Chris Anderson: Less Is More

By Edward Cone  |  Posted 10-31-2006

Chris Anderson: Less Is More

In early 2004, Chris Anderson, editor of Wired magazine, was asked to estimate how many of the 10,000 albums accessible via a Web-connected digital jukebox had at least one track played at least once per quarter. Sensing that traditional sales metrics, which would indicate an answer of about 20 percent, didn't apply to this Internet-enabled example, he ventured an extreme-sounding guess of 50 percent. "I was, needless to say, way, way off," he writes in his influential book, The Long Tail: Why the Future of Business Is Selling Less of More (Hyperion, 2006). "The answer was 98 percent."

That episode inspired Anderson to research the new rules of distribution and customer choice in the Internet age. In a Wired article, then a blog, and now in his book, he argues that familiar sales charts needed to be redrawn. From Inc. to eBay Inc. to NetFlix Inc., the data bore out his observations: The Internet allows companies to put a far wider range of goods in front of customers than a physical store ever could—and customers respond by buying not just the popular items bunched at the peak of the sales curve, but also the obscure products out on the long tail of the curve.

In fact, the less-popular items out along the tail might sell only a tiny number of units each, but when all these "onesies and twosies," as Anderson puts it, are added together, they amount to a large market. Long-tail products don't replace hits, they replace the monopoly of hits enforced by the limitations of physical retail space. The long-tail phenomenon deals mostly with electronic media and deep-catalog businesses like books and DVDs, but Anderson touches on its impact on manufactured goods and commodities as well. He spoke about the many long tails all around us with Senior Writer Edward Cone.

CIOI: Give us the short-attention-span version of the long tail.
Anderson: It's about life beyond the blockbuster, what happens to our culture and our economy as we shift from mass markets to niche markets. It's the recognition that one-size-fits-all is no longer a necessary model. It never suited any of us well, and now we have the option to treat individuals as individuals, so that one size fits one, or one size fits me. Increasingly, this is not only possible but is demanded by customers.

How have companies reacted to your book?
I expected the biggest opposition from the blockbuster businesses, the entertainment companies, Hollywood and all that. But the book's number-one market is Los Angeles. I was just there, talking to Hollywood companies, and it was standing room only. They totally get it. I thought they'd be in denial, hostile, but it's just the opposite. They're like, "Well, duh, but what can we do about it?"

One of the biggest misunderstandings is that people assume this means the end of the hit. Of course it doesn't. The long tail is a Pareto Curve [a projection that shows, among other phenomena, that 20 percent of products typically generate 80 percent of sales], and that's all about radical inequality, some products selling a lot and others not. Once people realize it's not about the end of the hit, but about the end of the monopoly of the hit, it becomes less threatening, and less counterintuitive. This isn't a theory, it's just measuring what's already happening out there. It's all based on real-world data, and no one's disputed the data. I don't really have to argue the thesis.

There are industries where I'm finding questions about how the long tail applies, rather than opposition. It clearly applies more broadly than people thought. I'm often asked to speak in an industry where I hadn't spotted the resonance. When Anheuser Busch created a division called Long Tail Libations, I realized that people were finding resonance in very traditional industries I had not anticipated. I've spoken to companies involved in heavy machinery, shoes, coffee, agricultural goods, consumer packaged goods. The long-tail concept makes perfect sense to them, so I've been on a serious learning curve about subjects like agricultural machinery, just so I can map the phenomenon to their businesses.

Next page: Seeing the long tail across industries—and markets

1 | 2 | 3 | 4

Seeing the long tail

across industries—and markets">

Do manufacturers need to embrace things like mass customization to get in on this?
That's less of an emphasis than it once was, largely because it didn't really work all that well. It turned out to be limited in its scope. In the 1990s we were looking for the Dellification of everything—and even Dell doesn't do that anymore.

Instead, we now recognize there are lots of products already out there, but they just haven't been easily available. We don't need all sorts of new niche products, because they already exist. We just need to make them available and findable to the customers who want them. And with the recognition that customer needs are all different, a company can make products that buyers can customize themselves. Let the consumer do a certain amount of customization work that isn't really feasible from the producer perspective, but is increasingly easy for customers to do on their own.

