther King Jr. Day, Thanksgiving Day, Christmas Day) and "Not Widely Observed" (Washington's Birthday, Columbus Day, Veterans Day). If a postal holiday falls on a Sunday, USPS observes it on the following Monday.

The clock starts after acceptance, not at the counter. Hand a parcel over past the local cutoff and the mailing day is not one of the counted days. Cutoff times vary by location and service, so check the schedule posted at your office. The next movement day becomes day one.

USPS in-window count is the number of Monday through Saturday days, holidays excluded, between the acceptance scan and the delivery scan, set against the window USPS quotes for that class. Two parcels in the same class and lane are comparable. Parcels in different classes are not.

How to read a quoted window

Class comes first, because counting rules differ more by class than most shippers assume.

USPS revises windows and product names from time to time, so confirm the current figure on the USPS service comparison page before you promise a date to a customer. The Saturday question changes rarely, and it is where postal counting pulls away from a weekday-only count.

A quoted window is a promise about a class of mail. It becomes a promise about your parcel only when your acceptance scan beats the local cutoff.

Moving a promised date takes arithmetic on a working calendar rather than a wall calendar; adding business days covers dated cases with different start conventions.

What the postal count cannot tell you

The count cannot separate your late drop-off from a network delay. Both show up as a missed window, and the tracking record seldom says which one happened.

It does not predict a lane either. USPS publishes class-level windows, not address-level ones, so a rate that holds for a Chicago loop can miss on a rural route.

Weather and volume break the pattern in weeks nobody can forecast. A December rate measured during a calm month will flatter your ship times.

The count also stops being the right instrument once a legal deadline is involved. A response date on an IRS notice follows tax rules rather than post office rules; see IRS response deadline for that count.

Attribution: whose calendar you are measuring

Whose weekend and whose holidays sit inside the number? Say so before you compare two rates, because a count built on USPS days and a count built on federal days will differ by roughly one day for the same parcel.

Cross-border shipments add a second calendar. A parcel leaving the United States on a Friday can meet a national holiday on the receiving side the following Monday, and the clock stops there instead of running. Country calendars compares per-country holiday sets.

For international mail, USPS hands the parcel to the destination post, and that leg sits outside the window you measured.

Example: a Friday drop-off with a two-day window

A seller hands 30 Priority Mail parcels to a Cleveland counter at 4 p.m. Friday, ahead of the cutoff, with a quoted two-day window and no holiday nearby.

Saturday is day one. Monday is day two. On USPS counting, the promised arrival is Monday.

Run the same parcel through a weekday-only count, the kind a federal office uses, and Saturday vanishes. Monday becomes day one and Tuesday day two. The two answers sit a day apart for a parcel that never left Ohio.

Federal statutes compute time the same way, excluding the opening day and treating weekends and holidays separately under 1 U.S.C. § 1. That is why a date copied from a contract calculator reads late when it is applied to a mail class.

When to stop measuring and decide

Once a week is enough. Thirty shipments in one class give a rate you can act on. Ten shipments do not; three late parcels out of ten is noise, not a trend.

Set the trigger before you look at the result. At 90 percent in window or better, keep the class and the ship time.

From 80 to 89 percent, ship an hour earlier or add a day to any date quoted for a hard deadline. Below 80 percent after 20 shipments in one class, change the class or add a day to every estimate you publish.

Then stop. The rate is a policy input, not a forecast for the parcel in your hand.

Holiday dates underneath the count deserve one annual check, since a missed closure shifts a whole week. Working days per year holidays covers that rebuild.

Common questions