by Mike Shatzkin

This conference is billed as a repeat of the June conference, when I delivered a speech called "What If?: Surprises Your Competitors Could Deliver Tomorrow Morning". Perhaps the surprise this morning is that I've changed the speech to a new subject, "Selling Books on the Internet". I hope the urgency of addressing this subject will be evident to all by the time we conclude. But we can start today the same way we started in June, with the story of the rise of the Ingram Book Company in the United States. We told this story in June to illustrate how a company could gain an enduring advantage by "preparing the surprise", to use the Ioan Tenner paradigm that we incorporated into our thinking when we prepared the 1996 research paper you are receiving today. We tell it today because it sets the stage for what we see as the big and imminent changes in international publishing that will be engendered by bookselling on the Internet.

This story takes place in the early 1970s, approximately 1972, when Ingram first employed the microfiche reader to provide better service to their bookstore customers.

Because of the sheer size of our country, wholesalers had a business for years filling reorders for the fastest-moving books. Before the Ingram microfiche changed the industry, wholesalers usually didn't really try to stock much breadth of titles beyond those fast-moving books. And the business climate between bookstores and wholesalers was characterized by mutual frustration and disappointment.

Wholesalers would find that they had in stock, and could therefore make money on, only 30% or so of what they received orders for. And bookstores, in those pre-digital days might have to wait a week or more to find out which 70% of the order they weren't going to get. This was normal. This was the landscape in which everybody lived.

Harry Hoffman was at that time the new president of Ingram Book Company, a little and, frankly, insignificant wholesaler in Nashville, Tennessee. Hoffman had once worked for the electronic hardware manufacturer Bell & Howell, and one day a former colleague of Hoffman's stopped by Ingram to discuss new applications for the microfiche reader, a clunky slide projector kind of machine that Bell & Howell believed had untapped potential with commercial applications. The value of the microfiche reader was that it permitted a huge amount of information to be stored on a very small piece of film.

Hoffman made the connection that this machine, if stores would use it, would permit Ingram to tell customers what was actually in stock on a weekly basis. So it would permit stores to select from what was really likely to be available when they placed orders, adding efficiency at both ends of the chain which should more than repay the investment in the machines.

As the business model actually evolved, the stores paid a nominal rental for the machine, which covered the hardware cost. And the publishers were charged a small fee for having their titles appear on the microfiche, which permitted Ingram to make a profit on the production and distribution of the weekly microfiche updates.

Michael Zibart of BookPage, who was Ingram's buyer at that time, remembers the ABA national convention when they first showed the reader to booksellers. He told me that they did a mailing before ABA and got about 100 responses, which was a lot. Then every person at the convention that saw it said they wanted it. Ingram knew immediately they had a winner.

Ingram's innovation propelled them into the top ranks of America's trade wholesalers. They are now preeminent. Ingram has continued to innovate with service improvements, becoming the key driver to push EDI use in the book trade and becoming the first to be exchanging inventory status and order information online. Although many smart things have been done by Ingram since the microfiche, they have been building since the early 1970s on the edge they gained when they became the first book wholesaler to offer timely information to their customers about what books were actually available.

There were dozens, if not hundreds, of wholesalers in America bigger in the trade market than Ingram was in 1970. There were none by 1980, and there are none today. That was their payoff for preparing the surprise.

And, at the same time, they permanently changed book wholesaling in the United States. Title breadth became essential to compete. Only Baker & Taylor, which was already the largest wholesaler in the States before the Ingram microfiche, has really kept pace with Ingram's hundreds of thousands of titles in stock on a consistent basis, although there are other important players. The result is that American bookstores can purchase most of the moving titles of most publishers from centralized resources, getting fast service and the convenience of consolidated ordering and delivery.

We will see how this capability will create a literally globeshaking impact on online commerce.

We were not aware, when our office first discovered and ordered a book from Amazon dot com in August 1995 that the site was then only one month old. There were at that time already dozens, if not hundreds, of bookstores that had created Web sites. We were immediately impressed with Amazon's searchable database, its very user-friendly and intuitive interface, and its obvious service attitude. Our order was confirmed by email, we were told how long it would take to get the book and delivery met the expectations they'd created.

Within a few months, Amazon was creating a sensation. According to the company's own press releases, sales have been increasing by 35% a month since inception. Outsiders have pegged Amazon's annual volume now at $17 million, from a standing start in 14 months. They have just received up to $10 million in additional capitalization.

Jeff Bezos, who created Amazon.com, comes from a Wall Street background. He had the business acumen to appreciate the enormous leverage afforded the online bookseller by Ingram's inventory and service. He selected Seattle as a home base to put Amazon virtually next door to Ingram's biggest warehouse. That gives him instant access to the books Amazon needs to fill most customer's orders. It also gives Amazon an inventory turn of infinity.