In the book you discuss a tangerine-colored KitchenAid mixer and a "long tail of color" that helps that manufacturer woo customers.
KitchenAid offers not just a quality product, but also this extraordinary range of colors. If you're redoing your kitchen, and possibly changing your appliances as a result, you can have a color-coordinated mixer and make a kind of aesthetic statement with your appliances. They are offering a kind of concierge service where your interior decorator can work with KitchenAid and actually give them paint chips, and they do a custom-colored mixer just for your, for some huge amount.

So does the range of colors increase market size, or just market share for a particular company?
In some industries there is a wealth-redistribution effect, where the top sellers lose in sales, and others gain. In other cases, you can actually grow the market. In the case of music, there's more music being made than ever before, and more music being consumed than ever before. The iPod lets you take music into your day, and extend available listening hours. In economics there's something called Say's Law, which deals with the idea of supply creating its own demand. If you produce something, you stimulate demand for that thing that was not previously measurable. Part of this long-tail work is the recognition that demand for variety and choice, which was depressed by distribution inefficiencies, is now measurable and real and large. The biggest thing we're seeing right now is the ability to tap latent demand in the marketplace for variety. We didn't see it before because we didn't know how much choice was out there.

Next page: How does the long-tail concept affect the supply chain?

1 | 2 | 3 | 4

How does the long

-tail concept affect the supply chain?">

Do all these products with low demand levels portend a revival of retailers maintaining massive warehouses and enormous inventories? Should manufacturers give up on just-in-time strategies?
Just-in-time turns out to be somewhat less dramatic than we thought it was going to be, for loads of reasons. Ask Detroit why they don't use the Dell model, and the answer is they're under all sorts of constraints. With cars, the number-one contributor to popularity is quality and defect ratio. And it turns out that the more variety enters the manufacturing process, the higher the defect rate. This outweighs the consumer benefit of choice, since people are not willing to accept lower-quality cars in exchange for more choices.

When we had limited channels, based largely around geographic concentration of demand, we were able to tap the most obvious demands, the top of the pyramid. The distributed demand, the onesies and twosies coming from outside the population centers, was sort of invisible, with no way to measure it or tap it. Now we recognize the long-tail customer outside the top-ten products. We see there's also demand for the next hundred products, but it's just not located in any one place. That demand is turning out to be such a big, interesting, new market that you almost don't have to get into the next level of supply-chain efficiencies, because there's so much demand for the variety of products that already exists.

It boils down to a simple equation: When you lower the costs of distribution you can offer more stuff. We've seen long tails emerge many times in the past, and each time they get longer. The Internet of 1896 was another network, the railroad system. It enabled the tapping of distributed demand with catalogs, and railroads that connected distributed demand to centralized inventory. We've hit the next level, but the new model doesn't require huge warehouses. Amazon does have some, but it's shifting to virtual inventory, with storefronts for third-party merchants. It bears none of the inventory costs itself as it facilitates the sale by giving small merchants global prominence, and giving customers a one-stop shopping experience.

What does the long tail mean for commodity businesses?
The very act of introducing variety to a market decommodifies it. This is why you introduce variety, so you can break the commodity cycle and increase prices, why there's been an explosion of variety on the supermarket shelf. That differentiates the customer, which allows you to charge a premium for some products, and in some cases to grow overall demand, but certainly to shift upstream and not just race to the bottom with an undifferentiated product. Classic commodity thinking is generally found in inefficient markets with inefficient measurement of consumer demand, inefficient distribution to consumers, inefficient production of variety.

We all know that one-size-fits-all is a gross approximation of actual need. But we fall back on it because it's what our markets have demanded. With much more efficient demand-measurement abilities, and distribution abilities, we can start segmenting and dividing the market. We can assume the cost of distribution is falling close to zero in many markets, and that the ability to measure is growing, thanks to an extraordinarily connected marketplace where word-of-mouth is now amplified by the Web.

