The disruption of media use is primarily driven by the big four American companies Google, Amazon, Facebook and Apple - but in the second row remarkable experiments happen. Spotify was with me in the past few months again and again with new experiments on the radar - time for a review article on the innovation experiments at the Swedish music service.
Innovation on the shoulders of giants
Spotify relies on the operating systems of Apple and Google and still manages to develop utilization models at many ends of the user contacts - based on marketing of user data, playing personalized content, the development of their own advertising network - to more and more also the production of their own content, For me, the service is an example of trying new things - in small completed experiments, always questioning the status quo, but also: Always on the shoulders of the GAFA giants. The need to develop other business models on the platform will continue to increase. Creativity and diversity can serve as models for other companies.
Master of algorithms
In my perception, Spotify has delivered a big promise as one of the first services: the recommendation of good music. It took a few years, but: The "Discover Weekly" list is almost always great (unless someone else listens in my stream).
On a technical level, "understanding" of music is highly complex, in addition to general genre, style, and beats, usage data, length of stay, location, and many more data points are also included in the calculation. In the meantime Spotify seems to have solved the complexity for a user, because: In addition to a price adjustment for family tariffs, Spotify wants to offer combined recommendation lists starting this fall - for example, with the family mix it should provide music recommendations for the whole family.
Spotify listening to non-music content.
- Daniel Ek, CEO & Founder Spotify
Word becomes more important
With the acquisition of the pimped-label Gimlet for an estimated $ 230 million and podcast-producing cloud software Anchor, Spotify also clearly demonstrates the importance of contributing to the future of the platform. CEO Daniel Ek estimates that more than 20% of Spotify's content will not be music. To the classification: The whole thing is a bet on the rising podcast market, as TechCrunch shows nicely.
Spotify experimentally experimented with algorithmically played playlists for the way to work ("Your Daily Drive") - in which then podcasts and music are played mixed: the personalized radio. With current Gema regulations, this experiment will probably not be tested so soon in the mass, but the cost of individual streams are far too high.
In January, Techcrunch talked about Spotify's plans to commercialize its own advertising and at the same time enter exclusive deals with podcasters - world's most successful exclusive podcast is "Feast and Fluffy" by Jan Böhmermann and Olli Schulz.
Gold media sees the subscription sales of streaming services such as Spotify, Apple, Amazon and Co. in Germany in 2019 at about the level of net advertising sales of all German radio stations, in 2020 even for the first time. According to the estimates of Vaunet figures, the revenue with Paid Audio in 2018 is slightly above the advertising revenues.
Storyline: Spotify becomes music television
With Storyline, Spotify tests a kind of social network in the tradition of Instastories ... Artists get the opportunity to offer "artwork" for their albums and songs, which can be excerpts from the videos, news and information - as an example, the terrible song "Cherry Cherry Lady "in the even worse version of Capital Bra (please do not ask how I figured that out) - on the smartphone there are video clips in the upright format that playfully adds a video experience to the Spotify audio service - and presumably so the dwell time on the platform still increased significantly.
Together with the music information portal "Genius" there is also the format "Behind the Lyricc", which provides information on artists and songs in the first few bars of the music (which has been around since 2016). Thus, the service is perfectly prepared for the jump from the phone to the smart displays of Google or Amazon - and could perhaps revive the music television of my generation.
Advertising at language assistance level
Spotify-based voice assistant advertising on the Spotify platform has been publicized since May - Spotify currently only runs content advertising on its own platform, users are referred to, for example, a playlist or podcast ( here a link with sample application). Pandora wants to start a similar experiment in 2019.
Data for brands
Spotify acts extremely data-driven and uses the information collected multiple times. For example, in collaboration with brands - on SpotifyforBrands.com, there are studies on Livestyle Generation Z and Millenials, Best Practice for advertising formats and a number of white papers - a great place for data nerds.
Spotify for the cigarette lighter - that's the idea of a Bluetooth dongles (called "car thing") that can control music and podcasts in the car by voice. However, Spotify also made it clear that it's more about collecting user data (only in the US) from a smaller group, so there are no plans for a corresponding product - too bad, I would be there. However, further tests seem to be planned - Spotify talks about possible "Voice Things" or "Home Things".
Content Curation - Playlists to choose from
For the last European election there was an obviously human curated playlist of artists from all European countries - creepy to listen to, but this experiment shows: editorial also wants Spotify offers and can supply.
Goodbye to the record collection
That I still find surprising: How complicated it is to build your own music collection, the "ownership" of music from earlier empathizes - I can search and play albums targeted, organize individual tracks in lists, using a song or artist create playlists let - but from the collection lets Spotify fingers. For me, it is not clear if this is more of a licensing problem, or if thinking in collections for other reasons is simply not "up-to-date" anymore.
Taylor Swift cleaning up
The inspiration for this article came to me when cleaning up - at the command "Alexa, play Taylor Swift", I probably came in a test group for radio and podcast experiments, because all five songs told a moderator anecdotes from the life of Taylor Swift, This has never happened to me on Spotify before, but it felt organic, made cleaning up more entertaining and I thought to myself: Actually like radio should be.
I certainly forgot some experiments in this article - who still notice which ones: I am grateful for hints!