When to Refund, Reship, Replace or Investigate

Start with urgency and reversibility

Some cases need immediate protection: a possible safety issue, a time-sensitive order, a clearly wrong item or a customer already waiting beyond a firm promise. Other cases benefit from one focused evidence request before a costly remedy is chosen. The aim is to avoid both extremes—making every customer prove too much, or reshipping repeatedly without understanding a recurring failure.

Refund when the commercial relationship should close

A refund often makes sense when the order cannot be fulfilled, the delay is no longer acceptable, the item is unsuitable and replacement adds no value, or the seller's policy and applicable consumer rules require it. For US online orders, FTC guidance requires sellers to have a reasonable basis for shipping promises and provides cancellation and refund rights when shipment is delayed; other jurisdictions differ.

Reship when a complete second parcel is the cleanest remedy

Reshipment can fit confirmed loss, an unrecoverable delivery failure, or a complete order problem where replacement parts are impractical. Before releasing it, correct the address or route issue that affected the first parcel. A reship on the same bad data or unsuitable service can repeat the failure.

Replace a part or unit when the failure is contained

A replacement may be better than a full reship when one component is missing or faulty, the rest of the order remains useful, and the replacement can travel safely and economically. Confirm compatibility and whether sending the part creates a confusing support burden for the customer.

Investigate when the next fact changes the decision

• Tracking is silent but still within the normal interval for that stage.
• The cause may be product, packing, transit or setup and evidence can distinguish them.
• Several complaints may indicate a batch-level issue requiring containment.
• The requested remedy is unusually costly, irreversible or inconsistent with the evidence.

Investigation should have a deadline and owner. "Waiting for the carrier" is not a plan unless the next escalation time and fallback remedy are defined.

Use a two-loop case record

• Customer loop: issue, evidence requested, promised update, remedy and completion.
• Operations loop: SKU/version, likely cause, containment, corrective action and later effectiveness check.

CPSC guidance for regulated products highlights the value of complaint, production and quality-control records. Even for ordinary goods, structured records help reveal patterns that isolated chat messages hide.

Do not publish case identifiers

A public lesson can describe the broad product category, destination country, non-identifying quantity and decision process. Keep names, addresses, customer contacts, product links, images, order numbers and tracking identifiers in the private case system.

Choose the remedy from cause, evidence and repeat risk

Choose the fastest fair remedy that does not hide a repeatable problem. Then close the operational loop so the next customer is less likely to need the same remedy.