The Cicada that Ate Five Dock

Richard Chirgwin has written in today's Communications Day (no online version)claiming that Chris Anderson's Long Tail thesis has been found not to apply in the real world. He wrote in part;

Anderson’s view of the supply side is essentially accurate: the incremental cost of stocking one more song in the catalogue is very small. But user behaviour didn’t conform to long-tail theory: on the Internet, the “head” is bigger and the tail is thinner than in the physical world.**

Without going into too much of the detail, Anderson's thesis is that under the broadcast media model with limited channels available that there was a natural tendency to "hits". Songs that became popular, for example, wre then played more often on radio thus crowding out othr songs. His theseis is that in the Internet model with effectively unlimited channels the hit effect will not operate.

Economically speaking the first effect is what is known as a network effect, or in the phrase used by Carl Shapiro and Hal Varian in Information Rules demand side economies of scale. This refers to any circumstance where the more users there are of a product, the more valuable it becomes. PC operating systems have this feature as the more machines there are running the OS the more other software will be designed for it.

Since the Long Tail thesis first appeared in Wired in October 2004 it has been incredibly influential. Policy makers and others have become sanguine about further concentration in "traditional" media as the thesis suggests that these traditional cotrols will increasingly be a smaller share of the over-all consumption. The rise of blogs and citizen journalists supposedly also fits this pattern.

This, however, conflicts with the observation that platforms in new media seem to have more extreme network effects that old platforms. Amazon dominates on-line book retailing, YouTube dominates on-line short video distribution and iTunes dominates on-line music retailing. Social networking sites most dramatically represent the network effect; the more people who use the social networking site the more valuable it becomes.*

At this point the difference between these Internet platforms and the historic media platforms is that the platform owner is not also controlling the content, in fact, to become a successful platform you want to provide the greatest connectivity to on-line content. However, other factors, especially the potential role of the infrastructure owners, could change that.

Certainly there is lots to be sceptical about in relation to the long tail thesis. It is important that the discussion focus on the distinction between he media content and the distribution platforms.


* There are limitations to network effects though, the first being congestion and the second being that not all users value other new users equally. This latter point was well explained by Eli Noam in his Telecommunications in Europe in Chapter 3 "Network Tipping: The Rise and Fall of the Public Network Monopoly".

** Richard chose to use as an example of the less popular musice a song from 1979 called The Cicada that Ate Five Dock. I thought I'd share it with you - it would take a lot more than the Long Tail to ever get this to challenge Kings of Leon.