What most people don’t understand about the long tail

The “long tail” is the buzz-phrase par excellence of the new media revolution. No presentation about how media is changing is complete without it. An image of the long tail, taken from our Future of Media Report 2007, illustrating the scaling of business models along the long tail, is below.

From a media industry perspective, the most important aspect of the long tail is that it illustrates an effective doubling of the size of the media market. The tail is as large as the head, allowing both production and consumption of media from small producers.

However, arguably the most fundamental aspect of the long tail is poorly understood by most people who use the term.

The long tail curve describes an intrinsically network phenomenon – it shows the distribution of the number of connections of each node in the network, from the nodes with the most connections to those with the fewest connections.

The familiar shape of the long tail is a “power law” distribution, based on an exponential function. The most fundamental aspect of a power law curve is “scale invariance,” meaning that it has the same shape, whatever its scale.

Indeed, it has been proven that for a network to have a classic long tail power law distribution, it needs to be a “scale-free” network. To be a scale-free network, its nodes must be able to have any number of network connections, from none to essentially infinite. If it is bounded on the upside, then as the networks scales upwards in size, its shape will change as the top end of it becomes limited.

Some networks are not scale-free. For example personal networks are not, because there is a limit to the number of people we can know. Airline networks are limited by the size of airports and physical constraints.

However digital networks, epitomized by the Internet, are scale-free, because there is no limit to the number of connections that nodes (webpages) can have. Media has become effectively a scale-free network, not so much because the top end is unbounded, as because production and distribution costs have become negligible and now many nodes (media producers) have very few connections (audience).

In Chris Anderson’s book The Long Tail, which sparked off the mania for the phrase, the issue of scale-free networks is not mentioned once. Those who have learned from this book and other popular sources do not understand the implications of the nature of the long tail.

People still commonly talk about the “head getting smaller” or the “tail getting longer,” however if it is truly a long tail, based on a scale-free network, this is impossible. The overall shape of the curve will always remain the same. And this, despite what many pundits go on about at length, is exactly what is happening in the media landscape. The overall shape of the curve is not changing. However participants within the media landscape are moving up or down the distribution curve of popularity, or entering or leaving the market. Irrespective of any individual shifts, or changes in market size, the shape of the overall distribution remains the same.

The many implications for media strategy include being able to scale business models effectively, focusing on increasing the value of connections at any point along the curve, and building businesses while accounting for the overall growth in the market. Of course there are many subtleties in what happens within the long tail of media, however most people first need to understand the fundamentals of how the long tail is based on scale-free networks, and as such it does not change in shape, only in size.