"Who Runs Spotify? Measuring the Power of Content Owners on Digital Platforms"
Digital platforms connect consumers with content owners or producers (e.g., music labels), mostly on paid subscriptions. To date, platforms replaced traditional revenue streams of content owners across many industries. For example, online streaming now accounts for more than 80% of total revenues in the music industry (2019). The digitization greatly disrupted the market interactions between content producers and consumers through platforms that substantially reduce users’ costs for the discovery of new content. For example, streaming platforms – such as Spotify – are typically “one-stop solutions” for consumers, guiding user attention by their decisions to promote the content of certain music labels on their own-curated playlists, or based on content recommendations on the platform (“economics of attention”). In contrast, the major record labels hold a wealthy pool of popular artists, allowing labels to threaten and finally harm revenue generation on the platform. This paper contributes to ongoing public debates about who is more powerful on digital platforms – the platform itself (Spotify), or the suppliers of digital content (music labels). Using data on more than 1.2m playlists collected from Chartmetric.com, we empirically quantify the strengths of players’ toolkits on Spotify – i.e., content popularity of music producers vs. “attention economy” – and illustrate how power proportions vary across music categories (e.g., genre, mood, activity). Our results provide evidence for a predominant position of music labels on the platform. While our estimates highlight Spotify’s effectiveness in guiding user attention through playlist promotions and recommendations, we find that compositions of the most popular playlists on the platform are systematically biased towards large fractions of major label-content that are not driven by consumer demand. As a consequence, we show that some playlists would even gain incremental listeners if reducing the fraction of major-label content.
*Co-authored with Max J. Pachali (Tilburg University)