We don’t know if you know but a few years ago, the major music labels had pretty sweet business. They profited from any single or album, not matter it was vinyl, CD or even cassette. It is not like this anymore. MP3, torrents, piracy, streaming, and video sharing networks killed this particular business model. What music labels did to stay afloat? They asked data analytics for some help. And they were not alone – a lot of businesses emerged and shaped their models by applying data analytics to music services. We are going to explain this.
Not so long ago, the music editors & producers working in radio broadcasting had a huge amount of classification work BEFORE any song got in radio playlists, hence charts. Such classification work is including genre, tempo, period, artist gender, music instruments, instrumentals, and much more. The modern radio broadcasting business still works like this but when music streaming services appeared, they had to fill a huge gap.
Musical Genome Project tried to make it by structuring music data by manual classification. Specially trained musicians studied each track, just like Netflix employs people to watch films and classify their content. Pandora developed this project and now you know how the first effectively marketed music recommendation engine was born. By collection up to 450 data points for every song, data analytics algorithm is capable of predicting which songs you rather enjoy to listen next.
Since music is unstructured data, machine learning is applied in order to get classification that matters. Combine it with predictive analytics and you`ll have the perfect music recommendation tool. Pandora was the first service like this, Last.fm become even better along with iTunes but Spotify has mastered this out with careful examination of what users listen, when, for how long and how often. Meanwhile, Spotify acquired the music analytics company The Echo Nest to beef up its analytics potential. This is a huge amount of data but proper data work is giving results that keep users logged in and submitting playlists. All of these relatively new businesses are possible not just because of the Internet but data analytics.
Let us get back to music labels. Since the records sales are highly compromised these days, one of the main profit sources are the live gigs. Jay-Z, for instance, arranged his UK tour locations based on Spotify data because the streaming service knows pretty good how many fans he has and where they live. This is how data analytics applied in music gives pure optimization efficiency resulting in improved sales.
It is not just this. Music labels are keen to know which band has rock star potential and which single has hit potential. This is where predictive analytics comes to help and turns the art of discovering the right artist into science. MusicMetric is analytics company has tracked the journey of thousands of artists from start to epic success. This way, their team spotted data signals distinguishing regular guitar heroes from future rock stars. On top of this, MusicMetric expanded their data journey to social media and their experts can identify who is buzzing the networks in most promising and right way. Such insights are helping music labels and stream services who has potential and which band is worthy to keep an eye on. As a result, MusicMetric were acquired by Apple.
Researchers from the University of Antwerp created an algorithm that was able to predict with relatively high accuracy the position that dance singles would get at the Billboard Dance chart. By analyzing all of the songs which made it on the chart between 1985 and 2014, the algorithm predicted which record will make it into top 10 with an accuracy of 65%. Such insight is worthy to keep in mind if you are in this business. Later, Spotify data team made an experiment with predicting the Grammy winners. Their data analytics solution guessed right 4 out of 6 winners. Not bad at all.
This is how data analytics is not just giving extra value and is improving efficiency into already existing business. It is about creating new services and business models, getting key insights and fitting in a better way into customer needs.
