The Long Tail: Why the Future of Business Is Selling Less of More
The Long Tail, by Chris Anderson ($25, Hyperion, 2006. 226 pages, tables, graphs, notes, index.)
Book review by Andreas Ramos, Insider-SEO.com
The Long Tail shows how a handful of online companies are using web technology to exploit the vast number of niche markets. Digital storage and digital delivery lowers the costs to insignificant amounts. Peer filtering and peer recommendation allows people to find whatever they want. Companies find that they can offer 1.5 million products, where nearly every item (98%) will sell at least once per quarter, thus generating 25% of their revenues in markets that are unreachable by mainstream offline business. This gives them a competitive advantage over companies that ignore niche markets.
Long Tail Economics
Anderson starts with several examples. eCast, an online music store, has 400,000 songs and 98% of these sell at least once per quarter. The same at Amazon, Rhapsody, and Apple iTunes: they have millions of items and between 95-98% of these sell once per quarter. These companies get as much as 25% of their revenues from marginal items that no other store can offer.
Traditional stores are unable to do this because they work under physical limits. A store on Main Street has a fixed amount of shelf space. They also have fixed costs: rent, salaries, and so on. Their accountants can calculate the numbers down to the dollars per inch of shelf space: every shelf-inch must sell for a certain amount of revenues per year. Otherwise, the store loses money and shuts down.
This applies to nearly every form of storefront business (supermarkets, dress stores, shoe stores, bookstores, record stores, restaurants, brake repair, dentists, and so on): they have a limited amount of space (even a Walmart has a limit; American shoppers will not walk a mile) and costs. They must maximize their sales per square inch. Therefore all traditional businesses concentrate on the "Top 100" products that produce 80% of revenues.
In contrast, digital technology removes physical constraints on a online store. Amazon offers five million books and they could easily offer 500 million books, even if many of those sell only once or twice per year, and still be profitable.
By removing the constraint of selling only the top 100, the new niche markets are created. Anderson describes how digital technologies led to the collapse of mass markets. Movies, TV, radio, newspapers, and magazines once has very large market share but their market share has dropped dramatically. The easy-to-reach large mass markets disappeared. Where the top TV show in the 50s had 90% audience share, the top TV show in 2005 had only 18% audience share. The new technologies (cable TV, the Internet, Sony Walkman, PDAs, MP3 players, etc.) allow people to find exactly what they want. The handful of mass markets turned into thousands of niche markets.
Anderson makes a good deal about statistics, numbers, and percentages. One can discuss whether it's 2% of products that generates 80% of sales, or 100 products that accounts for 80% of sales, and whether these numbers are diluted or concentrated by adding two million more products. Anderson predicts niche markets will grow so large that they will rival mass markets. However, his examples account for at best 25% of their revenues from niche markets. Every one of his examples makes the bulk of their revenues on top sellers.
Lee Gomes of the Wall Street Journal disputes Anderson's theory and numbers (WSJ, July 26, 2006). At the blog for the book (thelongtail.com), Anderson responds that one should look at titles, not percentages. He points out that for Blockbuster, the top 100 titles account for 69.4% of sales yet at Netflix, the top 100 titles account for only 38.8% of sales. Because Netflix offers twenty times as many titles as Blockbuster, it has a longer tail and thus most DVD rentals are in the long tail. But Anderson is confusing the issue. Each DVD in the long tail is only a single hair on that tail. Anderson starts by comparing absolute numbers (100 DVDs vs 100,000 DVDs) and then switches to comparing percentages (39% vs 61%). If we stick to a comparison of titles, Spiderman 3 will earn far more than A Story of Floating Weeds (a Japanese movie from 1934).
This is the core problem with long tail theory: the long tail appears to be huge, but the long tail itself is an artifact of statistics. It doesn't exist. It isn't a market. The long tail is actually tens of thousands of micro-markets, each with unique audiences, unique products, and very small amounts of money.
Selling to the long tail is very cheap per item, but cumulatively, it's extremely expensive. Amazon and Google require large buildings, electricity, cooling, computers, large technical teams, salaries, health insurance, etc. It costs billions of dollars to build the technology to sell to the long tail.
Anderson talks about sellers and buyers, but he pays little attention to a third group: the writers and artists who produce the books, music, and so on. Book authors, musicians, painters, photographers and so on also have their real-world limitations. They must pay their rent and food, buy some clothing, and so on. It's quite nice for Rhapsody to make 25% of their revenues on 1.2 million songs that sell only once per quarter, but a musician can't live on $2 or $3 per quarter.
(On a side note: I don't see why Anderson is using Netflix as an example. I am a Netflix subscriber. I pay a flat $15 monthly subscription and I borrow as many DVDs as I like. Netflix makes the same amount of revenues (my subscription) whether I rent DVDs or not. They actually lose money when I watch a DVD. It's most profitable for Netflix if I don't watch any DVDs at all and just pay my subscription. A rental isn't a sale, it's a loss. Netflix indeed sells DVDs, but this is a side business where they get rid of top-100s overstock. When Spiderman 2 came out, they needed (I'm guessing) 50,000 copies to cover requests from their subscribers. But after six months, the demand drops very fast, so Netflix sells off the surplus inventory. Netflix states on their site they have only a few hundred titles in used DVDs. Anderson's Long Tail doesn't apply at all.)
Who Is the Decider?: Experts vs. the Mob
The physical constraints of traditional markets force businesses to offer only top-100 items. In order to identify the top 100s, traditional markets had editors, experts, consultants, reviewers, researchers, critics, and so on who had expert training (such as degrees in art history, English literature, music conservatories, MBAs, and so on.)
