This Thursday at 1:00 pm Eastern (18:00 British/10:00 Pacific) we will be having our first symposium and workshop, to illustrate a few of the practical uses of music industry data in a variety of disciplines.
Guest speaker Andrew Ullah will present “Deconstructing the Billboard Hot 100,” a study of song attributes in the top 40s and how they correlate with commercial success, and Mathew Flynn will present a pedagogy of music industry data in MA Music Industry curricula. Please read their abstracts below.
Katie Straker (PhD candidate in Musicology, The Graduate Center, CUNY) will give a short presentation on industry data as a tool for developing theses (rather than merely empirically supporting them) in musicology, then lead an hour-long practical workshop to help scholars understand both the technical nuts and bolts of how to navigate MusicID’s interface and how to use the platform to find data in support of various applications.
Registration is free, and attendees will receive complimentary access to MusicID until July 1. Register on Eventbrite. We encourage attendees to log in to the platform and follow along during the workshop, so that you will best be able to ask questions as we go and obtain guidance in how to use MusicID in your unique research.
Andrew Ullah (USC, Marshall School of Business) – “Deconstructing the Billboard Hot 100”
In collaboration with other MBA students at USC, I analyzed Billboard Hot 100 data to build predictive models informing the likelihood of a song’s popularity using chart positions as our benchmark of success. Acoustic song data was gathered from Spotify’s open API to model song attributes such as tempo, key, duration and danceability, which were then joined with revenue data from MusicID Revenue to forecast the song’s expected earnings. Statistical models such as regression and decision trees were explored to retrieve the insights. Where it is widely known that attention spans are decreasing year over year with the ample amounts of competition of media content substitutes for music, shorter songs also maximize on the royalty payout model in streaming where 30 seconds equates to a unit of sale. Meaning, there is no marginal benefit for music creators when listeners listen to their songs beyond 30 seconds. This creates the opportunity for the composition of songs to be incredibly sensitive to the consumption platform as seen in Saweetie’s hit song, “Tap In,” with just over two minutes in duration and a large level of dance engagement on TikTok. This song moreover features a prominent sample within the first 30 seconds, to ensure listeners engage long enough to yield a unit of sale. Further song attributes need to be explored such as melody, chord progression to offer additional insights, but from our study duration, consumption platform, and royalty payout models are all core criteria that need to be considered when developing and collecting on recorded music within the digital ecosystem of today.
Mathew Flynn (University of Liverpool, lecturer in Music Industry) – “Challenges of Teaching and Using Data in Music Industries Curricula”
My research and teaching at The University of Liverpool focuses on music makers’ and music industry practitioners’ decision-making across numerous creative and commercial contexts. Increasingly big data is a key input into to strategic decision-making within and across music companies, and for music makers developing careers. Equally how data is stored, coded, accessed, managed, shared and used brings with it its own decision challenges. Students pursuing careers in the music industries need to prepare to work with data and understand its value in terms of making more informed choices. However, due to the proprietary nature of most industry datasets, this is a challenging subject to teach. Over the last year I have been developing an active role for data in the MA Music Industries curriculum I currently deliver, and this presentation will demonstrate the early stages of how I have attempted to do this, the challenges encountered, and progress to date.