What are the differences in the dialer types?

Do you have a documentation on the dialing types with their definitions.

preview
predictive
progressive
automatic

You can find this information here under Type:

Type

This is the type of campaign. There are four different campaign types from which to choose.

  • In campaigns of the Preview type, the calling records are submitted to the agents participating in the campaign. The agents review record information, dial destination numbers, and monitor call progress. Additionally, this type of campaign allows agents to be reserved for a requested callback without indication to the agent.
  • In campaigns of the Progressive type, the system automatically dials numbers from list records according to a fixed (preconfigured) dialing rate, monitors call progress, and connects successful (answered) call attempts to available agents.
  • In campaigns of the Predictive type, the system automatically dials numbers from list records according to currently optimal dialing rates, monitors call progress, and connects successful (answered) call attempts to available agents. The dialing rate is optimized to maintain the desired agent occupancy based on the statistical analysis of outcomes of most recent dialing attempts.
  • In campaigns of the Automatic (IVR) type, the system automatically dials numbers from list records, monitors call progress, and connects successful (answered) call attempts to a prerecorded interactive voice response (IVR) message. Note that because agents are not involved in campaigns of this type, it may require different kind of license and may have to be explicitly enabled for you by your service provider.

Note that the definition of some of the subsequent campaign properties will depend on the selected campaign type.

On the Automatic (IVR) type of campaign, has anyone done a deployment where you play one message if the answer is a voicemail greeting and another message when an answer is an actual person?

Tim, may I suggest using dialing rules based on disposition, and in the case of voice select a scenario to run when answered.