The virtual bookstore stocking concept Amazon employs has been matched by Bookstacks.com, which has been online since 1991, and Bookserve dot com, which is owned and operated by two sons of a former Ingram CEO. Bookserve is in shouting distance of Ingram's Nashville warehouse.

All three employ the same basic strategy. They offer a huge title database of available titles, a million or more from Amazon and Bookserve and over 400,000 at Bookstacks. None -- or practically none -- of the books are actually in their possession. They rely on Ingram, primarily, and then other US wholesalers and then, ultimately, the publishers, to get them the books. Their initial competition was to have the largest searchable database and the most user-friendly interface. Now they are competing to provide "service", which also involves defining what "service" is in this new frontier of bookselling.

Amazon has clearly the led the way, in p.r., in innovation, in ease of use. But all of the cyber-only bookstores we spoke to report sales growth at a rate roughly the same as Amazon's. And each has begun to define its own service personality. Bookserve, for example, has set up enormous databases and sources of supply for books in four foreign languages.

One of the earlier pioneers of online bookselling is Australian Simon Verdon of Open Communications, who began selling computer books from down under two years ago and opened a US office 18 months ago, some months before Amazon opened its site. Verdon reckons his business has grown by a compounded rate of 25% a month since he began in the Internet's stone age. Verdon was the first to use the virtual bookstore supply model on the Web. He told us he started Open Communications with an investment of about $1000, has fixed costs of under $100 a month to keep his Web page up on the server, and he pays for the goods he sells after he collects from his customer.

Open Communications' value contribution is its computer book specialization, to which it strives to add every day. At Frankfurt, for example, Verdon was looking for ways to sell out-of-print computer books and manuals, perhaps as file downloads. That value contribution adds up to a pretty good business model. It will be duplicated endlessly.

Similar growth patterns for online sales are reported to us by the independent bookstores that have developed a Web presence. The stores we've found, both in the US and in the UK, have relied on their own inventory for their Web business; they haven't expanded to include "virtually" everything, which they can, of course, easily do. We expect they will, although that realization is not yet prevalent among them.

Amazon as a retailer and Baker & Taylor as a wholesaler have tried to distinguish themselves from their competitors as the entity that will go out and get any books a customer of theirs wants. The degree to which the distinction is either useful or valid is somewhat open to question. If the last piece of Internet bookselling that isn't precisely turnkey is doing the special ordering with myriad publishers at the wrong end of the 80-20 rule, it seems likely that both wholesalers will offer a solution before long. Particularly when the collective system wakes up the potential to charge a "search fee" for hard-to-find books as a way to compensate for special handling. In effect, Amazon does this now by not discounting those books, or not discounting them as deeply.

Wordsworth in Boston has a site even older than Verdon's featuring their database of over 100,000 titles. They claim to have had one of the first 200 sites of any kind on the World Wide Web. Over the last 18 months online sales have grown at least 25-30% a month. Online sales now constitute 30% of their mail order business and 10% of the store's overall volume, and we're talking about a big, well-established store.

Powell's Bookstore in Portland, Oregon is a local chain of differentiated stores: one enormous one specializing in used books, one for technical books, one general trade, even a travel bookstore. They began a Web presence with the technical bookstore's database and have been adding in the others. Of course, their used book database adds a dimension other online stores can't match. Powell's reports online sales increasing at 50% a month over the 6 months before we spoke in September.

We had an email exchange with Willie Anderson of John Smith Bookshops in Scotland. They've been selling online for quite some time, particularly in the library market. They are attributing millions of pounds a year in business to their cyber-channel. Anderson stressed an attention to prompt response to email; they seldom allow anything to age as much as half-an-hour.

Britain's Internet Book Shop has always strived to present the widest possible array of titles, a la Amazon dot com, even prior to Amazon dot com. However, the original supply model of referrals to stores was too cumbersome for the consumer to accept. IBS has gone to the Amazon supply model as well in the past few months, with some rather startling immediate implications to UK publishers and their relationship to UK booksellers.

And, in the States, Barnes & Noble and Borders have announced ambitious Web plans to be unveiled sometime next Spring. They're very late, but they project the impression that they're spending the time and money to extend their book retailing dominance into cyberspace.

And every ground store we talked to with a cyberpresence finds itself investing more and more resources to keep up with their burgeoning online business, meaning it will grow even faster. Both Powell's and Wordsworth have as many as 15 people dedicated to their online sales effort at the moment, half of them simply filling the orders.

So this is a mushrooming market. Ingram knows that too and they also know they'll inevitably be the biggest beneficiaries of its growth. So they've been working on some things that are about to make it grow much faster.

What Verdon, Amazon, Bookstacks, and Bookserve demonstrated is that there are only two essential ingredients to run a bookstore on the Net: access to Ingram and other U.S. wholesalers, and a searchable database for the consumer to "shop" from.

