Big Data and Music

A lot of speculation goes one around the Western world about the life-cycles and sustainability of different industries, and the topic du jour seems to be the music industry. This debate is just like any similar debate regarding the arts in that it spans a broad spectrum of concerns: from “how do we maintain artistic integrity?” to “how do we make money?”. Of course, the latter end of the spectrum serves only to deliver the former. I’m sure. Questions that often befall our ears or eyes include: How do we stop illegal file sharing of music? Should we stop illegal file sharing of music? Is music owned by the performer, the writer, the producer, the studio? What about music that’s made experimentally? Who owns that? Can artists survive on measly royalties from online streaming alone?

In my opinion, the music industry is certainly at a cross-roads but this does not necessitate a decline by any means, but more a willingness to adapt to the new ways that humans are consuming, making, sharing and using music. Forgive the bluntness of my phrasing, but if you’re smart then I believe that you can truly make a name for yourself across the entirety of the aforementioned spectrum in the artistic debate.

As Big Data is our own topic of concern, let’s take a look at how big data is impacting the music industry.

When discussing Big Data and the music industry you cannot avoid Spotify. For anyone who has had their eyes, ears and brains directed elsewhere since 2008, Spotify is an online  music streaming service which originated in Sweden. Similar services include Pandora, which offers radio stations based on particular genres, Beats Music, which provides algorithm-based personalization, and to a lesser extent The Hype Machine and SoundCloud, both of which are largely used for discovering new, unsigned, underground music and obscure mixes by niche producers.

This article by Mark Van Rijmenam doesn’t conclude much about the implications of Spotify’s data usage, but it definitely gives some fascinating figures and facts:

Company	Spotify
Established	2008
Active users	24 million
Of which are paying	6 million
Facebook fans	3.7 million
Number of songs	25 million
Songs added daily	20,000
Playlists	Over 1 billion
Royalties paid	Over $500 million
User-created data per day	600 GB
Storage	28 Petabytes
Data Centres	4 worldwide

While simply looking at these facts and figures is astounding, it’s still nothing new. The pioneering stuff that we’re seeing here is really how that data is being used. Simply having all of this data is worth nothing if you can’t do something meaningful with it. Spotify developed their own open-sourced workflow manager called Luigi. They use this information to drive intelligent business decisions about where there product is going and how they should structure or restructure their business model in future. They are actually using the product to become proactive rather than reactive in their business decisions, an option that was entirely foreign in the not-so-distant past.

You may think that’s great for Spotify, but what about the artists. In my opinion, this has leveled the playing field. Like all the great cloud tools, it is a multi-tenancy platform that enables the little guys to reach the same audiences as the big guys. This means that I have spent my money thus far this year on the likes of Electric Picnic to see all of these guys, and a little gig in The Workman’s Club to see All We Are who are quickly becoming my new favourite band. All because I have access to all of them on the one application which only suggests music it knows I’ll like.

What about the integrity of music? Ultimately that is going to be a very subjective topic. Who am I to say that just because you don’t enjoy a certain combination of sounds that I enjoy, that you’re lacking in good taste. But the question becomes so much more when you realise that sometimes the idea of the music starts with a person but the production and even the outcome is entirely out of their hands. Take, for example, a fascinating project between Sue Murphy, an Irish art director and designer living in New York who recently collaborated with James Murphy (no relation) of LCD Soundsystem. They produced The US Open Sessions: Music Made With Tennis Data. James Murphy himself admits that he was not making the music. The music came from the data produced by the tennis matches which was then fed into an algorithm that produced the music. This raises a number of questions about whether this is art, music, or purely a publicity stunt by IBM, the US Open and anyone else involved.

Perhaps my optimism will damage me in the long-run, but personally I feel that big data is launching us into a new era of exciting audio, and while I will never neglect the classics, I can’t help but find my curiosity piqued.

Other interesting links:

https://datadrivendj.com/

https://www.business-cloud.com/articles/news/cern-turns-big-data-music