Content area
Full Text
Music streaming services like Spotify have changed the way consumers listen to music. Understanding what attributes make certain songs trendy can help services to create a better customer experience as well as more effective marketing efforts. We performed cluster analysis on Top 100 Trending Spotify Song of 2017 and 2018, using nine musical attributes, including danceability, energy, loudness, speechiness, acousticness, instrumentalness, liveness, valence, and tempo. The results show that music structures with high danceability and low instrumentalness increase the popularity of a song and lead them to charttopping success.
Keywords: cluster analysis, music, Spotify
INTRODUCTION
Music streaming services have revolutionized the way consumers listen to music, not only by lowering the costs but also by providing consumers with an endless library of artists from all genres and musical backgrounds. As of July 2019, Spotify, the leading music streaming service, provides access to over 50 million tracks to 232 million monthly active users, including 108 million paying subscribers. Spotify's payment model structures around a $5 monthly subscription fee that provides a user with unlimited, advertising-free experience. For an additional $5, users receive premium features including offline listening, a mobile app, enhanced sound quality, exclusive content, early album releases, and sound system compatibility.
In recent years, Spotify has allowed users to discover music and create exclusive playlists based on their musical preferences, favorite genres and artists, and even mood. This design has helped in eliminating a potential struggle for users in searching for an extensive database of millions of songs. To optimize such discovery and personalization, streaming services like Spotify not only rely heavily on recommender systems but also on human editors. A deeper understanding of the characteristics and use of playlists and how users create and maintain their playlists can contribute to better recommendations.
As these playlists become more customized based on Spotify's recommendations, certain songs begin to recurrently appear on "Top Song" lists resulting in their trending on the platform. For each song, Spotify provides audio features such as duration, key, and mode. This study intends to investigate whether the success of the trending songs is related to these attributes. The results would allow music streaming services to create better-customized playlists that reduce search time and improve the satisfaction of their users. The findings...