How Spotify’s Algorithm Manages To Find Your Inner Groove

Have you ever wondered how Spotify manages to recommend you that perfect song, playlist or even that ‘daily mix’?

The answer is simple — data. 

Spotify’s algorithm is always finding new ways to understand the kind of music one listens to — from the songs that are always on repeat to the favourite genre that one can’t let go.


Sign up for your weekly dose of what's up in emerging technology.

Not only is the algorithm monitoring the music history but also analyses the reason behind a person listening to a particular song or preferring a particular genre over the other.

Spotify’s algorithm is what sets it apart from other music streaming services. Where other music streaming services make use of paid playlists created by individuals and other communities, Spotify, on the other hand, relies on its algorithm in addition to free human-created playlists.

Download our Mobile App

How Does This Algorithm Work?

As of October 2019, Spotify had over 248 million active users. The algorithm seems to be working very well for the streaming giant among the competitions.

The first thing one notices on Spotify is the home screen, that’s where the recommendations start, and the home screen is governed by an AI system called Bandits for Recommendations as Treatments or simply known as BaRT. BaRT is the real reason why one doesn’t go on searching for an appropriate playlist to listen to Spotify. 

Of course, if one has been on Spotify for some time, it starts recommending songs based on the previous listening activities. But, this is not just is, BaRT also drops in fresh music that it thinks users will like in order to keep them out of the same listening loop.

Now two concepts come into play with BaRT — exploit and explore. The combination of ‘exploit’ and ‘explore’ is the key to Spotify’s recommendation. 

While exploiting, Spotify makes use of every activity produced by a user. Exploit usually makes use of the user’s listening history, songs skipped, playlists the user has created, social media activity on platforms and even one’s location to recommend music. While exploring, Spotify studies the rest of the world. It starts searching for playlists and artists similar to your listening taste and also looking at the popularity of the artists in that area whom you haven’t heard of or any other related works.

By taking all this data that has been collected from you over the decade, or as long as you have been on the Spotify, it presents their user with ‘Decade Wrapped Playlists’, and by taking the data set of different genres one has listened to over the decade will become the ‘Best of the Decade For You’.

The 30-sec rule

BaRT’s success totally depends on whether one listens to the songs recommended in the ‘shelves’ or ‘rows’ on their home screen. Spotify’s algorithm looks at the duration of the time one has spent on a song, and if it is for more than 30 seconds, then the platform takes it as a check on their recommendations. The longer one spends on a song or a playlist, the better their suggestions will get. So, if one doesn’t like the first 28-second of a song, it is better to skip it before the 29-second mark to never get to listen to something like that again!

Recommending New Artists by Analysing Audio

Recommending music liked by other people or from the data received from millions of music blogs which can be called collaborative filtering doesn’t seem to work in the case of a new artist. In that case, how do you think one gets a recommendation on a song that has just been released, that too by a new artist? The solution to this problem is Spotify analysing the audio itself, and training the audio analysis algorithm to learn to recognise different desirable characters to music. Spotify experiments could identify various aspects of songs like distorted guitars, and while others recognised more abstract ideas like genres.

Recommendations in Automatic Playlist Continuation

It is obviously has been observed that Spotify plays a song automatically after your current playlist ends, and this happens because of artificial intelligence too. The feature analyses the songs in your playlists and tries to predict the music that can be played next after your song finishes. 

The platform has taken this feature to the next step by releasing a ‘Million Playlist Dataset’ of user-generated Spotify playlists. This vast data set was released to help understand the behaviour of users on Spotify. The auto continuation of the playlist becomes a problem when one doesn’t get appropriate songs after the existing playlist finishes, so in order to solve this problem, the company invited AI researchers all around the world to present their solutions at The RecSys Challenge 2018.

Recommendations Using Locations

The data provided by users about their age and gender along with their locations are also used by Spotify in order to study whether a user’s taste of music changes after the individual moves to a different state. Along with that, how the age of the listener impacts the kind of music the person listens to.

“Every word anyone utters on the internet about music goes through our systems that look for descriptive terms, noun phrases, and other text,” Brian Whitman wrote, CFO The Echo Nest (Acquired by Spotify in 2014)

By studying the kind of music people listen to in particular areas and then putting this data against the group of people who have recently moved to a different area, is a move done by Spotify. With different overall trends, the Spotify team concluded that over some time, location does factor in when it comes to moving to a different area in a certain way.


With Spotify’s recent struggles and by struggles I mean, the stock hit that it took last year because of the emerging competitions like, Alphabet and Apple, it is continuously looking out for different areas to improve or trying to enhance the existing features. Making features like Discover Weekly more personalised is one of the areas where people probably don’t mind sharing their ‘musical’ data with the streaming platform. 

Spotify is also taking bold steps to improve it’s ‘Behind the Music’ feature, where it is looking to integrate real-time lyrics, and sort of make it karaoke-ish. It also plans on including a feature which can detect covers of a song, although it would prove challenging especially with a genre like Jazz. 

With other competitions like Youtube Music, Amazon Prime Music, direct competition juggernaut Apple music, and the other streaming platforms, Spotify is still keeping its top position intact, and the biggest reason is its algorithm.

More Great AIM Stories

Sameer Balaganur
Sameer is an aspiring Content Writer. Occasionally writes poems, loves food and is head over heels with Basketball.

AIM Upcoming Events

Early Bird Passes expire on 3rd Feb

Conference, in-person (Bangalore)
Rising 2023 | Women in Tech Conference
16-17th Mar, 2023

Conference, in-person (Bangalore)
Data Engineering Summit (DES) 2023
27-28th Apr, 2023

3 Ways to Join our Community

Telegram group

Discover special offers, top stories, upcoming events, and more.

Discord Server

Stay Connected with a larger ecosystem of data science and ML Professionals

Subscribe to our Daily newsletter

Get our daily awesome stories & videos in your inbox

Do machines feel pain?

Scientists worldwide have been finding ways to bring a sense of awareness to robots, including feeling pain, reacting to it, and withstanding harsh operating conditions.

IT professionals and DevOps say no to low-code

The obsession with low-code is led by its drag-and-drop interface, which saves a lot of time. In low-code, every single process is shown visually with the help of a graphical interface that makes everything easier to understand.

Neuralink elon musk

What could go wrong with Neuralink?

While the broad aim of developing such a BCI is to allow humans to be competitive with AI, Musk wants Neuralink to solve immediate problems like the treatment of Parkinson’s disease and brain ailments.

Understanding cybersecurity from machine learning POV 

Today, companies depend more on digitalisation and Internet-of-Things (IoT) after various security issues like unauthorised access, malware attack, zero-day attack, data breach, denial of service (DoS), social engineering or phishing surfaced at a significant rate.