A system and method are described for delivering to a member of an audience supplemental information related to presented media content. Media content is associated with media metadata that identifies active content elements in the media content and supported intents associated with those content elements. A member of an audience may submit input related to an active content element. The audience input is compared to media metadata to determine whether supplemental information can be identified that would be appropriate to deliver to the audience member based on that person's input. In some implementations, audience input includes audio data of an audience's spoken input regarding the media content.