Dealbreaker: "The Long Tail" must have gotten a few smirks when you first coined the phrase. What were some of the runners up?
Chris Anderson: I'm sorry to be such a stats geek, but it didn't occur to me that there were double-entendre implications until much later. The phrase comes out of statistics--the powerlaw is one of a family of curves that are said to be "fat-tailed", "heavy-tailed" or "long-tailed" distributions. It was the length of the tail (the huge increase in available products and choice) that was most important to me, so that was the natural word to choose, aside from the fact that it had the best connotations.
DB: The Long Tail has been applied to everything from national security and fashion to labor and education. Some critics believe these applications don't work, and diminishes the strength of your theory. How do you feel about this?
CA: The Long Tail is both a mathematical construct and a metaphor. In industries where I've done the quantitative analysis and have the data, I've tried to be crisp in defining what I mean by a long tail distribution and showing the evidence. In other industries where I haven't done the quantitative research, I've used the term more loosely to refer to the large number of markets and sectors where there has been a massive increase in "niche" products or non-traditional participants (from the fringe goods on eBay to the amateur contributors to YouTube). In some instances, such as the national security example, I've simply quoted other people who have applied the theory to their own fields of expertise.
Frankly, I think the theory works on both levels: as a crisp mathematical construct and as a looser metaphor. I see the same forces at work in many sectors, even including those you mentioned--democratization of production and distribution, which hugely lower the cost of participation and thus increasing the number of participants. But if people are uncomfortable with the broader metaphorical use, they're certainly free not to use it.
DB: Have you run into any Long Tail applications you’ve disagree with?
CA: Sure, it does get pretty loosely tossed around sometimes. The main confusion is that people sometimes call every powerlaw distribution (that characteristics ski-slope shape in my book) a Long Tail. Although it's true that powerlaws as a family are known in statistics as "long-tailed distributions", that's as an idealized mathematical function. In the real world, I've focused the term on data sets that show the effect of massive increase in product variety (as in music), or intrinsically diverse marketplaces (such as search) with thousands or millions of different "products". If you have twenty products and when you rank them by sales they show a powerlaw, that by itself isn't a long tail.
DB: How do you think web 3.0 is going to affect The Long Tail?
Web 3.0?! I don't even know what Web 2.0 means.
DB: You recently had an online debate about The Long Tail with Wall Street Journal columnist Lee Gomes. What was the biggest point of contention?
In the quantitative examples in the book, I'm very crisp about my definitions and methodology. I define the "tail" as being products not available in the dominant bricks-and-mortar retail in that sector. The advantage of this approach--and over a two-year research project, I explored many methodologies, eventually settling on the one that made the most sense--is that the definition is both rational and consistent. Unfortunately, Gomes used a different definition--he looked at the marketplace in percentage terms (1% of inventory account for X% of sales). Aside from the problem of comparing *my* statements with *his* definition, his method runs into statistical problems, which is why I didn't use it in the book. Basically, as the inventory in the marketplace grows massively, as is the case in something like music (iTunes can add a million tracks a year), so does the denominator in the percentage calculation. That makes marketplaces look even more head-centric, even though in absolute terms the size of tail is growing. It's just a statistical mirage, but it's an easy mistake to make.
DB: The effectiveness of the Long Tail seems to depend on "democratizing the tools of production." But let's be honest—not everyone's that talented. How can find the gold in the long tail without hitting all the junk?
Fortunately, we've got more help than ever. The rise of powerful "filters", ranging from Google to Netflix recommendations and even blogs as word-of-mouth vectors, make it easier and easier to find the good stuff and ignore the junk.
DB: You think that if the internet had been there to democratize the tools of production and distribution back in the 80’s, you'd be the bassist for your old new wave band, Egoslavia, instead of editor-in-chief of Wired Magazine?
I'm sure that my total lack of talent would have stopped me sooner or later.
On further examination of the buzz surrounding The Long Tail, Dealbreaker later followed up with Anderson with a final question:
DB: The reaction to your book has been mostly positive. But there's also been a choir of critics (the long tail of opinion about The Long Tail, if you will). It seems to get harsher the more mainstream the media outlet is. Why do you think that is?
You ask a good question and I wish I had an equally good answer. Obviously the marketplace has spoken with its pocketbook and the sales of the book have exceeded our most optimistic projections, so that's the "review" I value most. Broadly, the objections, such as they are, have broken into two camps: one that says that the theory's wrong and another that says it's obvious. That tells me I probably got the balance just about right.
Fortunately, the book is based not on speculation and hypothesis but on real-world data, extending over several years. People can quibble with my conclusions and disagree about the causes. But at the end of the day, the data speaks for itself. I think that's the real reason The Long Tail has resonated so strongly--it's simply giving a framework and name to a phenomenon that's clearly visible all around us. That's why it's hard for anyone to really argue that it's wrong, and as for the other objections, well, all the best ideas are obvious once you hear them!