Wednesday, March 8, 2006

So what's the long tail really about?

20060226_iphoto_longtail_1 Over the past couple of years there's been a lot of interesting work to study the effects of an inclusion of economic agents previously left out of specific industries. This set of people who were previously out of the loop is actually a very large group now described as "the long tail". While this echoes my views about the influence of masses of people in today's world, I was wondering whether the long tail is simply another way of naming the mass market: how is the "long tail" different? what sets it apart from traditional forms of retail markets? In fact, I believe that the single most important effect of infotech and the landmark achievement of the Information Age is precisely the inclusion of people who previously had no significant way of  influencing the larger system through their individual actions. That has fascinating implications for open business.



A lot of interesting content and an explanation of the name given to the long tail can be found on a blog dedicated to the long tail phenomenon. It is one of the sources of information I used to educate myself and understand what long tail is all about. Some of my conclusions may echo what is already published on that site and on the Internet; where I have identified such inspirations I will explicitly acknowledge them, else the acknowledgement is implicitly made here.



While I discovered only recently this notion and associated business models that have been floating around on the web, I initially felt reluctant to simply adopting new terminology without probing further into what it refers to exactly. You see I still remember the jabber about the need for a "new economy" back in the late nineties and it eventually turned out that economic theory had sufficient resources to describe what was going on in the information economy. As Hal Varian showed, the classical economic theories were perfectly able to describe and analyze the "network effect", only they called it "network externalities"... And of course, the tools developed by economists of the 20th century that should be used to speak about the networked economy were often "hidden" in footnotes in our textbooks at university of at the business school. The mainstream stuff that was taught was taken by many of us (myself included) as absolute "truths" and I must say I may have not paid sufficient attention to the prerequisites for applying a given economic model. It is much later that I read Hal Varian's and Jospeh Stiglitz's books, only to discover alternative ways of looking at the economy that were deeply rooted into traditional economic theory.



Initially I believed the "long-tail" was just another name for the retail market. Then I realized that there was more to it because of the impact of infotech on the structure and dynamics of business. Technology empowers the masses and gives them access to previously exclusive categories of goods and services: today everybody
can publish thought to reach anyone else on the planet, everybody can
produce music, start-ups like Skype and Jajah can deliver value-adding telephony services... These are not the exclusive hunting ground for large corporation with deep pockets (and I actually have some reservations as to the relevance of large organizational corporate structures in a networked economy, but that is beyond the scope of this post). Examples of long-tail phenomena:



  • an author no longer needs the approval of an editor and a publisher to have their texts delivered online to whoever wishes to read them. She can do that using Typepad, Blogger, Bloglines, Jotspot or any other publishing tool (depending on what the intent of publishing is).


  • a musician no longer needs to have the go-ahead from an art director working for a label to actually distribute music online without loosing their intellectual property rights. They can do that without any individual's approval using jamendo and let the listeners give them direct feedback.


  • an inventive microelectronics or microsystems engineer needs less and less of the large infrastructures of fabs or labs to prototype and even produce novel structures. The time is coming when individuals will be able to "print" pro-grade microcircuits and microstructures.


  • a journalist does not need the approval of an editor or a news agency to report events. They can do that directly on the Internet using blogs or open newspapers like Open Democracy.


  • a software engineer does not need a company or heavy infrastructure supporting a large project to assemble a first class development team and deliver top notch peer-reviewed software.


And as previously excluded individuals enter the loop, several things are happening. Firstly there is a drop in average quality of what is produced and I personally believe it is only temporary because of the learning effects of sharing content and methods with peers. Then the presence of a large number of individual agents who are able to take autonomous action and interact with their peers transforms the very nature of the system to which they participate: the system becomes near chaotic and the butterfly effect becomes effective. Finally, the incumbents react more or less violently as the new entrants put challenge them and claim a share of the activity, thus transforming the economics of the industry or the dynamics of the social and political system.



So what is the long tail all about and what sets it appart from traditional mass markets? How is the long tail a sort of "retail market revisited", a massmaket 2.0 to speak in current fad-language.



  1. volume not premium - the name of the game is to sell large quantities of goods or services carrying a small apparent ticket price (i.e. an offering packaged and priced in a way the buyer perceives as reasonable or affordable). An interesting example of this is the market for SMS on cell-phone networks: an SMS seems inexpensive because it costs only a few cents to send, but relative to the amount of data that is actually transferred it is the most expensive channel one can use at a whopping 900 € / MB sent! The same business model applies in some African countries where a consumer can buy an extra-small quantity of a good (say bread or toothpaste) at a seemingly low price, when in fact they are paying the extra-small unit and extra-high price.


