When delivery falls, first select the destination, operator and recent time window. Check the route sample count and delivery trend; a platform-wide average can hide a problem limited to one operator.
Follow a failing message through acceptance, submission, delivery receipt and any retry. The trace should identify the route alias, the operator and the failure reason at each step without exposing the carrier name to support staff.
If a route is degraded or paused, move affected traffic to a healthy alias and verify the next batch. Record the action so the customer and operations team can understand what changed.
Start with a time boundary. Record when the first failure appeared, the affected country and operator, and the count of final receipts. Compare a healthy period with the affected period using the same filters. A sudden change in one operator can be hidden by a stable country-wide average.
For a representative message, follow each event in order: accepted by the platform, submitted on an alias, acknowledged or rejected, delivery receipt, and retry or final failure. Separate a submission failure from a handset delivery failure. A missing receipt is not automatically a confirmed delivery failure; it may still be pending.
Check whether the route was paused or degraded at the time of the message, and whether failover was available. If moving traffic, test a small batch on the new alias and watch the resulting receipts before expanding. Record the operator, timestamp, alias and reason in the ticket. This gives support a concrete explanation and preserves an audit trail for the routing change.