July 26, 2006
The Wrong Tail: How to turn a powerful idea into a dubious theory of everything
Tim Wu in Slate:
Chris Anderson's The Long Tail does something that only the best books do—uncovers a phenomenon that's undeniably going on and makes clear sense of it. Anderson, the Wired editor-in-chief who first wrote about the Long Tail concept in 2004, had two moments of genius: He visualized the demand for certain products as a "power curve," and he came up with a catchy phrase to go with his observation. Like most good ideas, the Long Tail attaches to your mind and gets stuck there. Everything you take in—cult blogs, alternative music, festival films—starts looking like the Long Tail in action. But that's also the problem. The Long Tail theory is so catchy it can overgrow its useful boundaries. Unfortunately, Anderson's book exacerbates this problem. When you put it down, there's one question you won't be able to answer: When, exactly, doesn't the Long Tail matter?
The graph below [on the right, here] is the Long Tail in a nutshell.
This image accurately describes the demand for cultural products. In most entertainment industries (films, music, books, etc.) a few hits make most of the money, and demand drops off quickly thereafter. Demand, however, doesn't drop to zero. The products in the Long Tail are less popular in a mass sense, but still popular in a niche sense. What that means is that some businesses, like Amazon and Google, can make money not just on big hits, but by eating the Long Tail. They can live like a blue whale, growing fat by eating millions of tiny shrimp.
This insight goes only so far, but like many business books, The Long Tail commits the sin of overreaching.
More here.
Posted by S. Abbas Raza at 11:42 PM | Permalink
TrackBack
TrackBack URL for this entry: http://www.typepad.com/services/trackback/6a00d8341c562c53ef00d834a570ce53ef
Listed below are links to weblogs that reference The Wrong Tail: How to turn a powerful idea into a dubious theory of everything:
Comments
[Disclaimer - I've read the original long tail article, not the book]
Tim Wu's article is a good read. My only complaint is that it doesn't go far enough. Wu points out that by claiming that long tails are wagging everywhere you look, Anderson overreaches and threatens to make his catchy phrase meaningless. And I agree. But Wu seems to agree that there is a big area within the overlap of cultural/enterntainment goods and information technology where the Long Tail is operating. The image, he says, "shows, graphically, the difference between the mass culture we've had, and the folk culture we're bringing back."
But are we really bringing back folk culture? I believe (but don't have the figures at hand) that in recent years overall sales of books have been roughly flat, the number of titles has increased, and the sales of the top few have increased. The sales of those in the long tail have, if my assertion is correct, decreased. More Harry Potter and DaVinci Code and less other books. The mere presence of a book on a bookshelf, whether virtual or real, does not make it a sellable item. Some markets follow not power law curves but exponential decays, and in those markets the area under the tail is small.
Anderson's article, and book, has its cultural roots in open source software and perhaps the EFF. The core belief is that we can get rid of those gatekeeper/middle-person/censor types and get unmediated publication, and that wisdom-of-crowds type information sharing will take care of spreading information. I wish I could believe this, but I can't. There are informational flaws and principal/agent problems in "unmediated" publishing just as in mediated publishing. Instead of the gatekeepers controlling the publication process, they control the process by which people find out about published works.
An example from the open source world is the huge success of the MySQL database product, despite its marginal-openness. Meanwhile, other open source database products (PostgreSQL, Firebird) are marginalized. Why? Because there are increasing returns when it comes to information about these products. For all kinds of reasons, people want to work on and with popular products. There is no long tail in open source databases, or in many other areas of open source software.
But it is a great title.
Posted by: tom s. | Jul 27, 2006 12:43:22 AM
I believe (but don't have the figures at hand) that in recent years overall sales of books have been roughly flat, the number of titles has increased, and the sales of the top few have increased. The sales of those in the long tail have, if my assertion is correct, decreased.
I believe this is not true when you consider sales of used books -- a true long-tail phenomenon. Now publishers of new books have to compete much more with sellers of used books--which used to be bought only by frequenters of 2nd-hand book stores but are now easy for everybody to discover and buy.
I think the best way to understand where the long tail does and does not apply is with the iPod and iTunes. The marginal cost of adding additional songs to iTunes is very low and so the long tail applies. But the long tail absolutely doesn't apply to the mp3 player market where the competitors to the iPod are relatively few in number and struggling to hang on (and several have given up and exited the market).
Posted by: Slocum | Jul 27, 2006 8:24:18 AM
"I believe (but don't have the figures at hand) that in recent years overall sales of books have been roughly flat, the number of titles has increased, and the sales of the top few have increased. The sales of those in the long tail have, if my assertion is correct, decreased."
There are serious problems with the book publishing industry, but this example seems to miss the point of the "Long Tail" argument. The argument (and I, too, have only read the article) addresses the issue of the variety and depth of niche markets, and not necessarily the volume of sales of individual products within those markets. Therefore, if the hypothesis that the number of book titles has increased in recent years while overall book sales have remained flat is accurate, then it would actually appear to confirm (rather than contradict) Anderson's argument--suggesting that book sellers have found ways to make a profit with each title (outside of the big hits) selling fewer volumes.
(I am not at all certain that the hypothesis itself holds true in academic publishing, where the trend in recent years appears to have been to try to cut back on numbers of titles being published).
Posted by: crojas | Jul 27, 2006 10:44:00 AM
Post a comment