Answering machine detection can be a key part of your campaign. It filters out the answering machines and only connects a live call to the agent. Answering machine detection works by calculating the numbers of words and the amount of silence between them when the call is first answered. This method is not 100% effective but can detect almost all standard answering machine messages.
Notes on Answering Machine Detection
- AMD can only be used in Ratio/Predictive dialing modes.
- AMD will not detect 100% of the answering machines.
- Calls will take longer to get to agents because the system listens to the first few seconds of the call to determine if it is an Answering Machine or a live person.
Here are the steps to enable this feature:
From the Dashboard, Go to Call Center and select Campaigns
Under Show Campaigns, Click on the Action Button and select Settings
Click the Call Routing & Analysis tab.
Scroll down to Call Analysis.
Under Detect Answering Machine & Fax, choose either Advanced or Disable from the drop-down menu
- Advanced mode is about 60% to 70% accurate. Fewer answering machines will get through to your agents but you run the risk of a live call being flagged as an answering machine.
- Detection Background Noise - This is the audio people will hear while Answering Machine Detection is in progress. The default sound is a crackling noise with a cough to give the person the impression that a live person is on the other line. This reduces the chance of dead air people hear during detection, which increases connection rate and decreases hangups
- Detection Background Noise - This is the audio people will hear while Answering Machine Detection is in progress. The default sound is a crackling noise with a cough to give the person the impression that a live person is on the other line. This reduces the chance of dead air people hear during detection, which increases connection rate and decreases hangups
- Disable mode - the system will not filter Answering Machine Calls. But the Admin can still customize how the identify the call based on the time (Silence after greeting and the maximum length of initial greeting)
Note: We removed the "Standard" from the dropdown option. This change addresses a key issue that our customer base as a whole has brought up -- the number of answering machines that agents continue to receive despite turning on AMD. By changing the configuration, you will capture more answering machines that leak through.
This will reduce the amount of time your agent spends dispositioning calls that went to voicemail. In removing the Standard option, we remove the pre-configured options and logic of standard. This allows clients to determine what configurations work best for their needs instead of trying to manage call flow solely as a function of dial level.
If you would like to configure your Answering Machine Detection much like the Standard version, please click this link (How to set up Advanced AMD to function like Standard AMD)
For Advanced Mode:
Silence After Greeting (milliseconds): Allow the Admin to customize the Answering Machine detection logic to allow them to change it based on their preference. This value represents milliseconds, 500 = 0.5 seconds, 1000 = 1 second. If exceeded, the call is treated as a human. Lowering this will result in connecting more humans in exchange for more answering machines getting leaked through. The same applies vs versa. Optimal settings are between 600 to 900. Increasing this value will also increase the time between detection and calls connecting to agents.
Maximum Length of Initial Greeting (milliseconds): This value represents milliseconds. 500 = 0.5 seconds, 1000 = 1 second. This sets the maximum length of a greeting. If exceeded, the call is treated as an answering machine. The optimal setting is 1100. Increasing this will result in connecting more humans in exchange for more answering machines getting leaked through.
FAS Detection: Calls that have detected FAS (False Answer Supervision) will either have a disposition as FAS Detected [Disposition: FASD] or be routed to an agent [Route to Agent].
Silence Detection: Calls that are silent after the initial 2-second duration will have a disposition as Answering Machine [Treat as Answering Machine] or be routed to an agent [Route to Agent].
Skip Call Analysis On First Call: If this is set to Yes, the system will disable Answering Machine Detection for New Leads.
Leave Message when Answering Machine is detected: Voicemail Drop is Disabled by default.
Note: If you want to add this feature, please contact support.
Message Type - We have 3 Types of messages
- Basic Voicemail Drop - if this is selected, the system will always play the audio file selected in Message Played field.
- Advanced Voicemail Drop - if this is selected, the system will play a specific Audio File depending on the call count of a lead.
In the example above, if the system detects a Voicemail and the Leads call count is between 0 to 5, the campaign will play the "Broadcast Message1" Audio File.
if the system detects a Voicemail and the Leads call count is between 6 to 10, the campaign will play the "Please leave a message" Audio File.
- Smart Voicemail - if this is selected, you can choose to play different audio depending on how many times you have left a voicemail.
If for example you don't want the voicemail logic to start until the 3rd call, or you want it to stop after the 10th call, then you can add Call Conditions. This gives you the ultimate flexibility for your voicemail strategy.
Click on Apply Changes to save settings.
Note: This applies to the Conversational Campaign and Survey Broadcast Campaign.
See Also: