Music discovery, big data, and brands: making a better radio

We are in the midst of the second disruption of the music industry, with the rise of streaming services like Spotify and iTunes Radio. Yet music discovery hasn’t changed all that much since terrestrial radio. Although music industry revenues have remained flat over the last four years, the fear is that we’ll see a decline like the one we saw over the last decade when music sales fell from a high of $14.9 million in 1999. What can prevent the second plunge? The enormous amount of data created when we move music to the cloud.

I was excited to hear Paul Lemere, music technologist from The Echo Nest (recently acquired by Spotify), speak about music discovery at SXSWi. The Echo Nest is responsible for the music discovery algorithms in many of the major streaming services — Rdio, Spotify, and iHeartRadio are among the services they power. (Coincidentally, Spotify announced that day that it was acquiring the company. This caused somewhat of an uproar in the industry.) Paul’s job is to find new and innovative ways for listeners to discover music through data. His talk was about how to make a better, more personal radio.

Music Listeners are Not Created Equal

Only 7% of music listeners are actually truly passionate about it yet most of the tools for discovery cater only to this minority. There is no great UX for music discovery for casual or indifferent listeners. If the experience of music isn’t better than the radio, then they’ll just stick with the radio.

How Data changes the Listening Experience

The Echo Nest APIs explore the relationship of music to age, gender, location, taste (like “mainstream-ness” or diversity). By looking at playlists The Echo Nest can even track context (music you wake up to vs. music for “sexy time.”) This gives you an intimate (literally) listening portrait of a user. You can even make a better playlist for people you know nothing about by simply removing tracks that skew in any one direction over another.

How Does Better Music Discovery Change the Music Industry?

The obvious answer is that it will create more enthusiasts, who spend 10x the amount on music. But the less obvious and more overarching disruption happens when we begin combining the data captured by listening with the data surrounding the listener.

  • Combining social, music preference, and third-party data will link brand preference to music in a quantifiable way. This has obvious implications on how to use music to build a brand in an era where the bulk of musician revenues comes from brand partnerships.
  • Music and location, particularly in the place where we listen to half of our music, could mean the re-emergence of radio advertising in a programmatically driven world.
  • The rise of the Quantified Self could mean new ways for healthcare companies, who now need to market directly to consumers, new ways of reaching that audience.

Simply providing “Artists you may like” isn’t true discovery. “That’s a shopping list,” Paul says. We’ll be able to discover music in new ways, something we explored a few years back when we partnered with The Echo Nest and radio station KCRW to build the Matchmine app for iPad. The untapped potential of music data could save the music industry and reinvent others. Previously unknown insights about how we listen may turn into a revolution in advertising.

By the way: Paul’s blog is full of interesting ways to use the incredible amount of normalized data found in a song. my personal favorites: Girl Talk in a Box and the Infinite Jukebox. Mind. Blown.

Headline Photo Credit: bricolage.108