A pioneering predictive analytics platform has its finger on the pulse of the music industry
Just two short decades ago, the recorded music industry derived almost its entire global revenue from physical media including LPs and CDs. Fast forward to 2020 and music streaming made up the biggest slice of the pie by far, demonstrating tremendous growth over the last several years, overtaking and far surpassing downloads and other digital forms. Over 60 000 tracks are uploaded every day on Spotify – almost one per second. Physical media produces less than a third of the revenue brought in by streaming. A small portion is also provided by performance rights associated with live events and by use in movies or video games. Against this mind-boggling landscape of ways in which music is produced, distributed, and used, the EU-funded FuturePulse project has delivered advanced multimodal predictive analytics and recommendation services. These were developed-hand-in-hand with relevant music industry players to maximise the industry’s impact and income.
A rhapsody of musical accounting genius
The FuturePulse project targeted three key areas: record labels, live music, and background music. This software-as-a-service gathers and indexes music usage data from a plethora of sources. These include digital service providers such as online stores and streaming services as well as music charts, radio, TV, live venues, and social media activity. State-of-the-art automatic extraction tools for audio analysis deliver audio descriptors of music including mood and genre. Popularity metrics are computed for tracks, artists, and genres according to their activity in the different services, and algorithms predict and forecast future trends. FuturePulse’s highly intuitive and attractive web-based user interface makes it all accessible to music industry professionals.
Pioneering data-based forecasts identify winning revenue streams
The benefits to the three use cases are myriad. For example, record labels can easily assess the past performance of their current artists and predict future potential. Insight can help support events planning, allocating marketing budgets more effectively and scouting new unsigned talent. Concert promoters, festival organisers and artist managers who conventionally make decisions based on instinct will now gain data-based insight. For example, FuturePulse can shed scientific light on local trends in musical tastes. A popularity metric specially designed for the live music use case can help users identify the most relevant artists and venues at the city level based on past shows. Background music platforms will gain advanced tools for automating the curation of playlists to improve their service for stores and public venues. According to Gonçal Calvo, head of innovation at BMAT Music Innovators, “Most actors in the music industry are dependent on predictions – whether a track or an artist will become popular, or whether a festival or concert will sell enough tickets. Predictive analytics in the music industry is in its infancy and the potential is huge.” The music industry is seeking answers, not just more data. Understanding the context is as important as the tools the platform offers. FuturePulse is a flexible framework that can be tailored to meet different needs. Its predictive algorithms can be used far beyond the music sector. With its finger on the pulse of modern data analytics, FuturePulse’s future could not be brighter.
Keywords
FuturePulse, music industry, predictive analytics, streaming, musical, popularity metric, record labels, forecast, software-as-a-service
