Abstract: A software-implemented method for delivering advertising presentations in a wide area network (WAN) based on Voice Recognition analysis is described. Data captured through Voice Recognition and other analysis is used to determine which calls contain “success events.” These “success events” are typically sales or other forms of revenue-generating end-user actions. Calls which include “success events,” are grouped into a database, and data in that database is used to determine optimal placement of forthcoming advertising presentations, to maximize the yield of “success events” per every 1,000 advertising presentations shown to end-users. The result is a continuously optimizing method for advertising display across a portfolio of available advertising choices, tied to “success events” identified in the calls themselves.