Live streaming has brought about a tectonic shift in music consumption. Apps like gaana.com, Spotify
and Pandora have been at the forefront of this change.
Brian Hamilton, General Manager, Music and Auto, Gracenote (Nielsen Company) in conversation with
Pitchteam, discusses the importance of Metadata in the music industry today and the growth of the
music business in India.
How is streaming changing the algorithms of the music industry?
BH– With streaming, the music industry is enjoying an unprecedented view into music consumption, listening
habits and a better than ever understanding of listening preferences. In addition to accurate playback
counts, streaming services receive direct listener feedback based on like/dislike features, skips and
replays. This information is being used to identify super fans for concert pre-sales, early access to album
releases and direct artist-to-fan initiatives, among other things.
In addition, deep descriptive information about songs combined with playback information can help
drive personalized recommendations of new songs based on analysis of common musical traits found in
one’s playback history. The ultimate goal is to use all of these data signals, along with deep descriptive
metadata, to accurately predict the exact song you want to listen to in a specific listening context,
whether it’s a commute into work or a relaxing day at the beach.
With apps like Spotify and Pandora, live streaming has become convenient. What change has this
brought about in terms of their listenership? How important a role does Metadata play in the music
industry and also does it help in marketing a particular song or album?
BH– Streaming has significantly increased access to all of the world’s music in a way which was unthinkable
in the physical media era. Previously, music fans were relegated to a collection of 10-12 songs on an
album whether it was on a cassette tape, CD or digital download from iTunes. Now, we’re looking at
immediate access to millions of songs – we are literally walking around with 60 million songs in our
pockets – not including user generated content on platforms like YouTube.
Massive volumes of music content to sift through make discovery and recommendations even more
critical for a good listening experience. Today, music is typically categorized by artist genres (Rock, Hip
Hop, etc.). And while you can draw parallels between songs based on artist genre, deep descriptive
information about each song (tempo and mood) can help pick new songs from local artists and less
popular independent artists that may not normally appear in popular playlists.