What Exactly is Music Metadata and Why is it So Important?
Written by Joshua Cruz
When thinking of the key elements that make up our favorite songs, a distinct melody, catchy hook, and maybe even a funky bass line might come to mind. Although these are the reasons why we might listen to a song over and over again, that’s not why or how we first connect with music. Music metadata is what connects us to new songs and can help us find our next favorite tune.
As music listenership has become primarily driven by streams, a song’s revenue and popularity are directly tied to its metadata management. Flawed metadata can cause a song to be lost in the millions of tracks on streaming platforms, or worse, limit its creators’ ability to collect their earnings. With this year’s Grammy Awards all wrapped up, our Salt Flats metadata team gives an overview on the current state of music metadata.
As it relates to music, metadata is the information that pertains to a song file. Music metadata can be separated into three categories:
Descriptive metadata
Whether you are a label executive, playlist curator, or just a casual listener, most of us are already familiar with descriptive metadata. This type of metadata contains the basic song information that we already see when we open our music libraries, including artist, song title, cover art, and release date. Descriptive metadata is objective and based on the publishing information supplied by the artist or their representation.
Ownership/Performing Rights metadata
In an ever-changing music landscape, ownership/performing rights metadata is arguably the most significant type of metadata, as it specifies royalty calculations for each song. Music releases today tend to exchange hands many times over, cycling between songwriters, artists, producers, and publishers, who all have specified percentage stakes for their contributions. To paint a picture of the importance of this type of metadata, Mary J. Blige’s Good Morning Gorgeous, which was nominated for Album of the Year, credits 12 additional singers, 23 producers, 21 engineers, and 47 songwriters.
For some individual songs, these royalty calculations are straightforward, while others have more complexities. Let’s take two of this year’s Record of the Year nominees, “Bad Habit” by Steve Lacy and “BREAK MY SOUL” by Beyoncé. “Bad Habit” credits one producer, which happens to be Lacy himself, five songwriters, two engineers, and one mastering engineer. “BREAK MY SOUL” credits four producers, eight songwriters, three engineers, and one mastering engineer. Royalty splits are easier to calculate for Lacy’s work, meanwhile there is likely more negotiation and complication for Beyoncé’s.
Recommendation metadata
While the first two types are supplied by the artist’s camp, recommendation metadata is created by streaming services and other digital service providers (DSPs). These tags attempt to label subjective information about each song, album, or artist, including mood, listener demographic, and popularity. Each service has their own proprietary system of adding tags to curate related songs to their listeners. Think of Spotify’s “Daily Mix” playlists, Pandora’s Music Genome Project, and Apple’s “Hey Siri, play some music I like” feature. In the case of Spotify, song attributes include tags such as energy, overall popularity, and danceability. Whether they are hitting the mark or not, the goal is to create a personalized experience for each individual user.
Spotify is known for their recommendations, creating both human-curated and algorithm-based playlists.
Although such personalization features leveraging metadata are common, there are some imperfections in metadata management. The overarching issue is that there is no industry standard for collecting, verifying, or storing metadata, leaving room for inaccuracies, as information is transferred between parties. In the case of ownership/performing rights metadata, most metadata is manually submitted, making human error a major factor to consider. Not only could false information or a simple misspelling of a party’s name come off as unprofessional, but it could cost artists, producers, and labels some serious cash. Niclas Molinder, founder of music metadata company Auddly estimates that about 25% of music publishing revenue does not make it to its rightful owners due to metadata inaccuracy. This leaves billions on the table by some estimations.
Another important aspect of music metadata is the visibility that is provided to those who contribute to the making of a song. Having credits on a track provides validity and promotion for lesser-known artists and producers looking to land more projects. If someone has their name misspelled, or is left off the credits in total, their discoverability can be significantly hindered.
There have been attempts to create global centralized databases and collaboration tools for metadata, but all have fallen victim to various roadblocks. With 100,000 songs being uploaded to Spotify and other DSPs daily, artist revenue opportunities continuing to diversify, and global collaboration on the rise, the industry could benefit from some unification, as the importance of maintaining accurate metadata cannot be understated. Going forward, it will be worth keeping an eye on whether or not the industry comes together to create standard procedures for metadata management to protect artists, producers, and labels alike.
At Salt Flats we love music, but metadata is our real jam. If you’re interested in learning more about how metadata can help your business you should contact us to learn more.
Sources:
The Verge: Metadata is the biggest little problem plaguing the music industry
Soundcharts: How Broken Metadata Affects the Music Industry (And What We Can Do About It)
Playlist Radar: Spotify 101: Understanding the Spotify Playlist Landscape
Bandzoogle: 22 ways to make money as a musician
Hypebot: All The World's A Studio: Global Collaboration Will Transform How We Make Music Together
