Musicovery

» see musicovery.com as it lived, on the wayback machine

~ the eulogy ~

musicovery's mood-pad interface was not the input i wanted. i was on rdio between 2013 and 2015, then spotify after that, both driven by search and library rather than by mood coordinates. musicovery ran the whole 2006-2017 window as a small parisian project that never crossed into my rotation, and it shut down its consumer webradio on 2 january 2017 with a facebook post announcing the B2B pivot.

the founders were vincent castaignet and frédéric vavrille, both french programmers who launched the beta in november 2006. the pitch was a discovery interface that skipped genre and artist as inputs and used two axes instead: energy (calm to energetic) and positivity (dark to bright). you clicked somewhere on the pad and the webradio streamed songs tuned to that coordinate. under the hood was a system tagging tracks along about 40 musical parameters.

the site had a free tier at 32 kbit/s and a premium tier at 128 kbit/s. it added integrations with itunes and amazon for purchase links, and it kept a lightweight social layer for shared playlists. for its ten years it operated as one of the smaller european music-discovery companies, competing with pandora and last.fm on the discovery angle and with rdio and spotify on the streaming angle.

~ the dossier ~

~ the mood pad ~

the mood pad is the piece of musicovery worth remembering. the whole interface was one 2D space with two dials, and you interacted by moving a dot. every music service before or since has organized discovery around some combination of search, genre, playlist, and algorithmic feed. musicovery replaced all of that with one continuous emotional coordinate.

for the users who liked it, this was the whole point: you did not need to remember what you wanted, you just clicked a point on the pad that matched how you felt and the music followed. for the users who did not, the pad was too much friction against the actual thing they wanted, which was to type in a song title. musicovery bet on the first kind of user and it turned out there were fewer of them than the interface economics needed.

~ the pivot ~

on 2 january 2017 musicovery announced on facebook that it had closed the webradio and was focusing on providing its recommendation and playlist engine to other music services. this is a specific shape of platform death: the consumer product goes off the internet but the underlying technology gets licensed. musicovery SAS the company survived. musicovery.com the destination did not.

the B2B pivot is a common exit for small european music-tech companies of that era. mood-based tagging turns out to be genuinely useful as a component inside someone else's product; the challenge of building a consumer product around it is that the audience willing to input a mood coordinate does not scale to the size that streaming economics need. sold as an API to an operator with a bigger catalog and a bigger user base, the tagging engine works fine.

~ the long tail ~

the actual musicovery brand faded out of the consumer web after 2017. the URL now redirects to a small B2B site describing licensing options; there is no interactive mood pad on the front page anymore. spotify's own recommendation engine has since folded similar emotional-shape features into its track-metadata layer as inputs to its personalized playlists.

musicovery sits with rdio and songza and mog in the 2010s music-service graveyard: small, well-designed, run by a founding team with a strong point of view, killed by the economics of licensed streaming at scale. the mood pad was the interface no one else built and it lived exactly as long as its two founders could keep it running.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »