ALA TechSource Logo


curve Home spacer Publications spacer Subscribe spacer Blog spacer About  

The Long Tail Wags the Dog

Submitted by Tom Peters on July 7, 2006 - 11:18am

Graph of a long tailOn June 26th I caught the tail end of Chris Anderson's standing-room-only talk in New Orleans during the ALA Annual Conference. Chris was discussing the "long tail" phenomenon in the new business-and-economics climate of the Internet age. His book on this topic was just published, appropriately titled The Long Tail: Why the Future of Business Is Selling Less of More.

The long tail in a nutshell: when you give people hundreds of thousands or millions of choices—in books, music, movies, or other consumables—then create a graph, where the available choices are plotted (in descending order of popularity) on the X axis and the popularity of each item is plotted on the Y axis, the result is a rapidly declining, swooping line with a "long tail." In other words, a relatively small percentage of the available items will be very popular, but most of the items available will receive little use.

For years—decades, centuries—businesses focused on the head, not the long tail. Because order-fulfillment and distribution costs were relatively high, it made more sense to focus on the small percentage of high-demand items than on the larger percentage of low-demand items. Wal-Mart, which may have taken the traditional mode of retailing to its logical—some would say obscene—conclusion, fills its big-box stores with "heady" items. For example, the typical Wal-Mart stocks only a few thousand CD titles and probably only a few hundred book titles.

Anderson points out, however, that several innovative companies, such as Amazon, Netflix, and Rhapsody, have taken the long tail seriously and have discovered that anywhere from twenty-five to fifty percent of their total sales come from "taily" items when the long tail is well-groomed.Long tails on a short dog

It seems to me that for a long-tail enterprise to succeed, at least two essential ingredients are needed:

  1. The legion of customers must both perceive, and find, it easy to discover and experience items in the long tail. There must be a broadly experienced sense that millions of items are there for the picking, each equally available with just a few simple clicks of the mouse.
  2. The business must develop an easy, quick, and inexpensive means of order fulfillment.

One fascinating thing about this first wave of long-tail businesses is that—although computer networks, search engines, Web browsers, and secure e-tailing systems have helped them achieve the first essential ingredient—order fulfillment usually still involves shipping physical items in boxes. Efficient, effective, competitive courier systems are still essential to the early success of long-tail businesses. In fact, courier systems may be to Amazon, eBay, NetFlix, and others what the coal and steel industries were to the automobile industry in the third quarter of the 20th century. If fuel and distribution costs in general continue to rise, that may be the Achilles heel of these long-tail businesses.

If it is true that long-tail business models have entered the mainstream world of commerce and are here to stay, what does this mean for libraries?

  • Libraries were into long tails before long tails were cool. Any library stocking more than a few thousand titles (i.e., the vast majority of libraries) knows all about the long tail. In fact, most large libraries have collections that extend far beyond the utmost limits of the longest tail. In other words, many items in their collections have not been used since added.
  • Perhaps some libraries, in an effort to boost circulation statistics, have focused too much on the "heady" end of their collections. Rather than cater to the clamorers for Koontz, perhaps libraries should cultivate more long-tail usage.
  • If the long-tail phenomenon is here to stay, perhaps the 80/20 rule (that 20 percent of the collection accounts for 80 percent of the use) will become increasingly suspect. Anderson documents this in his book.
  • I still wonder if more libraries should experiment with patron-driven selection—point our users to millions of available books, perhaps through a mash up of information cobbled from WorldCat, Amazon, publishers, book jobbers, etc., then a user could actually purchase (or libraries could even lease access to) an item only when a user expresses the interest or need to use the full text. While we're at it, let's let them decide what format they prefer: p-book, e-book, audiobook, etc.
  • Perhaps librarians need to re-examine how we go about achieving the first essential ingredient for a long-tail enterprise. When the massive digitization projects come to fruition, it may become increasingly myopic to focus lots of resources on the local OPAC. The new rallying cry may become "OPACs suck, and I don't care," sung to the tune of "Jimmy Crack Corn."
  • We also need to concentrate on the second essential ingredient. Let's face it, currently most ILL systems and services are not going to take us to the promised long-tailed land. They are just too unwieldy and inefficient for both the users and libraries.

Technorati tags: audio books, audiobooks, E-Books, ebooks, Long Tail, longtail, libraries

Comments (12)

The refrain to that song

The refrain to that song would be, 'We're going to digitize!' Seems to me another angle to this long-tail-thang is the concept of low overhead and fast delivery (e.g. service) through colocation. We have half a tail, in terms of ownership; we don't have the other half, which is the ability to offer these items very quickly to a huge public. In that case the local OPAC is therefore not even just sucky, but an obstacle, as I think you're suggesting.

