Spotify Algorithms

  

Spotify Algorithms

Spotify is a music platform with over 190 million users. Once a week, users receive two playlists in the discover section of the music browsing function, one of which informs people about newly released music from artists that they listen to, including a few new recommendations based on an algorithm, and the other suggesting new music that the user may like based upon what they frequently grace their ears with.

The results are surprisingly accurate, with many users shocked at how well their music platform understands their listening needs. While this may seem like voodoo, the results are achieved using digital algorithms which are constantly being refined to be as accurate as possible. With so many music streaming platforms available, all of which have access to most of the music out there, Spotify has had to do something to stay ahead of the competition and to set them apart from the rest. This is why this algorithm is so important for retaining their paying customers.
The program was initiated in June 2015 and grew quickly in popularity, so much so that when there was a glitch in the server which delayed the release of the weekly playlist in September 2015, many users expressed their dissatisfaction on social media. The brains behind the algorithm boast that it is also beneficial to the musicians putting their media onto the database, claiming that they have the technology to find the twenty users who would enjoy the most diverse of compositions out of the millions of users who rely on the music service.
The ingredients in this digital concoction begin to make sense, the more you dive into the subject matter. Spotify begins by pooling data from other user's playlists. Those users who have some sort of a crossover with your own music taste are likely to enjoy similar pieces of music, and by pooling large amounts of data together, this can be refined down with unfathomable accuracy. Playlists from all levels are entered into the algorithm, from celebrity users to the playlists of your hairdresser down the road. Naturally, more popular playlists have more importance in the algorithm. To put it simply, if you share two songs with another playlist, but a third song is on the other playlist which you haven't heard, Spotify will recommend the third song. The reality is more complex than this, but it should give you an idea of how your music platform understands you so well.
To make it even more accurate, Spotify creates a digital profile of your individual music taste, refined into not just sub-genres but micro-genres, the names of which you may well have not heard of. Using this and adding the element of shared interests with other users, Spotify uses the open-source software Kafka to bring both elements together, combining this with collaborative filtering, and produce a list of songs that you may like and that you haven't listened to yet.
This opens up the potential for artists and labels to bribe Spotify so that their songs appear on the frequently played discover playlists. A spokesperson for Spotify claims that, despite having many requests, this is not the case, although sometimes it seems that songs can pop up in multiple playlists all at the same time.
Of course the picks aren't perfect, and often while a user might really enjoy one or two songs and like a few more, there will be a few on each discover weekly playlist that the user will not enjoy. The fine tune your weekly playlists, you can make extra effort to add songs that you do like to playlists in your own library. You should also skip the songs you don't like, as if songs are skipped in the first thirty seconds, the algorithm counts this as a rejection and will include this data for the next playlist. Exploring outside of the mainstream music and “going down the rabbit hole” is also noticed by the algorithm, and the more you explore micro-genres, the more this will influence what comes up on your playlist. The result is that the music platform will be providing more accurate music for you to listen to every week, and you will be supporting the smaller independent artists who rely on music streaming to provide an income.

 

up Top