But in the world of infinite niche markets, technology allows users and customers to collectively filter, recommend, and refer products to each other. Amazon tells you that people who bought this book also bought that other book. These recommendations are automated and unsupervised. Users create their own recommendations. No need for elitist experts who assert their narrow political agendas to keep others out of the market.
This sounds nice, but it isn't true. It's a dream of populism that everyone can make meaningful choices. The US population is 300 million people. If it were true that everyone is equal, then we would have 300 million authors or 300 million musicians. However, society isn't "flat" (where everyone is equal). Human societies form into several dozen clusters with experts at the top of each cluster. The 100,000 new books every year in the USA are produced by less than 0.03% of the US population. Only one person in every ten thousand writes a book. This isn't because it's difficult to write books. Blogs require very little work, yet only a few hundred thousand people write the blogs that everyone else reads.
What Is Anderson Up To?
It's a good question to ask why Anderson wrote this book. Chris Anderson, John Battelle, and others all came out of Wired magazine, a magazine that made a great deal of money for a short time in advertising the dotcom boom. Battelle proposes that advertising on blogs will allow everyone to get rich. Anderson thinks that an infinite number of niche markets will let everyone will get rich, or at least everyone will be able to buy the things they want. They are trying to reinflate the dotcom bubble. They call this "Web 2.0", which is the 90s' "new technology will make everyone get rich quick" stuff from George Gilder all over again.
Is this possible? Will everyone get rich? Not really. Long tail economics applies only to a handful of companies. These six to ten companies will be the universal supplier for each digital genre: music, TV, video, books, images, etc.
But maybe six is five too many? They all do essentially the same thing: they sell digital content. So exactly why do we need six companies to deliver the same item?
Back in the early 90s, I wrote a list of rules in computering. One rule: There Is No #2 in Computering. When a product establishes itself as the market standard, competitors disappear. #1 has no effective competition. There's Microsoft Word. What's the #2 word processor? Who cares? As for the operating system, there's Windows. Apple's Macintosh has 1% market share. What's the #2 spreadsheet? The #2 email software? This happens because users want compatibility. Whatever offers the best compatibility (regardless of quality) between various computers (home, office, pocket) and file formats will win. The other products disappear.
Thus the future for long tail economics will be a very short list: it will be only one company and it will get 90% of the market, 90% of the products, and 90% of the revenues. Whether this will be Amazon, eBay, Google, or Yahoo remains to be seen. Amazon and eBay already have the infrastructure to procure, offer, collect payment, and deliver millions of items. Long tail economics won't have long tail retail; there won't be an infinite number of companies selling to the long tail.
Will long tail economics ever become fat tail economics? Or will it be a fringe business for decades, available only to a handful of major dotcoms? The book is about ecommerce, yet ecommerce accounts for only 5% of the US market, and it will be a very long time, easily decades, before it reaches 25%. This means old, boring, dusty offline commerce carried out in brick buildings will continue to be 75% of revenues. Long tail theory won't ever apply to the bulk of commerce.
Anyone going into business has to make the same calculations as the starving musician: either play to a niche market and never get more than a small fraction (with high risk of failure) or go after the big, easy top-100s market with high volume and large market share. For example, if you want to distribute CDs, where would you sell? Online or in a brick store? Walmart offers only 4,500 titles in a very small music section. Of some 30,000 new albums every year, Walmart carries only 750 albums (only 2.5%), yet Walmart single-handedly accounts for 20% of all music sales. Why bother with niche markets, when you can distribute through Walmart?
Digital content has a problem: it is easy to copy. Anderson points out that Apple has sold 42 million iPods and one billion songs (p. 175). That works out to about two CDs of music per iPod. Copying and sharing is rampant among teens and twenty-somethings. As he puts it, this is a very unimpressive business model.
Anderson writes about digital music, TV, and books, but he avoids any mention of the product that is most suited for long tail economics: pornography. Porn is easily delivered digitally, the anonymity and privacy of the web allows private consumption, and there is an almost infinite array of tastes and perversions, such as Furries and Pony Girls. Perhaps his next book will be The Economics of Tail?
As for writers, poets, artists, and musicians, there is little in long tail theory for them. Anderson admits the vast majority of self-published books won't ever earn any meaningful revenues (p. 77). I would add that people should self-publish their books, music, and art, but not to make money. They should do this because it's meaningful to them or they can share good information with a small group. However, that is no longer economics, but a form of self-expression.
I strongly recommend that you read The Long Tail. The book is well-written, with many interesting facts and anecdotes about music, business, and ecommerce. The book will be widely read by the industry and many will accept it as fact. It's important for the large dotcoms to know about niche markets and understand how to go after them. By reading this, you'll understand what they are up to.
Although Anderson's numbers may be off, in general, he is right: because web technology allows ecommerce sellers to carry massive inventory, they can sell to the long tail. This offers an entirely new market. However, there are limits: it's extremely expensive to build the technology. At best, the long tail market will account for modest fraction of revenues (but this is not to be dismissed; if a company can tweak out even a 1% advantage over competitors, it will get ahead). Finally, long tail economics is dependant on the economics of computering, and this means only a handful of companies will compete and survive in this market. Due to economies of scale, they get the profits, but the writers, musicians, and artists who produce the long tail won't get anything.
Chris Anderson writes about logarithms, log charts, and exponentials, but if you want a better explanation of how these work, I recommend Lazlo Barabassi's book Linked (see andreas.com/FAQ-Barabasi).
About the reviewer: Andreas Ramos is co-founder of Position2.com, where he is directs the company's SEO and PPC strategies. He co-wrote Insider SEO, a book about SEO and PPC. He is a Google Certified Adwords Expert and a beta-tester of upcoming Google and Microsoft PPC products and services. He works with long tails every day and helps his clients to market their products and services on the web. Contact him at andreas @ Position2.com