Wednesday, March 29, 2017
Home » Big Data » The Convergence of Big Data and the Music Industry

The Convergence of Big Data and the Music Industry

The music industry is constantly changing. The way we consume songs have evolved, with portable players and phones replacing compact discs (meanwhile, there’s been a resurgence in vinyl), and how artists record and release music has also evolved along with the industry. Something else that’s constantly revolutionizing the way we handle business is big data. When the two converge, we can draw some fascinating conclusions.

For one, the music industry has more of a pulse on what people are listening to than ever before. And not just the what, but the where, when, and in what format. Spotify had a fun campaign last year where they crunched the data of their listeners to creative interactive billboards. For example, they discovered one listener listened to the Justin Bieber song “Sorry” 42 times on Valentine’s Day, prompting the company to ask “what did you do?”

Have you ever heard a song on the radio and started subconsciously nodding your head or tapping your toes? That’s not an accident. Big data can be used to predict what songs will be popular. In fact, data scientists from the University of Antwerp in Belgium developed an algorithm that was successfully able to forecast whether a song would be a hit. The research looked at dance songs from 1985 to 2014, though they also ran the hits from 2015 through their tool. The algorithm predicted a 65 percent or higher probability of a hit for the entire Top 10, and over 70 percent probability for 6 out of 10 songs. When it calculated the dance songs on the Top 10 of the UK Official charts, each song received a 68 percent or higher probability, with the Pitbull/Ke$ha duet “Timber” getting a 90 percent. It’s goin’ down, indeed.

“Do You Believe In Magic?” by the Lovin’ Spoonful is about the power of music to supply happiness and freedom, both to those making it and those consuming it. Just think about the first time you heard one of your favorite songs – you may recall exactly where you were or what you were doing. In the past, discovering a new song was a matter of hearing it on the radio or having a friend or family member suggesting you check it out. That’s certainly still a way to find new music, but big data has taken it a lot further.

The “Music Genome Project” on the website Pandora uses trained musical analysts to listen to songs and dissect them, using up to 450 characteristics. According to the site, “these attributes capture not only the musical identity of a song, but also the many significant qualities that are relevant to understanding the musical preferences of listeners.” In essence, big data is able to classify which songs are similar and thus may be appealing to fans of a certain artist.

Meanwhile, Spotify recently acquired MIT Media Lab spinout The Echo Nest. While it similarly classifies music based on a number of characteristics – such as tempo, time signature, danceability, and tone color – it’s much more automated. In addition to utilizing algorithms, it crawls the web to find data on artists and tracks, which it then analyzes. Spotify has used big data extensively, including using the database info to predict the winner of the Grammy Awards by analyzing streaming data with 4 out of 6 predictions turning out correctly.

Amazon Prime and its service Amazon Music Unlimited leverages big data to give consumers a myriad of personalized music choices along with algorithmic playlist creation and provides this platform on many devices. With the incorporation of voice commands for playing music via the Alexa Voice Service, additional data points will be available for the service to offer consumers more personalization using big data analytics.

It’s exciting to see where big data has taken the industry, but I don’t think it’s done just yet. It certainly seems that a big data strategy is necessary to be competitive in the music industry and many of these companies would not even exist if it weren’t for big data analytics and cloud-based technologies. The increased use of cloud-based services with near real-time big data analytics and the likely incorporation of additional data feeds such as social media patterns which, when coupled with machine learning, will continue to evolve the technology used in the music industry. These algorithms will continue to be perfected, and with more artists on the Internet than ever before, it will help new singers and bands get discovered. That’s a good thing. You may not believe in magic, but believe in the power of big data.

About David Lucky

David Lucky
As Datapipe’s Director of Product Management, David has unique insight into the latest product developments for private, public, and hybrid cloud platforms and a keen understanding of industry trends and their impact on business development. David writes about a wide variety of topics including security and compliance, AWS, Microsoft, and business strategy.

Check Also

Here’s How Big Data Can Help You Fill Out Your March Madness Bracket

The month of March brings plenty to get excited about: the end of winter, flowers blooming, warmer weather, and of course, the NCAA Tournament. Every year, millions of people across the country fill out their brackets in their quest for bragging rights over their friends and family. If you want to win your pool, big data can help.