Rdio

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

~ the last words ~

rdio launched in closed beta on 4 august 2010. it was made by the skype co-founders janus friis and niklas zennström, who had also previously made kazaa. they had sold skype to ebay in 2005 for 2.6 billion and were now in the music streaming business.

the product was a paid subscription music streaming service. unlike spotify, which had a free ad-supported tier, rdio was paid only. the pitch was that paying users would get a better experience: cleaner interface, no ads, real music curation, social features that worked because nobody was on the free tier looking for free music.

rdio launched publicly in the US in october 2010. it expanded slowly to other countries. the design was widely admired. tech reviewers consistently said rdio was the most beautiful music streaming service. the audience was small but real, mostly tech-curious users in the US.

spotify expanded internationally faster than rdio. apple music launched in june 2015 with a much larger marketing budget than rdio could match. by mid-2015 rdio was clearly losing the streaming war. on 16 november 2015 the company announced bankruptcy. on 22 december 2015 the service shut down. pandora bought rdio's technology and engineering team for 75 million dollars but did not keep the rdio brand alive.

~ the record ~

~ what made it good ~

rdio's interface was clean. the typography was thoughtful. the album art was prominent. the playlists felt curated rather than algorithmic. when you opened rdio, you felt like you were in a music app made by people who cared about music, not in a tech product trying to maximize engagement metrics.

the social features worked because the user base was small and engaged. you could follow other users. you could see what they were listening to. you could subscribe to their playlists. for a small audience of music-attentive users, rdio felt like a music club rather than a generic streaming service.

the recommendation engine was decent. not as good as spotify's algorithmic playlists became after a few more years, but better than apple music's at launch. rdio tended to recommend things adjacent to what you were already listening to, which was useful when discovering new music in a familiar genre.

~ the strategic problems ~

the paid-only model was rdio's biggest strategic disadvantage. spotify in 2010 had a free ad-supported tier that let users listen for free with limitations. people would sign up for spotify free, then upgrade to spotify premium when they got tired of the ads. rdio required a subscription from day one. the conversion funnel was much narrower.

international expansion was slow. spotify launched in europe in 2008 and the US in 2011. rdio was largely US-focused for the first three years. by the time rdio expanded to europe and other markets, spotify had already taken the audience.

marketing budget was a real factor. spotify raised hundreds of millions of dollars and spent aggressively on user acquisition. apple music launched with apple's full marketing power behind it. rdio had a smaller fundraise and could not match the spending. as the streaming war shifted from product quality to marketing reach, rdio fell behind.

~ the music license problem ~

the deeper structural problem was the cost of music licenses. streaming services pay per-stream royalties to record labels. these royalties consume most of the revenue. spotify in 2014 was paying about 70 percent of revenue back to labels. rdio's economics were similar. operating margins were thin.

this means streaming services need scale to be profitable. spotify's hundreds of millions of users let them negotiate better rates and amortize fixed costs. rdio's smaller user base meant worse rates and higher relative costs. the math kept getting harder as the user count gap widened.

this is also why most streaming services have died or been acquired. only spotify and apple music (subsidized by hardware sales) and youtube music (subsidized by youtube ads) have survived the economic gravity of music licensing at scale.

~ how it ended ~

rdio announced bankruptcy on 16 november 2015. simultaneously they announced that pandora would acquire the technology and engineering team for 75 million dollars. the rdio brand and service would be wound down.

existing rdio users were given about a month to download their playlists or migrate to other services. the official shutdown date was 22 december 2015. on that date the rdio apps stopped working and the website went dark. user playlists were preserved for download for a few months after, then the export tool was retired.

pandora kept the engineering team and used some rdio technology in their own product. the rdio brand was retired. some of the design language influence is visible in subsequent music streaming apps but rdio itself is gone.

~ what we lost ~

the prettiest music streaming service. spotify is functional but visually crowded. apple music is clean but locked into the apple ecosystem. youtube music is utilitarian. rdio's design was the cleanest of the streaming era and nothing has matched it since.

the paid-only model. rdio's argument that paid-only would produce a better experience was probably correct from a product quality perspective. it was wrong as a business model in a market where free tiers dominated user acquisition. that lesson was learned by every streaming service that came after, all of which have free tiers.

the case study in design-led streaming. rdio's user research and design work shipped with the product and was then absorbed back into the team that went to pandora and various other companies. the design language influences are diffuse but real. the original rdio app, with its specific typography and album-art-first layout, is still cited by music app designers as a reference point.

@}-->--

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

« previous  ·  random  ·  next »