A failure is shown on the message, inside the conversation: Failed to send, the explanation where Lynomia Chat recognises the code, WhatsApp's own words, and a Retry button where retrying could possibly work.
That is precise and it is easy to walk past. A conversation containing a failed message looks like any other conversation in the list.
Finding one again
If you know the customer. Search for the contact and open the conversation. The failed message is in place, in the timeline, where it was.
If you do not. Two places are worth looking:
| Where | What it tells you |
|---|---|
| The campaign's own report, under Campaigns | which sends failed, for a campaign. See WhatsApp campaigns |
| The operator's diagnostic, on the server | the recent failed outgoing messages on a WhatsApp inbox and what each refusal said |
The second is the operator's, not yours, but it is the honest answer to "which messages failed on this number this week" and it is worth knowing it exists before you spend time reconstructing it by hand.
Listing them
Conversations → Filter → Message delivery → Equal to → Failed to send. The condition asks whether the conversation contains a message with that delivery status, so it finds the conversation however long ago the failure happened and wherever it sits in the thread.
It combines with the other conditions, which is where it earns its keep:
| Add this | To get |
|---|---|
| Status is Open | failures still waiting on somebody |
| Inbox is your WhatsApp number | failures on that number only |
| Assignee is you | your own |
| Last activity is 7 days before | the ones that have gone quiet since |
Save it as a folder. Once the filter is right, save it, and the queue is one click away from then on rather than something to rebuild. See work in the inbox.
One honest note: a conversation whose failure was retried successfully still contains the original failed message, so it still matches. The filter answers "something failed here", not "something is still broken here" — open the conversation to tell the two apart.
Two habits make the rest less costly:
- Watch the composer, not just the message. Where the 24-hour window has closed, the composer is disabled before you type — that is the failure prevented rather than recorded.
- Read a campaign's report after it finishes, rather than treating completed as delivered.
Who can do this
Any agent sees failures in the conversations they can open. A campaign's report is an administrator's. The server-side diagnostic is the operator's.
Limits
- No notification is raised when a message fails. The agent who sent it sees it; nobody else is told, so the filter is something somebody has to look at.
- The filter does not distinguish a failure that was later retried from one that was not.
- A failed message is not retried automatically, ever, by design — see error 131042 for why that matters.
- Retry is withdrawn where a retry would simply repeat the refusal — see error 131049.
- A failure older than a day cannot be retried from the conversation. Send again instead.