  2. emergence not prescription - collective wisdom (or folly) emerges as a result of individual actions not as a consequence of deliberate design by a central authority. That is what actually makes folksonomy more powerful than taxonomy and I think folksonomy will become a major tool of knowledge and information management in business in the years to come.


  3. individuals not masses - people are empowered to participate as individuals and not required to subject themselves to organized or institutionalized groups. This dimension carries important consequences for the way politics are organized and I expect the power of parties to decline, perhaps even to shrink to the point of allowing some forms of direct democracy to take root as modern versions of the Swiss votation system or the agora of Ancient Greece.


  4. social networks not solitary wolves - the beliefs and opinions of individuals are shaped by the interactions they have with their peers and at a higher level that also means a fundamental change in the way public opinion is shaped. This aspect of the long tail makes it necessary to review traditional public relations models and methods and also influences the ways in which market research and polling are done.


  5. tribes not segments - consumers belong to active and interactive groups that have a life of their own; they are by no means inert categories and that is one of the reasons why segmentation along the lines of demographics cannot yield the desired results as shown in an article  titled "Rediscovering Market Segmentation" published in the Harvard Business Review in February 2006.


  6. peer influence not promotion - consumers are more likely to buy and use goods or services when somebody has provided recommendation rather than when the offering is pushed in their direction by traditional promotional methods. There is also evidence (research of the MIT Media Lab publication I believe but I have lost the reference) that a buyer's propensity to buy a peer-recommended good increases if the person recommending the product or service belongs to the same social group or tribe.


  7. word of mouth not ads - reputation spreads through buzz and individuals' expressed opinions, not through traditional advertising. Buzz can be created with text or fun content like for example a funny ad made available on the web (see one of my previous posts for a case). Traditional ads obey an industrial age mass logic: they are designed for broadcast not narrowcast and their impact is defined by repeated exposure. In a world of scarcity of attention such repetition is difficult to achieve and the effect of one impression is washed off very quickly by other content claiming the consumer's attention. Word of mouth uses the recommendation principle outlined above to build a lasting image for the product or service, although the image is also very vulnerable because any deviation will be noticed (the web is a space of openness and transparence) and commented.


  8. standard deviation not average - looking for an average value or a common denominator for such a large and diverse mass of individuals on deveral aspects of their behaviour in the world is very difficult if not impossible. Generally accepted truths about a tribe will not hold for long and may not even be valid for most of the tribe's members (even though the tribe is defined by shared characteristics). In fact, even if an average characteristic exists there is much more business value in identifying and finding a way to catter to characteristics that are peculiar to a small number of individuals in the tribe (I guess that tends towards the concept of segment of one that was ahead of its time although extremely insightful in my opinion).


  9. recursive not sequential - data that describes the behaviour of an individual or a group, data about the activity of a system like for example project progress tracking, must stem from the behaviour itself rather than be demanded from the user as and administrative prerequisite to provide service. For example it is way more powerful to derive relevant content for an individual by extracting intelligence from the way that individual is tagging content online rather than by asking them to choose relevant predefined categories of content on a web page. And project follow-up would seriously gain in accuracy and relevance if the project indicators emerged from the actual work performed by participants to a project rather than produced based on data specifically requested from participants sequentially (i.e. after their performance of work as planned).


I personally consider this (r)evolution to be a source of major business opportunity for people who care to seek and dare to understand. And the sources of inspiration are quite numerous fortunately: the creators or Jajah, the entrepreneurs behing jamendo, the people running OpenDemocracy, AlterNet or Agoravox, Lawrence Lessig, Joël de Rosnay...



Now, I find exciting the fact that we are dealing with a dynamic market of individual agents able to act in an autonomous and yet relational-interactive way such that emergence is a driving force shaping markets and social systems. That is because I have been researching complex adaptive systems for a while (without going too deep into the mathematics of that field) and because I have devoted ample time to the study of human psychology in the past 18 months or so. Combining the concepts of complex adaptive systems and the science of psychology is exactly the right mix to successfully develop business in this space and in this age.



1 comment:

  1. Good Post! Thank you so much for sharing this pretty post, it was so good to read and useful to improve my knowledge as updated one, keep blogging.

    Selenium with python Training in Electronic City

    ReplyDelete