Has it actually been two and half years since I started researching the Long Tail?! The first complete data set I got was from Rhapsody in March 2004, which I used in the original article here. I then got updated data in March 2005 and again in December 2005, the last of which I used in the book. Now I've just done the first pass at analyzing the March 2006 data, which I presented at an ADA meeting in Philadelphia last week.
Here's the slide from that presentation. The Long Tail of music, as defined as downloads of music not carried by the average Wal-Mart (using a relatively generous figure for Wal-Mart's typical inventory), has grown from 26% of Rhapsody's business to 30% over those three years, as Rhapsody's inventory has grown from 500,000 tracks to 1.3 million.
Talking to executives at the meeting, I also got some good data on how the various music markets compare in their degree of hit-centricness. Here's that slide ("P2P" refers to the major file-sharing services, as measured by BigChampagne):
Nearly half of Wal-Mart's music business is selling the top 100 artists. On Rhapsody, it's less than a quarter. This is a great example of the Long Tail at work.