Spotify and Big Data: Algorithms and Data Processing Methods

By: Ermira Salihu, Ilir Henci, Lorik Muçolli, Mohit Shrestha


Spotify is a Swedish music streaming service with 100 million active users as of June 2016 (Statista, 2016). It provides copyrighted music and podcasts from music labels as well as independent content creators. With over 30 million songs in its catalog, Spotify is one of the biggest streaming services in the world today. As an online virtual service, its services are widely spread in  most geographical areas and has little to no geographical limits. One of the core features of Spotify allows you to create playlists that can be made public to the entire world for other Spotify users to consume. It is also a very fragile service due to copyright infringement and complaints from artists and record labels. This led to artists and record labels to request that their music be withdrawn from the Spotify. However, Spotify did not lose its popularity and remains a strong service that successfully utilizes its data and consumer information.

Big Data in Spotify


Big data has a significant place in Spotify. The uses of big data in Spotify primarily concern the classification of the musical tastes of each individual user, and subsequently, in recommending music that may cater to the tastes of said user. In addition to that, Spotify users can sign in with their Facebook credentials and Spotify makes use of data accessible through the use of Facebook’s public APIs to make user experience even more streamlined. Spotify can do this by analyzing and processing trends, Facebook likes and shares, and views. The processed data is then used to conveniently suggest music to the user on Spotify. Moreover, Spotify ads are heavily personalized based on what genre of music the user prefers.

Challenges and Threats


The biggest challenges of Big Data for Spotify are data quality and data security. Spotify must attempt to ensure that all of the data portrays accurate consumer information.  In order to do so, Spotify could use data algorithms in order to “clean” the data sets it collects. Data security is also an obstacle that should be taken into account when big data is implemented. With all the data collected from users, there are significant privacy risks associated with storing all the data before it is processed. Modern day network security does work against the majority of attacks, but there are still cases where a vulnerability might be exploited, case in point, back in 2014, when Spotify had to roll out a whole new Android application just to patch a bug which actually led to one user’s personal information being compromised (The Guardian, 2014).

A very useful implementation of big data is in the field of marketing, especially those of online ads. Spotify does have ads, but the targeting of advertisements could be better. Such is done with the use of machine learning and processing other information the user provides to give the user advertisements he or she may actually be interested in. Advertisements must be personalized, and at this time, Spotify only uses information gathered from music preference and taste. Spotify has the opportunity to gather more information on users from other mainstream media that it partners with. Such opportunity must be taken as soon as possible due to the changes that are occurring to the online world. Privacy is becoming an ever bigger concern due to the tendency of the users to demand stricter privacy policies. This limits Spotify’s ability to gather information on said users and provide accurate and desirable advertisements.

As for work distribution, we plan to separate our work based on the project outline as mentioned in the first lecture, with each person being assigned one of the four sections in the SWOT analysis, namely, strengths, weaknesses, opportunities and threats. This, we believe, will ensure a fair and equal distribution of work done per member. Apart from that, we all will be citing sources, thus providing evidence for all the facts we gather. Finally, we shall all work on a conclusion based on the information we collectively gathered. This way, each of us will be able to have a clear idea of the project and  be efficient representatives of our project.

Preliminary Bibliography

Statista. (2016). Monthly active Spotify users worldwide 2016 | Statistic. Retrieved October 02, 2016, from

Gibbs, S. (2014). Spotify hack leads to rollout of new Android app. Retrieved October 02, 2016, from

Datafloq. (n.d.). How Big Data Enabled Spotify To Change The Music Industry. Retrieved October 01, 2016, from
Spotify. (2014). Analytics at Spotify. Retrieved October 01, 2016, from

2 thoughts on “Spotify and Big Data: Algorithms and Data Processing Methods

  1. A big fan of the music streaming service I am a bit disappointed that you left many aspects out.

    The history of the service (traditional to big data), a comparison to a competitive service (apple music, tidal) in the usage of big data and also in why Spotify is so much more successful and how big data is utilised to create personal playlists for each viewer (daily and weekly mix). That being said, if you work more on those aspects, I believe that this topic has a lot of potential. 🙂


  2. Since I use Spotify myself, I found this project idea very interesting. Maybe you can further elaborate on certain aspects to give a full picture of the service.
    When you talk about how Spotify uses Big Data you could add a practical example to illustrate the process like e.g. the connection of the different Spotify Services (the way you use the Radio service affects your recommendations for Discover Weekly).

    As a challenge you mentioned that “Spotify must attempt to ensure that all of the data portrays accurate consumer information.” This is important for the data analysis, to be eventually able to give accurate recommendations. For the perfect customer experience it is further crucial that the music recommendations are consistent across all Spotify Services. This could be a challenge, given that people might use the Spotify Radio differently than Discover Weekly.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s