the jukebox has a long tail // 16 Oct 2004

Chris Anderson's piece The Long Tail, already linked to shreds, brought a smile to my face with a quote from Ecast's Robbie Vann-Adibe. Robbie and I worked together at Viant, and I was on the Ecast team when Viant put together Ecast's original business plan, operating model and working prototype of their digital jukebox...

Meet Robbie Vann-Adibe, the CEO of Ecast, a digital jukebox company whose barroom players offer more than 150,000 tracks - and some surprising usage statistics. He hints at them with a question that visitors invariably get wrong: "What percentage of the top 10,000 titles in any online media store (Netflix, iTunes, Amazon, or any other) will rent or sell at least once a month?" ... Most people guess 20 percent, and for good reason: We've been trained to think that way. ... But the right answer, says Vann-Adibé, is 99 percent. There is demand for nearly every one of those top 10,000 tracks. He sees it in his own jukebox statistics; each month, thousands of people put in their dollars for songs that no traditional jukebox anywhere has ever carried.

I can't tell you how happy that makes me. Captured in there is the whole point of Ecast: thousands of tracks, delivered transparently on demand, giving barflys access to nearly whatever they'd want to hear at any time. And the jukebox is the perfect place for a long tailed catalog -- at one point we had modeled (inebriated) people paying slightly more for tracks that they typically wouldn't find in a corner bar.

And layered on top of the "any song, anywhere" benefit for users is the data component. Back in 1999/2000 when Ecast was blazing contractual trails with the major labels ("You're going to distribute digital versions of our catalogs to bars?"), the folks on the other side of the table didn't quite get the value of the data Ecast would be collecting. I have to believe that's changed: Ecast is sitting on geographically tagged social purchase data. As the hit business both concentrates and dissipates, being able to predict the lifetime value of a recording means more rational investments in artist development. It may not be as richly detailed as the data from Soundscan or the iTunes Music Store, but if I were in A&R or artist marketing at a major label, I'd sure as hell want to layer jukebox stats into my forecasting models. Music spreads socially, after all...and what's more social than feeding a fiver into the jukebox?