We now have the ingredients necessary to get to the next level of granularity, to look at the marketplace not as one homogenous mass, but as millions of individuals. Each individual now has the capacity to express their interests, and each producer now has the capacity to measure that interest, so that, depending on the industry, companies can address those interests with different products.

Next page: Measuring the value of the long tail

1 | 2 | 3 | 4

Measuring the value of

the long tail">

What kinds of tools are available for measuring particularized levels of customer interest and demand?
The tools we've been creating over the past decade let individuals express their interests, and we're at the very beginning of developing filters that help sort out these vast markets. The ultimate filter is other people. We trust other people more than institutions. That's always been true, but we haven't been able to find each other. Now in every industry there are blogs, and discussion groups, whether people in the industry know it or not. One thing I hear all the time: "There are no blogs in our industry." And I'll say, "Have you looked?" There probably is a conversation out there and they are just not aware of it.

We give blogs high weight because they're authentic and real and unbiased. Then we get into recommendation engines and collaborative filters, reviews. Word of mouth has gone from a person-to-person, face-to-face experience to something done in public, and it remains public and measurable pretty much forever. That is a fantastic tool for companies. Instead of guessing, and depending on feedback from a fraction of customers, they can access the relatively undistorted, honest feedback that is happening outside their walls among the customers themselves. It's a very measurable marketplace, beyond simple sales.

The lens today is search. We are training a generation—the instinct to use Google to research a company is pretty hardwired in anyone under 30. Google measures links, the expression of respect and interest of other individuals. It's kind of the ultimate filter for every issue. Your brand is no longer what you say your brand is. Your brand is what Google says your brand is, what individuals think your brand is. When Dell can spend hundreds of millions of dollars talking about its great customer service, and a couple of bloggers can make "Dell hell" show up on the first page of search results, the power has shifted from institutions to individuals.

This matters for product development, too. Lego brought in its enthusiast community and ended up with a better product by getting influential customers on board from the start. It's a model that could be applied to almost any company. It's easier and easier to find out who your influential customers are, the bloggers, the guy who shows up first in a Google search. Previously there were user groups, but they were formalized, you had to be a member. Now a business can measure who's influential by simply looking at Technorati. Who out there is really going to determine the success of your products? You can say, come help us, let us share some insights into why we're developing this product, how we're developing it. You can measure and predict what's going to be successful without having to guess.

How does the long tail affect companies as consumers, not just as producers?
One example is software, including the emergence of things like's AppExchange Software has totally suffered from the one-size-fits-all model, and the emphasis on big verticals, with the only customization to satisfy individual company needs happening via extremely expensive integrators and consultants. That model is clearly reaching the limits of its ability to scale. One problem is that it doesn't suit small businesses. Expensive software with even more expensive customization works at the level of large and medium-size business, but even there it's hard to address the individual differences within a company. It's very hard to scale down to the smaller companies that have very divergent needs.

The rise of software as a service allows something new, and not just a marketplace where three developers in India can reach users at a bicycle store in St. Louis. Hosted software, software as service, is tapping the reality of what's been happening all along, with software being customized for unique, individual interests with things like Excel templates and macros. There has been no real way to share that learning within companies, and it's generally been discouraged by IT departments that can't support that kind of variety in their user base. But now individuals can trade best practices and freely share what amounts to customized versions of generic software with each other. We're just in the beginning days of this, but the model is very interesting, in that it really suits the niche customer's needs by sharing a common platform that doesn't require executables and expensive installations, and that improves without any effort from IT administrators at each company.

Now niche suppliers can reach niche customers. When you realize that even the SAPs of the world can't get much more market share in Fortune 500 or Global 2000, growth has to come with the little guys who can't support and can't use the one-size-fits-all approach. The requirement for all of them is to provide a kind of self-service, do-it-yourself, very cheap and easy platform for people to find software that suits their own specific interests, to develop software that suits the interest of a few. The focus needs to shift from a dozen markets of millions, to a million markets of dozens.

Some of these changes will take time. One thing we are learning is that change does not happen at the pace of technology. We assume that change happens at the pace of Moore's Law, but it doesn't. Change happens at the pace of generations.

1 | 2 | 3 | 4