Diagnosing carrier rejects: reading the codes your provider sends back

When a campaign's connect rate falls off a cliff, the carrier is usually telling you why — you just have not read the message yet. Every rejected call carries a hangup cause the provider sent back. Group those causes, find the one that dominates, and you know whether the problem is your caller-ID, your number format, or the trunk itself.

Where the carrier's verdict lives

The Hangup Cause Report shows the carrier hangup causes for every call going out of your systems, and it lets you define which specific causes to display so you can isolate one at a time. The Carrier Log Report sits next to it, showing all dial attempts and their response codes with a downloadable log. Together they turn a vague "calls are failing" into a counted list of reasons from the Carrier.

flowchart TD
  A[Connect rate dropped] --> B[Open Hangup Cause Report]
  B --> C[Filter to dominant cause]
  C --> D{Which code}
  D -->|403 Forbidden| E[Fix CID or trunk auth]
  D -->|484 or format error| F[Fix number format]
  D -->|503 overload| G[Slow pacing or call carrier]
  D -->|488 media| H[Fix codec offer]

How to filter to one cause

1. Run the Hangup Cause Report for the affected campaign and the window where things went wrong. Read the counts, not the individual rows — you want the cause with the biggest number.
2. Define that one cause in the report so only those calls show. Now confirm it is concentrated on one trunk, one CID (caller ID), or one number pattern rather than spread evenly.
3. Cross-check the Carrier Log Report for the same window to see the raw response code alongside the cause, and download it if the provider asks for evidence.

What the common rejects mean

• 403 Forbidden — the carrier refused the call outright. Almost always a caller-ID it will not accept, an unregistered source IP, or a number you are not allowed to dial. This is the most common cause of a sudden cliff.
• 484 Address Incomplete or a format reject — your dialed digits do not match what the carrier expects. Switch the campaign to E.164 formatting or fix the dial prefix so numbers leave as +1XXXXXXXXXX.
• 503 Service Unavailable — the carrier is out of capacity or rate-limiting you. A SIP 503 Service Unavailable in bulk means slow your pacing and talk to the provider; it is rarely something you fix on the box.
• 488 Not Acceptable Here — a media or Codec mismatch. Offer only the codecs the trunk supports.

Reading the spread before you change anything

The mistake people make is reacting to the first failing call they see. One 503 is noise; ten thousand 503s in an hour is a story. Always read the counts across the whole window first so you change one thing for one reason. If two causes are both large — say a pile of 403s and a pile of 484s — treat them separately, because they have different fixes and chasing both at once just muddies the result.

Also watch where a cause concentrates. A reject that lands evenly across every caller-ID and every number pattern usually means the trunk or your source IP, not the data. A reject that lands on one CID or one area code means the data, not the trunk. The Carrier Log Report download is useful here precisely because you can pivot the raw rows by trunk, by CID, and by destination to see the pattern your eyes miss on screen.

When you do hand evidence to the provider, give them the unique IDs and the timestamps for a handful of failing calls. A carrier can find a specific call far faster from a unique ID than from a vague complaint that calls are failing, and that shortens the time to a real answer.

Once you can name the dominant cause, the fix is concrete: change the CID, fix the number format, or move the traffic to a healthier trunk. For the broader method see the troubleshooting playbook, and for a field-by-field tour read how to read the Hangup Cause Report. Carrier auth, trunk health and CID rotation are the kind of thing a managed VICIfast box keeps off your plate.