Spotify: What does it mean to be the best data analytical DJ?

7

October

2021

No ratings yet.



Share of music streaming subscribers worldwide in the 1st quarter of 2020, by company. (Source: Statistica, 2020)

You might have noticed, but over the last decade, the leading music platform is the Stockholm based company Spotify. Over the year, the company had established a competitive advantage that allowed the company to compete with some of Silicon Valley’s most innovative companies.

How does Spotify prepare its DJ set?

For Spotify users, music is organized not by genre or style (and certainly not by album, a concept that has become more obsolete) but by mood, activity, and “musical keywords.” The user is provided with several icons under the “browse” section that led to musical selections. In the Spotify language, these are referred to as “hubs,” and they are depicted by clickable square thumbnails.

Spotify’s music vision is clearly to come as close as possible to the subscriber’s mood and personalities preferences. It also means that Spotify is not just a place to listen to the music you want; it’s also a place to learn about it as you navigate through a sea of attractive icons that respond to clicks with a variety of auditory options. To put it another way, Spotify is primarily a music discovery platform.

What tools does this DJ have?

It’s hard to know how Spotify’s music discovery engine works inside and out. This is due to the fact that the system does not operate in any single way for any one person at any given time. The underlying algorithm is also a well-guarded trade secret.

To quote the words of one of the designers of Spotify’s recommendation system, the success of the model rests on communicating that its repertoire is both complete and properly maintained – that it has achieved a unique balance of “scale” and “care” (See Whitman 2012).

Spotify develops ‘meaning classifiers,’ which are predictive ‘machines’ that evaluate audio data and forecast community and personal reactions. The company’s goal is to interpret this extra-signal data computationally and connect it to the signal in such a way that future sounds may be anticipated.

What’s the role of a DJ?

On Spotify home’s page, you can find the following statement:

“With Spotify, it’s easy to find the right music for every moment. Choose what you want to listen to, or let Spotify surprise you. Soundtrack your life with Spotify”

The questions that arise are, then, what do we expect from our musical platforms? To what extent do we want to control the music that we are listening to? Spotify’s commitment is to be able to surprise their subscribers, the extent to which it is the case can be debated. At last, what do we expect from our DJ?

References:

Chodos, A. T. (2019). What does music mean to Spotify? An essay on musical significance in the era of digital curation. INSAM Journal of Contemporary Music, Art and Technology, 1(2), 36-64.

Whitman, B. (2012). How music recommendation works — and doesn’t work Brian Whitman. Retrieved October 7, 2021, from https://notes.variogr.am/2012/12/11/how-music-recommendation-works-and-doesnt-work/

Whitman, B. A. (2005). Learning the meaning of music (Doctoral dissertation, Massachusetts Institute of Technology).

Statista. (2020). Subscribers share of music streaming services worldwide Q1 2020. Retrieved October 7, 2021, from https://www.statista.com/statistics/653926/music-streaming-service-subscriber-share/

Please rate this

4 thoughts on “Spotify: What does it mean to be the best data analytical DJ?”

  1. Great topic Antoine! As a fellow Spotify user I’ve never considered how the playlists are made. In answer to your question I think Spotify gives the user the possibility of doing both options, firstly by controling their music through their personal library and the search button but also lack of control when we choose a random playlist. From the platform and from my ‘DJ’ I expect familar and unfamilar music, afterall if we kept getting recommended music we knew it would be quite boring, no?

  2. Hello, Antoine.

    Thank you for writing such an informative blog. It’s fascinating to see that Spotify has 30 percent of music stream subscriptions, which doesn’t surprise me at all. Actually, I listen to Spotify three times a day and it never bores me. It would be interesting to learn more about Spotify’s approach because the platform continues to surprise me. Personally, I believe it’s fantastic that Spotify continues to surprise us with new music linked to our favorites.

  3. Very interesting topic Antoine. Also I have been a frequent user of Spotify and I never questioned the control it potentially has on the music that I listen too. And to be honest, I am fine with it as long as Spotify keeps (positively) surprising me. What I would be interested in, is knowing whether the predictive machines and recommendations are solely based on evaluaton of audio data, forecast community and personal reactions’ or if Spotify has a secret agenda of pushing certain artists, labels, and/or genres.

  4. Hi Antoine, you touch upon a very relevant topic. I was very surprised by the large share of Spotify users, but even more surprised by the high market share of Apple music. Personally I solely use Spotify, it’s convenient to control your own music, but sometimes it’s comfortable to let the Spotify DJ pick your Music. Moreover, I appreciate the mood playlists. It introduces you to new music which could surprise you positively. I would be interested in how this algorithm works out and more important how they profit from these algorithms. What kind of power will a company like Spotify have in the future when it is able to determine which music you appreciate. Therefore, I think for us as a customer we like the convenient features of Spotify. However, we should be aware of how such a company could gain power in our music preferences and how the market could be manipulated like that.

Leave a Reply

Your email address will not be published. Required fields are marked *