A Genre-Based Analysis of New Music Streaming at Scale

Abstract

The rise of on-demand music streaming platforms and novel recommendation algorithms have brought a transformative shift in music listening, where users have an effectively endless supply of new music to discover. This study aims to understand novel music streaming patterns, operationalizing novel music as new music releases that are new to everyone, from the perspective of genres. Leveraging tracks, users, and streaming data from Spotify, we empirically analyze streaming patterns of 282K new music releases. We find that new releases in genres that often serve functional purposes, such as classical music for relaxation, are consumed less. Surprisingly, new releases in Pop generally do not exhibit high consumption rates despite being characterized as popular music. From examining 1 million users’ historical genre preferences, we observe that users’ new release preferences are distinct from their overall music preferences, although past genre affinities and proclivity for newer content can predict new music consumption. Our findings present important implications for new music recommendation strategies on music streaming platforms and broader contributions to understanding the dynamics of music discovery.