Digital music has transformed how people buy music, but the most popular albums have grabbed a larger market share since the iTunes music store launched in 2003. The Top 10 had a 5.09% share of album sales in 2004, according to Nielsen SoundScan (2004 was the first full year iTunes operated in the U.S.). The share fell to a low of 4.27% in 2006, rose to a high of 6.17% in 2010, dropped slightly to 6.12% in 2011 and fell to 5.59% last year.
It's not a paradox so much as a rewriting of the theory of the long tail. Fewer people are buying Top 10 albums -- just as fewer people are buying fewer albums in general -- as analyst Mark Mulligan noted at his blog.  But the Top 10 has bucked the trend seen elsewhere on the album chart. Popular albums are doing relatively well, all things considered.
Less popular albums -- everything outside of the top 10 -- have not fared so well. While the top 10 titles gained market share worth 1.59 million units since 2004, albums ranked 11th through 40th lost market share worth 6.74 million units. Albums ranked 11th through 20th went from a 3.44% share in 2004 to a 2.64% share in 2012. The next 10 ranks, 21st through 30th, fell from 2.77% to 2.04%. It continues that way through the top 200 ranks.
Put another way, popular albums have remained popular, while less popular albums have ceded market share to the flood of albums that gained market access during the age of digital distribution. If you were a Top 10 artist in 2012, you sold an average of 1.7 million fewer units than a Top 10 artist in 2004, but your share of album sales was 0.5% higher than it would have been in 2004. 
On the other hand, artists outside the Top 10 have lost market share. The worst place to be was between 11th and 20th. Albums in that group lost 0.81% market share, worth 2.55 million albums from 2004 to 2012. The next-worst group was 21st to 30th. It lost 0.73% market share worth 2.31 million albums.
Where did that lost market share go? Less popular albums have benefitted from digital distribution and Internet marketing and have taken market share from popular artists. With fewer brick-and-mortar stores in existence, any one album has less command over physical shelf space. Although the notion of digital shelf space does exist -- iTunes has limited space on its pages, for example -- a digital retailer can carry tens of million of songs. Consumers have access to more titles and can buy anything they desire.
Digital distribution has also brought the rise of individual track sales. While nearly all albums have lost sales over the years -- a typical album in the top 100 has lost dropped 60% from 2004 to 2012 -- a portion of that deficit is now made up for by a la carte track sales. And, as I pointed out last month, the top digital tracks have also claimed a bigger market share over the years. But whereas only the top few albums have claimed a bigger market share over the years, all tracks in the top 200 have increased their market share.
If there are simple storylines here, they go something like this: consumers have become diverse in their album buying while the albums in the Top 10 have snared an even bigger share of overall sales. At the same time, digital consumers have continually gravitated toward the most popular digital tracks as sales have continued to rise each year. In the background are the combating forces of declining album sales, rising track sales and increasing revenues from streaming services.