A shipping policy is working when customers receive accurate promises before ordering and the business consistently follows those promises when orders ship, run late, change, or are cancelled. Measure actual order outcomes against the exact customer-facing wording, then investigate exceptions by product, destination, carrier, and fulfillment route. A low complaint count alone is not proof that the policy works.
Turn policy wording into measurable promises
Inventory every active statement about processing time, shipment time, cutoff times, destinations, carriers, charges, tracking, delays, cancellations, and refunds. Include advertisements, product pages, cart and checkout text, order confirmations, marketplace listings, help pages, and support scripts. Give each promise an owner, version, effective date, and operational data source.
For U.S. sellers covered by the FTC Mail, Internet, or Telephone Order Merchandise Rule, shipment promises need a reasonable basis. The FTC also describes duties that may apply when a seller cannot ship within the promised time, including delay notices, cancellation choices, consent in some cases, and refunds. Review the FTC business guide and its prompt-delivery summary. This operational guide is not legal advice, and other laws or platform rules may apply.
Build an order-level scorecard
For each order, store the promise shown at purchase, promised ship date or window, actual ship date, destination, product class, fulfillment location, carrier, delay-notice time, customer response, cancellation time, and refund time when relevant. Calculate the share shipped within the stated window, but keep the underlying counts and do not combine materially different product or region groups into one reassuring average.
Track the number and rate of orders that required delay notices, whether the notices went out before the promise expired, whether revised dates were supported by current information, and whether cancellations and refunds followed the documented process. The FTC guidance recommends retaining what a delay notice said, when it was provided, and how the customer responded.
Test consistency across the customer journey
Run synthetic orders for common, remote, restricted, backordered, preorder, split-shipment, and high-volume scenarios when those cases exist. Compare the product page, checkout, confirmation, support response, and fulfillment-system result. A policy fails even if the policy page is correct when checkout or advertising makes a different promise.
Review support contacts about late delivery, unclear charges, tracking, cancellations, and refunds. Classify them by root cause rather than treating every contact as a writing problem. A rise may reflect stale policy text, incorrect inventory, carrier disruption, a broken notice workflow, or an unrealistic operational promise.
Set decision thresholds and actions
Define thresholds before reviewing results. For example, require investigation when on-time shipment falls below the internally approved target, when a notice is late, when any customer-facing surface conflicts with the approved version, or when refund evidence is incomplete. These are business controls, not FTC-prescribed percentages.
End each review with a decision: keep the promise, narrow it, change the operation, improve the notice workflow, or escalate for legal review. Preserve the data, policy version, reviewer, exclusions, and corrective action so the next review can distinguish improvement from a changed measurement method.
Sources
- Business Guide to the FTC's Mail, Internet, or Telephone Order Merchandise Rule — Federal Trade Commission
- Selling on the Internet: Prompt Delivery Rules — Federal Trade Commission