When your campaigns use a sound file longer than 15 seconds, Callfire automatically switches to the newer engine. We hope you are able to experience the difference our beta customers have.

So, what changed?

Old AMD:

Callfire would listen to the customer as they picked up the phone. During the first few seconds, AMD decides whether it’s listening to an answering machine, voicemail box, or live person. In the case of an answering machine or a voicemail, CallFire AMD would wait until after the beep and leave your recorded message.

New AMD:

While playing your message, AMD listens for a voicemail or answering machine. Any such indication will interrupt the playback of the live answer message, and start the answering machine message after the beep. To ensure accuracy, CallFire’s AMD engine requires a 15 second message, or longer, to make it’s decision.

The result?

In some cases, the accuracy of Answering Machine detection jumped to 95%!   This is staggering compared to the 90% accuracy in prior revisions of CallFire’s AMD.  I hope the new engine increases your ROI as it has for our alpha clients already.

For you API developers there’s a way to switch your campaigns to the new answering machine detection as well! Take a look at the “useAnsweringMachineDetectionType2” function in the CampaignService API.