iTunes downloads /= internet sales

Many niches have their own websites, social networks and distribution methods. Have these been included in the data?

Also, low sales doesn't mean no niche market. Maybe bands are gaining fans online and reaping profits offline.

Felix, London, UK

I think it's always very obvious to the keen observer when someone just doesn't want something to be true.. Any long term commitment is a good one. This is a battle between greed and patience. The internet is a baby and under estimating it's future will only create a monster.

Dustin, Winona, MN, USA

Mr. Foster didn't explain the data set well. If 80% of all revenue came from a mere 52,000 tracks, with 3 million tracks registering at least 1 sale, that destroys the 80/20 rule to 80% of revenue from 1.7% of product. That's a very big but very short tail! Would be nice to know the % from albums.

Dale Mahoney, Boise, Idaho, USA

Not to be contrary but Chris didn't suppose that the ending for the Island for Misfit Toys song, in which every item produced will have a home, was true. He pointed out that niches could be served profitably because of consumer co-location trends on the web and the resulting lower SG&A costs.

Ted Wright, Atlanta, US

Nowadays long-tail musicians can't rely on mass distributors to sell their music. You just can't be found amidst a sea of other singles out on iTunes.

I don't think this shortening of the long-tail will last. Companies like Audiolife are helping these musicians get their music to their niches.

Ricky, Van Nuys, CA, USA

Let's not forget that amazon has a lot of success with Long Tail selling because it has a good way of helping people find other books in genres they are already reading. "Rock", "Pop", "Alternative" can only go so far as descriptors. It's difficult to find new-to-you music!

Laurie, Boston, MA, USA

The top selling list is 52,000 items long, but in total over 170,000 items have been sold. This seems to back up the long tail theory, rather than critique it. The fact that there were over 1.3m items in total is a moot point. The infinite shelf space the internet offers has enabled this.

Daan Jansonius, Bournemouth, UK

I imagine the shape of the sales distribution will vary widely based on the quality retailer's ability to present compelling choices. Netflix surely rents more long-tail movies than Blockbuster, and Amazon surely sells more long-tail books than a typical bookstore.

Ed Kohler, Minneapolis, MN, USA

Surely the FIRST big challenge to the long tail fad is everyday experience?
That, and the fact that 2,500 years after Croesus, markets have evolved to follow Pareto's Rule (buy from specialists and experts) rather than the long tail (buy from your next door neighbour), because that what people want

Ian Kemmish, Biggleswade, UK

i disagree with Will's premises and conclusions. there is no business model yet to monetise all music traffic across ISPs. there are over 40 million tracks available on social networking sites alone - not the 13 million quoted. How would Will's stats look if all music downloaded earned money?

Robin Millar, London, UK

A point missed here is before download, if you liked only a couple of songs by an artist. Then unless they were released as singles you had to buy the album.
Now you can buy the 2 or 3 songs you like without having to pay for the junk that's left.

Keith, Chester,

When I was a lad record shops focussed on the top 40. That's all they stocked. Over a year about 250 singles. Now Page tells us it is the top 52000. The new tail is longer. But he also shows that the supply tail is even longer again. Due to the new competition niche marketing is tough.

Cam Rungie, Adelaide, Australia

Emusic.com offers millions of tracks, but limited guidance through its endless catalogue - the consumer is left to invest hours of time searching for needles in a haystack. Boomkat.com represents a proper solution, quality control strong enough to mean casual browsing won't be a waste of time.

Rich, London,

Anyone who's delved into the worldwide network of second hand books enabled by Amazon's databases has seen the Long Tail in action. Perhaps the absence of record sales reflects marketing issues rather than a failure of the concept. Still, as saying goes, every exception tests the rule

Mike Tanner, Auckland, NZ

The reason for this (at least in Classical music) is because CDs are often still better value than downloads. I'm happy to pay to download MP3s, but they must be quite a lot cheaper than buying a CD: the MP3 format is lossy, there are no notes, and I can't sell it on if I don't like it.

Alex, Stoke-on-Trent,

Long Tail does exist just give people tools to navigate it and discover (Amazon is the proof). I would very much like for Mr Page to do similar research on the amount of tracks listened at Last.fm. Mr Page's research just shows that Long Tail just still underutilized lacking discovery tools.

Elena Lvova, Moscow, Russia

So what was true last year will always be true? Is that the thesis?

One can't draw long term conclusions about human or market behaviour based on one year's data.

Keith S, Winnipeg, Canada

If 80% of sales came from 52,000 tracks, that means they came from about 4,000 different albums. Well, I'd bet that in HMV's physical stores 80% of their sales come from about 100 albums. Ergo, the long-tale theory has some substance.

Ali C, Warsaw,

Far from disproving the Long Tail theory, this new research actually supports it. The data has been misinterpreted it seems. Of the 1.23m albums available, 173,000 were bought. A typical High Street store only stock the Top 100, meaning a "long tail" of sales online of more than 172,000 albums.

Graham Jones, Reading, UK

The Internet certainly gives musicians the freedom to put there music on line and to be bought, but they forget that the money spent by record companies promoting bands on the airwaves is where the public here the stuff. It's great if you're an already established band not if you're not.

Winston Smith, Wigan, GB