Longing for the day ...

"Debunking the Long Tail"?

The Register published an article some time back about new research that claims to "debunk the Long Tail". It takes a critical mathematical approach mainly to the precise shape of the curve - is it a Pareto curve, as Chris Anderson originally suggested, or is it a log-normal curve? Anderson posted a response here. There's another good response from Benoit Felten, who was at Page's presentation here.

This analysis adds to a growing body of literature suggesting that the Anderson's maths in The Long Tail might be flawed, and, rather gleefully picking small holes in it. Gerd Leonhard posted a rebuttal of this some time ago. Leonhard makes the very good point that no study of existing data can be considered comprehensive, let a lone conclusive, because the data are all flawed by market mechanisms. Page's critique is fair enough and I'm still waiting to find a copy of Page's full report so I can think about it carefully. What I object to is the ridiculous journalistic fixation with "rebutting" the idea of a Long Tail.

Now, to me it's very simple: the music I have released in the New Media environment is selling. Had I tried to put that into old-school retail stores or (God forbid) try to get record company support for it, it would not be selling (I know this because it didn't). I also know I'm not the only one who has had this experience ... there IS a Long Tail effect, no matter what the numbers tell us about aggregated sales. Anyone who claims to be "debunking the Long Tail" is full of shit and probably trying to build a career based on knocking someone else off. The Register has been at it for years because controversy is the cheapest form of publicity they can get trying to get into the US market where Wired (which Anderson edits) dominates ... beat up your rival and people might care.

Whether the tail is as long as Anderson originally speculated, whether it's a Pareto curve or a log-normal curve, whether it applies to sales only or streams and file-swaps as well ... these are, as far as I'm concerned, fairly trivial details that a bunch of propellor-heads will devote the next ten (or more) years to figuring out and will actually mean very little in the real world. Looking at the other of Page's work that is available, he seems to be one of those propellor-heads: all numbers and no meaning.

The new study is also based on fairly limited data, as was Anderson's. Anderson admits this, The Register doesn't. The new study (like a bunch of others with the "debunking the tail" handle) starts from a bunch of assumptions about the Long Tail theory that Anderson has repeatedly said he doesn't make. The Long Tail is NOT as comprehensive a theory as the debunking people like to claim it is. It's easy to disprove a claim that nobody made - Rhetoriticians call this a "straw man" argument.

Yes, Anderson says it's hard to make money in the Tail - but that was always the case. His book is about the value of aggregating the tail where the marginal cost of administering each item is close to zero. At no stage to I recall him saying that anyone could make a living from peddling a single low-selling item within the tail. That's a no-brainer and admitting it does not equate to "effectively admitt[ing] game over for his Long Tail". My Musowiki project is built in the expectation that each page will contribute to a Long Tail of page views - no page or even section on its own will make much difference, but the aggregated effect should (I hope) prove sustainable. That's not possible in any other medium and that's Anderson's point, no matter what the number-crunchers tell us.

For the record, I'm not making a living in my musical niche either but I've opened up a fairly neat revenue stream that I had previously been denied and this has allowed me to increase my catalog and devote some resources to the Holy Grail of niches: promotion. It's pretty obvious, not just from the new study, that there IS still a constraint on the length of the Tail, but that that constraint is no longer shelf space, it's now governed by attention economics - that's where the career is for researchers (and journalists!) in the area. For example, as Felton Benoit points out (and as I noted years ago in my thesis submission), the limited advertising space on the iTunes home page promotes the attention paid to big-sellers at the expense of the Tail. If I was an industry mid-player I'd be ignoring all the Long Tail debate and learning as much as I can about attention economics.

Now, I'm going back to making more music, knowing full well that the Long Tail will work to help me capitalise on what I make and that if I get everything else right I might be able to increase the resulting revenue and build a business in this environment on my own ...