Debating the Long Tail

I’ve just read an article from Anita Elberse titled “Should we invest in the long tail?”, published in the Harvard Business Review (no direct link, google for it). Based on what appears to be a very rigorous and extensive study, the author reports conclusions which seem to go in the opposite direction of what stated in the famous book from Chris Anderson “The Long tail”.

In brief, these are some of the main observations (but the full article does deserve a reading):

  1. “Rather than bulking up, the tail is becoming much longer and flatter [...] The importance of individual best sellers is not diminishing over time. It is growing.”
  2. When looking at who is responsible for the growing volume of business we see in the tail, they found out that “a disproportionately large share of the audience for obscure products consists of relatively heavy consumers”; as such, “it is unlikely that a truly significant shift in media consumption will take place.”
  3. “Consumers of the most obscure content are also buying the hits. Although they choose products of widely varying popularity, top titles generally form the largest share of their choices.”

So, should we, or should we not, invest in the long tail?

Although I appreciate the thorouogh analysis made in the article, and I like some of the remarks made, we should be critical on one aspect: the long tail theory has become so dominant that we have forgotten what “head” and “tail” really meant just a few years back, when the better supplied shops and markets rarely had more than a few thousand titles. If we now consider “head” the top 10% of over 1,000,000 titles (that’s an amazing 100k!), haven’t we gone in what used to be considered the long tail just a few years back? :)

Before taking one side in the debate, you may want to read Chris Anderson’s thoughts on this article (linked from the article main page).

4 Responses to “Debating the Long Tail”

  1. Matteo Dell'Amico says:

    Here you can find Chris Anderson’s response to the article, as well as a link to the original HBR one.

  2. Matteo Dell'Amico says:

    (duplicate comment, sorry)

  3. After reading the article, I have to say that I find the linear scale of the graphs quite misleading. The main point I got from reading Anderson’s book is that extending storage space to a basically unlimited size accounts for a substantial percentage of the total possible sales.

    Analyzing the data of the 1 million tracks sold from Rhapsody, Elberse says that the “top 10% of titles accounted for 78% of all plays, and the top 1% of titles for 32% of all plays”. This means that tracks #10’000-#100’000 have a 46% selling volume, and tracks #100’000-#1’000’000 still have a 22% market share (by reference, the number of items at a Wal-Mart are around 50’000 in Anderson’s estimations). This is not in contrast with the observation that very few best-selling items account for an equally relevant part of the market (in fact, a power-law distribution explains both phenomena). The point still stands: find a way of offering a lot more items than traditional stores at a very low cost, and you’ll find an important new market share… even if the last decile (#900’000-#1’000’000) of your items sells basically nothing.

    Another point that I feel needs to be addressed is related to the silent assumption of “one business model fits all”: maybe, long tail items are not well-suited to the “sell tracks for 1$” business model; maybe those items can be more profitable if played freely from, say, a webpage, with the business model relying on advertisements for revenue.

    A final consideration is that a phenomenon does not need to raise (lots of) money in order to be beneficial to society: we have many examples of this on the Internet. Since we are researchers and not business people, I feel this distinction is very relevant.

  4. KDD forum just started to discuss the problem