The long tail refers to the small parabolic tail that marks the end of the consumer demand curve. It is an all-but-given fact that in a competitive market with elastic demand, some goods within the market will be in more demand than others. This is often related to the utility of the goods in question. Unfortunately, utility is often proscribed based on limited (and occasionally incorrect) information. Take, for example, the music industry. (I'll avoid my own personal anti-RIAA screed here. This is purely a documentation of a new trend in the industry.)
A| \U2 --- the long tail
N| \ V
At utility point 1 (U1) we have the artists, songs, and albums most in demand - the Nellys, Linkin Parks, and The Rolling Stones of the world. At utility point 2 (U2) we have the artists who are in moderate demand - artists such as The Chemical Brothers, Stephen Malkmus, and Wynton Marsalis, who sell between 100,000 and 1 million units a record would fall into this category. Finally at utility point 3 (U3) we have artists who would be classified as either niche artists (such as IDM specialists Boards of Canada and spoken word activist Cornel West) or artists who have virtually no marketing or publicity in mainstream circles (such as longtime punks Nomeansno and electroclash DJ Felix da Housecat). These people at U3 are part of the long tail.
What is important here is that despite the fact that the demand for products that fell into the U3 was nonzero, many of the retailers upon which the music industry relied upon for distribution did not supply these products. This was primarily due to the economics of physical space - why stock a record or CD that may or may not sell when you can instead stock a record or CD from the U1 demand point? In fact, this frequently led to overstocking of records perceived to be at the U1 point, and a failure to stock anything at the U3 (and sometimes U2!) point of demand.
In short, the long tail was relegated to mail order and diehard dedicatees of unpopular music (in the demand sense), who often operated at a loss to bring U3 music to local areas.
With the rise of the Internet, the music industry has, of course, developed a new business model. This model, which is based on selling digital copies of songs and albums, takes up no physical space. To offer up any track, whether it be a U1 track or a U3 one, takes up approximately the same amount of time. There is no tradeoff of utility here, no mutually exclusive curves in conflict. A label can offer all the songs, all the time, and at a fraction of the cost of physically supplying a retailer with those songs.
So what's the result of this new economic model?
Amazingly, companies such as Rhapsody, Amazon.com, and iTunes are learning that sales of products along the long tail actually exceed those of the high-demand products. Pareto's 80-20 principle does not hold true. According to ECast CEO Robbie Vann-Adibé, 99% of the top 100,000 audio tracks at his site have been downloaded at least once in the past month. That's 99,000 unique songs!
But this is really only the beginning of the long tail. At Rhapsody, they've found an actual nonzero demand for their top 200,000 tracks ... and the top 300,000 .. and the top 400,000 ... and the top 500,000 tracks! What's the lesson learned? "A sale is a sale." This is the new secret of the record industry, and perhaps its saving grace amid its recent anti-consumer lawsuits and constant attempts to perpetrate its own demise. And, amazingly, until the new spaceless model of music selling, this simply wasn't true. The costs of producing a record and generating physical copies of the record were constant, but the results were decidedly on different orders of magnitude.
Consider the following fact: the average Barnes & Nobles store has approximately 130,000 titles. But over half of Amazon.com's book sales come from books outside of its top 130,000 titles! At Netflix, over a fifth of their sales come from outside the top 3,000 titles. The statistics go on and on, and in a fairly logical manner. A lover of niche cinema and underground music is more likely to turn to the Internet to get their fix. Having a model that embraces these niches as equals to the more popular films - and why not, given the low point of entry costs - guarantees an increase in sales. The statistics bear this out.
Even more amazingly, the long tail begats a new breed of music and movie lover. We've now returned to the initial crux of the argument, the utility of a good based on available information. Here on the Internet, the interconnectivity of content guarantees that tastes will be assuaged in more nuanced and satisfactory ways. Searching for Madonna might lead you to Janet Jackson, who leads you to Taylor Dayne, who leads you to The Pet Shop Boys, who lead you to The Magnetic Fields. When Stephin Merritt is just 4 clicks away from Madonna, you know you have a system that works. The signal-to-noise ratio of major studio and major label marketing simply vanishes in the face of widespread popular approval and/or thrashing.
So here is the final suggestion to be successful in the new economy: offer everything. Get the rights to everything you can, beg, borrow, buy, have artists name a price and then accept it without hesitation. The long tail is no longer predicated by the demand of the users, but by the supply of the businesses. Those who take advantage of the new truth will reap its benefits. Ladies and gentlemen: welcome to the age of the endless shopping aisle.
Much of the information and general ideas for this writeup come from an article in the October 2004 issue of Wired magazine entitled "The Long Tail."