4 Ways One Big Database Would Help Music Fans, Industry

MySpace Music’s acquisition of imeem revealed, for the umpteenth time, a major problem with the music business — not piracy, nor the ongoing failure of pure-play ad-supported music services, but rather that each digital music service grows in its own data silo. This impacts us in all sorts of negative ways. In this latest example, […]

bigdatabase

MySpace Music's acquisition of imeem revealed, for the umpteenth time, a major problem with the music business — not piracy, nor the ongoing failure of pure-play ad-supported music services, but rather that each digital music service grows in its own data silo. This impacts us in all sorts of negative ways.

In this latest example, some imeem users are upset that their carefully constructed playlists are gone, for now anyway. MySpace Music plans to reconstruct them on its own site(s), but it admits that the job will be arduous and time-consuming.

The solution to this and other problems dogging the music industry could be forehead-slappingly simple: one big, free, public database with, at the very least, song titles in one column and unique identifiers in another. When online and mobile music services build their own content databases out of the labels' catalogs, they would have incentives to use the same numbers to identify each song, for the reasons laid out below. Music services already apply their own unique identifiers to songs in their catalogs, so the use of numbers is not the issue — they just need to be the same numbers.

This database would have to be free, readily available and totally transparent, visible to music fans and industry people alike, because the barrier to entry for startups to use the system would have to be zero. Open source software making use of the data set, available on the same website, might encourage services to use the numbers.

This idea would work, and ReverbNation and MySpace's own iLike site have already proved it to an extent, leading off our list of four ways one big database would help the music industry:

1. Paying Artists

SoundExchange has at times in the past been a bit lax about finding artists owed part of the money it collects from webcasters and satellite radio stations. Integration with other databases is changing that, albeit slowly, to SoundExchange's credit.

The multifunctional artist-oriented site ReverbNation, used by many to track tours, e-mail fans and the like, has so far compared about half of the artists in its database to artists in the SoundExchange database by matching artists in its system to SoundExchange's database of unpaid artists by artist name, rather than a unique identifier. Then, it's up to SoundExchange to validate the artists and send out their money, according to ReverbNation co-founder and COO Jed Carlson, who said matching up the rest of the data will take about another month.

Out of the 250,000 artists in its database that ReverbNation ran against SoundExchange's database, it discovered 5,500 who were owed money by SoundExchange, which previously claimed it couldn't find them. This is not pocket change. So far, SoundExchange owes the 5,500 artists ReverbNation has found approximately $4 million, according to Carlson.

The same goes for MySpace Music's iLike, which likewise ran its artist database against SoundExchange's database and found even more money sitting on the shelf. "SoundExchange has been holding money in escrow for artists, but lacked the contact info for thousands of lesser-known artists or labels." iLike co-founder Ali Partovi told us. "We matched up our database with theirs and identified over 8,000 artists representing over $8 million in royalties to be disbursed. iLike is contacting each of these artists to let them know they have a royalty check waiting for them."

It's an admirable effort, but one that would be made far easier -- not to mention faster -- if SoundExchange, iLike and ReverbNation used the same unique IDs for songs. And in many cases, time is of the essence. SoundExchange's charter requires it to store records on who's supposed to get paid for three years. After that, the money owed presumably gets disbursed among those artists who have registered in SoundExchange's database to get paid for digital and satellite streams.

Some might wonder whether the music industry really wants a fully transparent system that zips the right amount of money straight through to the people who deserve it, but that's another question. (Inefficiency may be a good thing for some, but not most, of the parties involved.)

relational_database11

2. Easing the Inevitable Acquisitions of Digital Music Services

Music startups tend to go under or get bought. When MySpace Music purchased imeem, users of that service were upset to see the service itself disappear, of course, but particularly its API (more on those later) and their playlists, which MySpace Music is still porting to its own music service(s).

If MySpace and imeem had built their music systems with the same song-identification numbers, they would be capable of creating what's known as a relational database, where one column acts as a sort of hinge between two databases (image above). In that case, porting every single user-generated playlist from imeem to MySpace Music or the company's iLike property would be a simple matter of making the two databases talk to each other in the shared language of those unique song identifiers.

Imeem itself owned a database that could form the basis of exactly the sort of big, public database I'm talking about here. Open-sourcing the database could solve the sort of problem that now dogs imeem's new owner, MySpace, of converting playlists from one music service to another.

Snocap, — co-founded by Shawn Fanning of Napster fame in 2002 — was purchased by imeem in 2008. SnoCap has long powered the artist- and label-created music stores on MySpace band pages and elsewhere, but its main strength could be its song-identification database, which includes audio-fingerprinting technology that can identify mislabeled songs and copyright-holder contact information.

MySpace Music told us it did not acquire Snocap for, and has no plans to use, its database going forward. As the company is finding out now, it's no small matter to integrate one music service into another without shared song identifiers. Perhaps the database belonging to Snocap, which is currently offline and apparently without an owner, could form the foundation of a big database that would make future digital-music-service acquisitions easier on everyone involved.

3. Better APIs and Mashups

Application Programming Interfaces are enabling a new music distribution system that sends music to fans through increasingly circuitous routes. Mashups are combining two or more services -- Twitter and YouTube, Facebook and iLike, and tomorrow's geo-aware social music apps.

It's becoming increasingly clear that APIs are a real engine for growth in music usage. But as with the integrations between SoundExchange and ReverbNation and iLike, the companies that use them still have to do a fair amount of legwork if they want to include the same song from multiple services. If music services used the same numbers to refer to the same songs, APIs with hooks into multiple music services would be easier to build and more effective for consumers and for music distributors.

4. Subscription Services Would Be Way Less Scary for Consumers

Aside from the thought of actually paying for music, one reason some consumers are wary of music subscriptions such as Rhapsody, Napster, MOG, the paid version of Spotify, the paid version of the former imeem, and the like is that they might decide they don't like the service. Or, if the service changes its terms or goes out of business, users can't port all the songs they've added to their accounts or playlists they've created.

Shared, unique song identifiers would enable subscription services to offer a new clause in their user agreements: that subscribers have the right to port their whole subscription — or cloud-based account — to another service, should they please. This would lower the barrier to subscription in the minds of potential customers, but the only way it is going to happen is if all these services use the same numbers to identify the same songs.

Subscription services might bristle at this at first -- why would they want their customers to be able to escape to another service? But from the consumer's point of view, a subscription — or cloud-based music collection — isn't worth much if you can't take it anywhere.

See Also:

Photo: Flickr/natita
Image: Depicting a relational database, from Flickr/Andréia