The Answering Machine Message Field: What to Put In It

The Answering Machine Message field is a small box with a lot of options behind it. It holds the prompt VICIdial plays when a call is sent to a machine, whether by answering machine detection (AMD (answering machine detection)) on an auto-dial campaign or by an agent clicking the Answering Machine Message button in the transfer conference frame. Getting the value right is what turns a detected machine into a clean voicemail drop (Voicemail drop) instead of dead air.

A plain audio filename

The simplest valid entry is the name of an audio file that lives in your audio store, or a prompt ID you recorded yourself. If you dialed 8168 to record a message, the system reads back a prompt ID starting at 85100001, and you drop that ID straight into this field. You can also use the audio chooser link next to the field to pick an existing prompt. Record the message twice end to end so a clipped start still lets the listener hear it once.

A TTS prompt

If text to speech (TTS (text to speech)) is enabled on your system, you can put a TTS prompt here instead of a recorded file. That lets you generate the spoken message from text rather than recording audio, which is handy when the message changes often.

Per-lead audio with DYN

You can build the filename from lead (Lead) data using the DYN flag. For example, DYN--A--user--B-- for agent 1234 looks for a file named 1234.wav in your audio store. That pattern lets a single field resolve to a different recording per agent or per lead field, so you do not need one campaign per voice.

STC for multi-file messages

When one message is built from several files and you run out of room in the field, put the files into a settings container and enter STC followed by the container ID. VICIdial then uses that container as your answering machine message. This is the clean way to stitch together a greeting, a body, and a closing without cramming everything into one box.

LTT for custom builds

LTT is the advanced option for systems with custom programming that construct the message on a per-call basis. If you use the Agent API vm_message function, enter LTTagent in this field. Only reach for LTT when you actually have that custom code in place.

How the field is read

flowchart TD
  A[Machine detected or agent clicks button] --> B[Read Answering Machine Message field]
  B --> C{Value type}
  C -->|Filename or prompt ID| D[Play audio from store]
  C -->|TTS prompt| E[Synthesize speech]
  C -->|DYN| F[Build filename from lead field]
  C -->|STC container ID| G[Play files from container]
  C -->|LTTagent| H[Custom per call message]

A few practical reminders. The field only does anything when a call actually gets routed to a machine, so it works alongside the AMD send to action and the AMD send to vm exten settings. Keep audio in GSM 8k-8bit mono or WAV PCM 8k-16bit mono so it plays cleanly. And the same value applies whether the route is automatic or an agent triggers it from the transfer frame, which keeps machine handling consistent across your dial method (Dial method).

For the full routing picture, read our AMD and CPD complete guide, and see the end-to-end drop walkthrough in how to leave a voicemail drop.

Want voicemail drops configured for you on managed VICIdial? See VICIfast pricing.