Wednesday, June 19, 2013

The Amazing Predictive Ability of iTunes Radio

The Amazing Predictive Ability of iTunes Radio

I was never much of a fan of Pandora. I started using it in the early days, and the predictions just weren't very good. I found myself using skip to the point of being shut out almost every time I tried to listen. I got a Zune Pass and never looked back. The Zune Pass evolved into an Xbox Music pass, and then I just started listening to Spotify for free instead. These whole predictive playlists just weren't cutting it for me. Then, I decided to start using iTunes Radio.

I wasn't listening to a simple radio channel on my favorite artist; I'd expect any service making a station based on my favorite music to deliver good tracks. A few nights ago, I saw this commercial:

The music seemed like a cool mix of an oldie vibe with the modern stuff I generally dislike and avoid, so I figured, why not give it a shot. It was Lily Allen. I listened to the song on Spotify at work, and then listened to the rest of her music. A fair bit of it was good, I thought.

But, I don't pay for Spotify, so I couldn't listen to more in my car on the ride home. "Well," I thought, "I've got this iOS7 beta... I guess I could give iTunes Radio a shot." I started up Lily Allen radio.

I didn't hit skip once.

I've listened to radio stations on Pandora based off of Alter Bridge and Halestorm, perhaps my two favorite artists, and used skip liberally. It just wasn't all that great.

But here I was, listening to a station based off of an artist whose music was stylistically far out of my comfort zone, who performed a genre I traditionally hate. And iTunes Radio nailed it.

I don't know how their algorithms work, what factors they consider when building them, but it seems apparent to me that Apple is in a position to do this custom radio station game better than any other company on Earth.

When you really think about it, it's not surprising. They've been doing their Genius playlists for years now, gathering vast amounts of data about what people have in their libraries and how often they listen to it, perhaps even data about what combinations they like best. They have all that data for you, too, so they already know what you like.

But they've got another step up from that, too. They have iTunes. Apple can tell what music people who enjoy song X like enough to actually pay for it. That could be huge, if they're leveraging it.

Apple has more data about music and music tastes than anyone. So this fall, when you get your hands on iTunes Radio, and wonder why you should bother using it instead of Pandora/iHeartRadio, etc., consider the fact that they just might know more about your music taste than even you do.

Source: http://whitenoise.gizmodo.com/the-amazing-predictive-ability-of-itunes-radio-514020593

zimmerman charged bonobos

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.