Good point, Karen. That huge

Good point, Karen. That huge sucking sound may be the imminent rapid decline of retail shopping space as the primary locus of retail transactions. I've never been in retail trade, but, compared to delivering goods and services out of a warehouse or online, the cost per sale has to be pretty high. I don't see much future in shopping malls, but perhaps that's just wishful thinking on my part. Ryan, yes, I agree. As I was working on this blog post, I discovered lots of thoughtful essays out there about long-tail business and economics, which seems to be about as far along as the library 2.0 concept in coalescing. There are lots of good riffs on the 'long-tail-thang' as Karen puts it, which I perceive as good and healthy. I agree too, that libraries have lots of 'cultural capital' right now. Most people like and trust libraries. If we could somehow measure and plot cultural capital, would it currently be rising, holding steady, or declining? I hope we don't squander that CC in the next few decades. We need to monitor these business trends and technological developments, then apply them, blended with our strong sense of ethics and social responsibility, to libraries. Mark, yes, many of the pieces already are in place. But the overall experience is still pretty klunky and spotty, IMHO. Should libraries begin to merge, because the costs of delivery (especially electronically) are dropping far below the costs of building and maintaining physical libraries, which I suppose can be considered in the context of this discussion as a type of retail space? I don't know. Let's discuss!

Tom, Good stuff and reflects

Tom, Good stuff and reflects my thinking too. On your last bullet point. I think there's a real chance to improve the whole delivery thing. We're right on some kind of cusp in libraries. On the one side there's the more traditional ILL systems, which have been getting faster, but it's still days or weeks into the patrons hands (and as old studies say, anything over 10 biz days and the patron has forgotten they ordered it). On the other hand we have numerous shared catalog systems (here in WI an example is the 6 county South Central Library system with 50+ libraries) where reserves and delivery are one day or less for materials on the shelf. And with 50+ libraries those 'long tail' items are on the shelves. So we CAN do the delivery part in regional settings within a shared catalog environment. We also CAN identify 70+ million bib items and their locations imediately. The trick is to merge that delivery approach to the identification approach. NCIP??? One huge shared OPAC for the whole world? 'Discuss among yourselves'. .

I think the 'long tail' is

I think the 'long tail' is being mashed with another 'long tail' in the world of economics -- the Long Run Average Cost curve (LRAC), which shows average costs on the y and production on the x. The LRAC used to be a sort of 'U' shaped curve -- meaning that it is good to expand production up to a certain point and then, because of management etc. it gets more expensive to expand (and so you don't). Because of the productivity benefits of IT, very low costs, network effects and endless possibilities for new and improved markets, I think the curve has gone from the U to a 'long tail' as well. This means you can expand about as much as you want and always get economies of scale. Can you say 'Google' anyone? The LRAC 'long tail' enables folks like Amazon to focus on the 'long tail' of products and still make a profit. It is a very exciting and scary proposition. Exciting because it means better services for more people. Scary because it means that the market may be able to handle many of those things that government services used to handle. The call for libraries may be a call for ethics and social responsibility in this world.

This is a good discussion. I

This is a good discussion. I agree with Michael that libraries should be experimenting more with other ways of getting the right book to the right reader: patron-driven acquistion, POD, digital books, etc. This comment from Michael got me thinking: '...that thinking ignores the reality of local politics -- public libraries and the communities they serve want that top 20% of materials ....' If the communities served by public libraries want the 20 percent of 'heady' (popular) materials, who's buying/renting all this taily stuff from Amazon, NetFlix, and others? Michael, are you saying that communities don't want taily stuff from their local public library? Let me throw out a very tentative thesis that I've been mulling over: The 'localness' (of funding, politics, usage) of most libraries (public, school, college and university, special) may actually hurt libraries more than help them in the next few decades as more people become accustomed to using and expecting global information/communication networks.

Oh, I agree that older and

Oh, I agree that older and less popular titles comprise the long tail, but I think those titles were never popular even when new, which is the example I was trying to illustrate. Many libraries, even cooperatives, cannot purchase unpopular new titles, let alone house old unpopular titles. The numbers are simply too huge. Old and formerly popular titles, maybe, but old and relatively unpopular titles, almost impossible for the libraries in my region.

