Consumer satisfaction with new content discovery through algorithmic and human recommendations

Empirical studies on digital platforms, such as streaming music services, often rely on aggregate data, which overlooks significant questions like how users discover items that align with their preferences. This research utilises a unique dataset that tracks the daily consumption of 9,778 random premium subscribers of a major European music streaming platform, who found 4,136 distinct new songs via algorithmic recommendations or human-curated playlists. Using the number of repeat organic streams as an indicator of preference matching, we demonstrate that algorithmic recommendations generally enhance song discovery, particularly for tracks released by lesser-known artists. Conversely, human recommendations perform better when introducing songs from artists unfamiliar to the user. While algorithmic recommendations can, in specific contexts, create imbalances and skew sales distribution, they ultimately assist consumers in finding items that suit their preferences in streaming services.