Understanding User Behavior on Online Music Distribution Sites: A Discourse Approach

We analyze user behavior on two music distribution websites, SoundCloud.com and Last.fm, using a computer-mediated discourse analysis approach. The broad goal is to infer patterns of use that can inform the development of system/site designs to facilitate communication among online users. The specific goal is to analyze the actions users perform through comments posted to the sites. Based on a manually-coded sample of 49,324 comments, we address two research questions: What are the most common speech acts used in comments on the music distribution sites, and how, if at all, do design differences – between the two sites and between the two commenting modes available on SoundCloud – affect speech act use? From these comparisons, we identify site purpose (SoundCloud as a music sharing site, Last.fm as a music streaming site) and commenting mode (regular vs. timed comments) as important factors that contribute to the characteristics of user comments, and suggest implications of these findings for the design of music distribution sites.