There are a number of way to do that, depending on the nature of the message or announcement. They all use the "Wait for" action in the Action Management. They are:-
1. Wait for tone (if the message is a tone prompt).
2. Wait for energy (If the message is not a tone. e.g. a speech. The system does not care what exactly the speech is. It just tests for the presence of voice energy)
3. Wait for speech matching (here user must have the wave file already saved in the directory, so that it will be used to compare with the incoming speech.)