We are witnessing the constant and ongoing change of the music industry today. We consume songs quite differently than before; we take our music everywhere thanks to the portable speakers, iPod’s, smartphones and so on. Also, the way artists record music and the different ways they release it has drastically changed as well. Another thing that changes equally in today’s business world is big data. When these two unite, some great conclusions come to our mind.
One such conclusion is that the music industry takes care of what people are listening. This was not the case in the past. The music industry also takes care of where and when people are listening t music, which formats they use the most and so on. Spotify is a company which discovered that one of its listeners listened to Justin Bieber’s song Sorry for more than 40 times. This certainly made them ask what did the poor guy do?
Remember the last time you heard a song and automatically started nodding your head. Don’t think that happens by accident. Big data today can even predict which song will become popular. For example, at the University of Antwerp, Belgium, developed an algorithm which managed to calculate which song will be a hit that year. During the research, a selection of dance songs was taken in the period between 1985 to 2014, plus some hits from 2015.
According to the algorithm, there was a 65% for a hit to reach Top 10. And when it calculated the dance songs already in Top 10, each song had more than 65% probability.
Remember the first time you heard your favorite song? You may remember the place you heard it for the first time, or what were you doing at that moment. Back in the past, finding a new song was either by hearing it accidentally on the radio, or some of your friends would suggest you to hear it. Of course, this still happens, but big data found its way into this as well.
You may have heard about the Pandora website and the Music Genome Project . They use trained musical analysts and they listen to songs and analyze them based on 450 features. By doing this they capture the true identity of the song, they can classify the songs by their similarity and what some people may like depending on their favorite artist.
In the meantime, Spotify acquired The Echo Nest. It is more automated than the Music Genome Project but it still classifies music according to tempo, danceability, tones, and tempo and so on. Additionally, it searches the web for data about artists and tracks and continues with the analysis. Spotify uses big data broadly and even uses its database to predict who will win the Grammy Awards. The best thing is that out of six predictions, four are correct.
Another huge company which uses big data extensively is Amazon Prime, specifically Amazon Music Unlimited. It uses big data to offer its user personalized music choices, using the algorithm to create playlists and works on many different devices. Big data will also expand with the use of Alexa Voice Service and using voice commands to play music. This and similar services will provide more personalized results to its users.
According to James Longman from Audioreputation, seeing where big data is being used in the music industry is exciting and it will develop even more in the future. If they want to stay competitive they can’t do it without big data. Matter of fact, most companies would disappear without big data and cloud-based technologies. If you add social media to the big data analysis and cloud-based technology, plus AI and machine learning, you can’t even imagine in which way the music industry will evolve. The existing algorithms will be developed and perfected further, and with the increasing number of online artists, new singers will be discovered in a second. So if you don’t believe in fairy tales, make sure to believe in big data analytics.
