Answering Machine Detection
Can Vital Interaction tell whether a call was answered by a human or machine?
Vital Interaction uses a feature called Answering Machine Detection (AMD) which can roughly determine if an outbound call has reached a human or if it has reached an answering machine.
Today’s technology does not allow for 100% accuracy in determining if it is a live person or a machine, yet it does work as designed in the vast majority of cases. Thus, in certain cases the answering machine message may be played for a live person and vice versa. Additionally, the timing of the recorded message on the answering machine may not be exact. In other words, the message may start too early or too late. As a result Vital Interaction plays the message multiple times in order to ensure that the full message is delivered, no matter what.
Please note, all automated voice providers must deal with the technology imperfections of AMD.
Even with the imperfections of the technology related to AMD, a vast majority of medical practices have chosen to use automated voice as an important part of their patient communications / engagement process.
How AMD works
Answering Machine Detection listens to the first few seconds of a call and analyzes the audio. There is no consistent signaling difference between a call picked up by a human or a machine, so Vital Interaction relies on analyzing the sound patterns during the first few seconds of a call.
When a human answers a call, the typical pattern is to say “Hello” and then wait for the other party to respond “Hello”. Basically, you can think of it as sound, followed by silence. Although some message machines or services also now simply answer “Hello”; still others say nothing while just emitting a beep to indicate a recording.
However, the typical pattern for voicemail is to continue speaking and say something like “Hi, you have reached the Smith family we are not in right now…”. This is constant sound with no silences. It is this pattern that AMD is listening for. Sound followed by silence means human, constant sound means voicemail.
AMD relies on measuring a greeting against typical speech patterns. Because of this, it will not work 100% of the time. When the message is delivered as though it is to a live person, without waiting for the recorded message, then partial message delivery may result. We repeat every message 2-3 times to ensure that the entire message is delivered even if the AMD made a wrong choice.
Vital Interaction Customer Satisfaction Team