It's worth asking if some

It's worth asking if some libraries do really well at the medium tail, which is what I see Mark Beatty talking about: the 24-hour-window of second-tier items. Out here, we have almost no conception of that--it's far too feudalistic--but I've lived in states where consortial sharing was taken seriously. Also, Michael C., in an article I just read about the Long Tail, older and less popular (or should we say niche) movies comprise that long skinny part of the tail. But I agree the acquisition model needs to change--as does the circulation model! How many decades before people look back and ask, you mean libraries bought multiple paper copies of books... in the thousands no less?

I'm not convinced that

I'm not convinced that either large libraries or large shared catalog systems are the answer. If that were the case, Amazon would not have the huge long tail advantage that it currently has over such very large chain bookstores as Borders or Barnes and Noble. While it's easy to say that a large system of cooperating libraries can stock the long tail of titles, that thinking ignores the reality of local politics -- public libraries and the communities they serve want that top 20% of materials and they don't care to wait 24 hours, not to mention waiting a week. And ILL is far to cumbersome and expensive a beast. I also think we need to be careful not to confuse older titles for long tail titles. To use books as an example, the long tail is comprised of new books that have small publication numbers. The long tail isn't old, formerly popular, titles. No matter the size of the cooperative library system, it cannot hope to stock that long tail of titles. I'm convinced that in order to properly serve customers in that long tail we will need to radically change some of our acquisition models. The idea of pushing purchasing to the library customer may be one approach worth trying. Printing on demand, used book purchasing in place of ILL, and digital books are also good places to begin.

Libraries may store the

Libraries may store the things in the long tail but if you look at circ data it's likely very similar to normal stores. The problem is the stuff in the long tail isn't always easy to find. The big thing that is changing the 'long tail' is search. Amazon and the rest take advantage not just by fulfillment but by fixing their search so it helps people find these things. Libraries are a bit behind. O'reilly is also moving to take advantage of this and have shown it with their Safari offering. Here's a post on it: You should also probably read Lorcan's post on the long tail as it probably gives a good overview of where libraries fall short:

Tom, I agree that our local

Tom, I agree that our local users are buying that taily stuff (love that expression) from Amazon and Netflix, but it's in diverse and small numbers, probably not sufficient to justify the library buying it, and there's our problem. I really think local libraries are hostage to the community's desires (not that that's a bad thing, it's just our reality). We can choose to stock unpopular items (taily things) but it will have a bottom line impact. Obviously this is no place to have an argument about the purpose of the library, but in the end we need to be at least popular enough with our local citizens so that we get funded, and the more people that use us the better politically situated we will be. There's a balance we must shoot for, and hopefully everything we choose to do will fall within our mission, but the local political realities cannot be ignored.

Yes, Jim, I seriously wonder

Yes, Jim, I seriously wonder if the historical bread-and-butter business of libraries--providing books to a local service population--can survive what seem to be some fundamental changes underway. Perhaps we should emphasize the quality of the services provided by localized libraries. The rub there, however, is that increasingly quality of service in offline and online environments is determined (at least in part) by how well the service provider is able to finesse user demographic info and user behavior into a truly useful personalized experience--while still preserving patron confidentiality. Even on this front, librarianship faces some serious challenges. When Google was negotiating with the G5 libraries (Michigan, Stanford, etc.) and perhaps with others as well, if the G5 libraries had had the will to pose this question to librarianship as a whole (or even to the core leadership of librarianship), how would they have initiated, conducted, and conluded that discussion within our profession? One basic problem here is that it is Google (billions in ready cash, lots of talented employees) 'against' libraries, which often are fractious fiefdoms.

Google and Amazon are

Google and Amazon are already positioning themselves to be able to provide 90% of the books ever published (the long tail for books) as ebooks. They are scanning every book they can get their hands on, and will continue to do so. As Kevin Kelly reports in his recent NYT Magazine article, 90% of books are either in the public domain (15%) or out of print but in copyright (75%). All it will take will be some good 'orphan works' legislation passed by the Congress to free up the OP books, and Google and Amazon will be in business serving up the long tail of books at modest prices (perhaps for free in the case of Google). As for libraries, it looks to me like we missed the boat. When Google came knocking on our doors, we should have said, 'No thanks, we'll do it ourselves.' But we didn't and now it may be too late. Public libraries have thrived for the past 150 years on the scarcity of books. For decades, books were just not available, or very hard to get, especially in rural and small town America. Now they are easier to get, but still scarce in the sense of being too expensive (a luxury item for most people). When all books are easily and cheaply available (probably within a shorter time than you think), public libraries will need to have found some new ways to add value to their communities, because our bread and butter business will be on the wane. The long tail is not good news for libraries.