A system that is capable of receiving and interpreting a user's spoken commands is also coupled to a television interface device that controls video programming content displayed the user's television. The user can speak commands to cause certain programs to be played, and to control various functions of the television. The system can track and analyze a user's video viewing habits. In addition, the system is able to determine which of multiple users might be interested in a particular video program, and the system can play marketing messages regarding that video program to those users that might be interested in viewing the video program.