Address Verification Before Shipment, Not at Delivery

Short answer: A bad address has a cost that triples the longer you wait to catch it. At checkout it is free to fix; before the label it costs cents; after a return-to-sender it costs 2–3× postage plus the product and a refund. The discipline is a pre-shipment verification pipeline — validate at checkout, flag ambiguous orders, re-verify in batch right before labeling — so the carrier only ever receives a clean address.

The cost curve is the whole argument

Stage you catch it Cost Reversible?
Checkout (customer corrects) $0 Yes, instantly
Before label (API flags it) cents Yes, before money spent
Carrier accepts bad label full postage lost No
Return-to-sender 2–3× postage + product + refund No, and the sale is gone

The later the catch, the less reversible and the more expensive. Verification is simply moving the catch to the cheapest, most reversible point.

The pre-shipment pipeline

A practical pipeline has four stages, and the last one is the gate that matters:

  1. Checkout validation (customer-facing). As the customer types, call the API and surface the standardized address or a one-tap confirm on ambiguity. Fixes most typos at the source.
  2. Order flag / hold. If an address is ambiguous and the customer did not confirm, flag the order (or soft-hold it) for review instead of letting it flow straight to fulfillment.
  3. Batch verify before labeling. In your fulfillment job, re-validate every address immediately before calling the carrier to buy the label. This is the last line of defense and the one that protects your postage spend.
  4. Hand the carrier a clean address. Only addresses that passed all three stages reach the carrier API.

Stages 1 and 3 catch different things: checkout catches what the customer can fix; pre-label catches what passed checkout but fails carrier deliverability.

Why "at delivery" is already too late

Once the parcel is moving, the address is locked into the physical world. You cannot edit it, you cannot re-route cheaply, and a failed attempt still costs a delivery attempt. The carrier's "address correction" or return flow is priced for exactly this late discovery — which is why it costs more than the check that would have prevented it.

This connects directly to the other checks in this series:

All four belong in the same pre-ship gate.

Build vs buy

Building a validator means owning postal datasets per country, handling format edge cases (APO/FPO, PO boxes, new construction), re-certifying CASS/DPV yearly, and maintaining uptime at checkout. Unless address data is your product, that is overhead. An API that is already certified, maintained, and internationalized plugs into the same four-stage pipeline for a fraction of the upkeep.

Where AddressGuard fits

AddressGuard is built for exactly this pipeline:

FAQ quick takeaways


Disclaimer: This is fulfillment-process guidance, not carrier or legal advice. Carrier fees, address datasets, and validation rules change; confirm current behavior with your carrier and address provider before relying on any specific check.