Disclosed herein are systems, methods, and non-transitory computer-readable storage media for broadcasting audio tweets. A system broadcasting audio tweets receives tweets via telephone devices, wherein each listener hears a telephone call of a broadcast on the telephone devices. Each received tweet is associated with one or more tags specifying tweet content, which is used to analyze received tweets. The system presents a display of analyzed tweets for selection to a host. The host selects one or more tweets and broadcasts the selected tweets as part of the broadcast to the telephone devices. Automatic speech recognition can be used to automatically convert received audio tweets to text to enable tagging and analyzing the tweets.