Misunderstanding the Long Tail

The current long tail discussion at ETech is interesting, but people are still missing something important, something that I think Chris Anderson underemphasizes, or maybe doesn’t even get himself.

We’ve always known there were lots of people not reached by typical companies, but it is not true that companies were somehow ignorant of those potential customers: They just couldn’t afford to sell to them. We have that discussion all the time in venture capital, with partners getting jumpy about portfolio companies selling into consumer markets, mostly because it has always been too tough to reach these people cost-effectively. “Cost of customer acquisition” is the favored phraseology.

But that is no longer the case — in some markets. You can cost-effectively reach infrequent buyers who want fringe products: You do it via customized ads (Adwords) and net-based distribution. And that is what is interesting about the long tail. Put another way, while the long tail is fun and growing, it isn’t new: It’s just that until recently it wasn’t economical to worry about it.


  1. The Long Tail

    I suspect that it is a question of emphasis rather than of understanding. Paul emphasizes here the *demand* side of the long tail. Most of the long tail discussion that I’ve seen focuses on the *supply* side.

  2. I’m not sure whether Chris makes that point, but the point is 100% right. I made the same point recently when talking about blogging being the killer app of community – because the low cost of personal publishing now makes it feasible to go after the long tail:

  3. Misunderstanding the Long Tail

    Paul Kedrosky makes the point today that there is some misunderstanding about the significance of the long tail – he stresses that part of the power of the idea is that the lower cost of some types of online customer

  4. Ben Tanen says:

    Paul, I think you are right on in highlighting this. The truly exciting developments captured in the notion of The Long Tail are to my mind as follows:
    1. digital production and distribution capabilities enabled by commodity technologies make it possible to keep (or put) less popular items “in print” (where in print is defined broadly. Item: print on demand books are ironically somewhat of a laggard still here).
    2. digital communications and relevance technologies (the net itself, web pages, google, RSS feeds) make it possible to efficiently and effectively match “buyers and sellers” of above (where buyers and sellers is defined broadly. Item: early success of eBay — on line commerce and offline goods — is a proof point here)
    In my view, BOTH of these are necessary elements of what is referred to as The Long Tail.

  5. Also, you do it by offering user-contributed content.
    CDDB/Gracenote will have track info for obscure CDs which wouldn’t be there if the content were provided by the major labels.
    Or consider Wikipedia, which covers a far wider array of topics than a regular encyclopedia would, because anyone can contribute an article on their own particular fetish.

  6. Paul,
    First – I greatly enjoy your blog. It’s a daily read for me.
    Second – Your comment about the Long Tail is correct. I’d put it this way: The Long Tail distribution chart doesn’t show what goods are *actually* being supplied, because then the Long Tail wouldn’t appear. Rather, it shows what goods consumers theoretically desire if it was economical for vendors to supply them.
    This has important implications. Chris’s work has focused on the increase in observable demand when the Long Tail is unlocked: all of a sudden, lots of “unpopular” goods are being sold. But the impact on overall demand is less clear, because there might be a substitution effect away from “popular” goods to “unpopular” goods.
    In other words, the unobserved, theoretical demand curve depicted by the Long Tail differs from the actual curve of goods being shipped in two ways: first, as you point out, there’s no observable demand for Long Tail goods. But second, the observed demand for “popular” goods may also be overstated.
    To use Amazon as an example: Amazon unlocked the Long Tail in books, but overall industry book shipments have hardly risen over the last 5 years. That must suggest that there’s a substitution effect when the Long Tail becomes actual.