Music by Analytics: Is Big Data Killing Creativity?
Last year saw a record number of people turn to digital streaming to get their musical fix, with 20 million subscribers signing up to Spotify by June 2015. Music streaming gives users access to sea of new music for the price of a CD, it’s no wonder music streaming services have gained immense popularity.
From a consumer’s perspective, it’s a real-world musical utopia. Streaming features are not the only way music streaming is revolutionising the music industry; it has also opened the doors for big data and music analytics. While music analytics possesses immense potential to facilitate growth in the industry, it may also have an adverse effect; fueling a culture of engineered production that threatens musical creativity. To understand how this relationship plays out, it’s essential to understand how analytics is changing the face of the modern music industry.
Music analytics is a relatively new industry that uses data collected from digital platforms such as music streaming services to uncover how music is consumed and how to predict future trends. The past few years have seen a frenzy of companies, including Pandora and Spotify, buying up Analytics companies such as Next Big Sound and The Echo Nest, respectively. Despite it still being early days, music analytics are now applied to a range of applications, from when to release albums, to setting tour routes.
Before analytic companies came to the forefront, the industry’s methods for predicting musical trends remained relatively the same. Talent scouts scoped the next big sound and radio charts measured the popularity of individual tracks. The ultimate evaluation for the popularity of music in past decades has undoubtedly been the album sales metric. This metric, however, fails to capture any data past the point of purchase. Album sales tell us a lot about what music people are buying, but nothing of what is being shared around by friends, made into mix tapes, which songs are on repeat and which are skipped.
Music streaming lifted this veil and provided new insight on how, when, and where music is consumed, shared and talked about. Record companies now have unparalleled and quantifiable data on what millions of people are listening to all over the world. Music analytics keeps a finger on the pulse of the music industry, but by doing so it may be killing its heart; for wherever there are numbers, there are also formulas.
The pursuit of big hit records with broad appeal has inevitably brought an element of homogeneity to the songs we hear on the radio. Research suggests that the last 50 years of Western pop music has varied very little from the ten most popular chords. Although our musical sensibilities are geared towards the same old patterns, recent years have seen a darker turn for the trend.
In his book, “The Song Machine”, New Yorker staff writer John Seabrook chronicles his years around the heavyweights of pop music production. Rather than reminiscing of the rock n’ roll lifestyle, Seabrook instead focuses on the shift towards an ‘assembly line’-like method of production that is responsible for much of today’s music on the radio. According to Seabrook, unlike the hits strummed out by band members on Abbey Road, today’s melodies are precision engineered en masse to stick in our heads.
It might be surprising to learn that for the past decade, the majority of top 40 songs are written by just four powerhouse producers; Max Martin, Stargate, Ester Dean, and the controversial Dr Luke. Haven’t heard of any of them? Well, if you’re a fan of Taylor Swift, Adele, Katy Perry, The Weeknd, Pink, Justin Bieber, Rihanna, Drake, Maroon 5, Nicki Minaj, Ke$sha, or even the Back Street Boys, chances are your favourite songs are among their work. These four producers have perfected the science of making catchy beats, to the extent that Seabrook estimates that one in every ten they make go on to become a hit. This recipe for success has lead to an all-out domination of the airways, with the same hit songs getting the majority of available airtime while forcing other songs out of rotation. The financial impact is that 10% of songs released generate 90% of the revenue; making it increasingly hard for emerging artists and producers to compete.
Using data from 75 million + music streaming listeners to improve the efficiency of the song machine record companies will undoubtedly tighten their grip on the charts. Which begs the question, is this really such a bad thing? Easily digestible songs with a catchy beat and lyrics everyone can sing to have always been a powerful factor in the unifying nature of music.
While it may be harder for independent artists to reach mainstream success without embracing the mechanisms of the hit factory, streaming services have created a space where artists can reach audiences anywhere in the world. Music analytics, too, has the power to benefit niche musical markets, uncovering the value of previously hidden segments and making it easier for artists to gain recognition.
In short, the further homogenisation of mainstream music does not necessarily spell doom for musical creativity; as long as music streaming and music analytics continue to generate opportunities for fringe artists and genres. Unlike radio, streaming services are giving listeners the chance to explore their musical tastes, and that can only be good for musical diversity. Music analytics’, like all big data programs, objective is to uncover what markets actually want. As long as there are those who want a different sound, there is no reason musical creativity and the song machine cannot co-exist.