The database question is not entirely straightforward. Ingram uses Bowker's "Books-in-Print", which it does not have the right to sell on to stores. Baker & Taylor has built their own proprietary competitor to Bowker, so they can offer their stores access to the database. However, so far charging for the database has not been the wholesalers' path to profits; they win by providing the books. As of our last conversation prior to this speech, it seemed Baker & Taylor had not solved the conundrum of how to prevent its database from being used to sell Ingram's books.

Ingram, with 350,000 titles in warehouses across the US, sees the value of building a bridge to make Internet bookselling from a virtual supply model simple and easy. Baker & Taylor is disadvantaged by inventory that provides less reliable supply, but they see the same thing. Both wholesalers are developing a variety of turnkey options to put title databases and their inventory at the disposal of Web booksellers.

Many observers estimate that we'll see $1 billion in book sales on the Web in the next 3-to-5 years. Of course, if it is $1 billion in three years, it may be $3 billion in five years.

To give some perspective, we'd estimate the total US book market to be about $30 billion at retail. That says the total online business will amount to somewhere between 3-and-10 per cent of the size of the US market in the next five years.

Servicing the Internet retailer should be an item at the top of the strategic list at every publishing house. I'd like to discuss three aspects of Internet retailing this morning and make some recommendations.

First, we'll spend some time exploring the vexing conflicts between working with intermediaries and pursuing direct sales in the world of online book commerce;

Second, we'll take a look at what Internet book retailing means to international sales and the consider the impact on the current practices of dividing territorial rights;

Third, we'll talk about how a publisher should approach maximizing his share of the growing online book sales market.

In stark contrast to the bullish experience of the retailers, and though many of them have tried, publishers generally have little success to report selling on the Net. Many publishers are first attracted to the Net by the illusion of an expansion of margin. But, of course, few publishers are well set-up for single copy distribution. Those that are know they need an order of a minimum size to be profitable, and that the customer service demands of dealing with the consumer put a real dent in those apparently expanded margins.

From the beginning of cybertime, the threat of publishers by-passing the intermediaries to sell direct to the consumer has annoyed booksellers. It still does. A recent string on an ABA email list had many booksellers complaining long and loud about one big publisher's efforts to recruit direct customers from site visitors who completed a "tell us your thoughts" questionnaire. The most offended bookseller had provided links to the publisher for an author tie-in, and found its own direct customers being recruited away by that publisher.

But the booksellers really shouldn't worry about publishers selling direct online, any more than they should worry if some publisher is silly enough to start a chain of bookstores with just his own books in it. Indeed, that idea has been tried from time to time with about as much success as most publishers have had selling books from their own Web sites.

The bookseller has an authority and presumption of objectivity that the publisher can't ever achieve. The consumer doesn't want to see one publishers' offerings on a subject; he wants to see them all. And he wants recommendations he can trust. The publisher's only two conceivable competitive advantages are price, if he chooses to discount, or certainty of supply, which other publishers will demonstrate can virtually be assured to the bookstores in a cyber-world anyway.

The publisher can employ the competitive advantages only by very unfriendly acts: squeezing the booksellers, including current customers, on margin; or consciously failing to meet what will be perceived as an acceptable standard of service.

We recommend that publishers make their decisions about when and how to sell on the Web guided by two principles:

1. The publisher's first responsibility must be to the customer on its Web site who wants to order a book. The publisher must make it as easy as possible for the customer to order the book.

2. Without compromising that obligation, the publisher should strive to take as FEW direct orders as possible.

Let's consider how pursuing these objectives in tandem would play out. Of course, the publisher would generate orders by its Web activity. That's often the point. So now we get somebody ready to order, what do we do?

We click them to an ordering screen.

A bookseller's ordering screen.

Which bookseller?

That one of the most dynamic operational and sales questions we will face over the next several years. And all publishers will ultimately face it, although most now are simply avoiding it, or calling it an unsolveable question.

The nut of the situation is this. In order to work effectively with booksellers online, you must be able to link with them, in both directions. They need to link to your page so they can use your promotional and informational copy; you need to link to theirs to close the loop on an author promotion at their store. But they won't want to link with you if you are competing for their sales.

Today what a publisher should do is work with all the existing accounts that have reached a threshold of competence with online bookselling. You can strike any number of bargains about how to work with them, but all sorts of opportunities will open up when you start to discuss how you can direct business to your retailers. What I would foresee as a future model is a rotation of referred retailers, the way Web sites now rotate advertisers.

And positioning yourself that way removes the most significant obstacle to using your Web capabilities in synch with the growing body of Web retailers.

Of course, geography will also become a very logical component of the referral decision as the universe grows of retailers who are competent online. And someday soon linking the store inventory databases with your own will permit you to know exactly where each book is in stock.

In the short run, requiring the stores to have a breadth of stock is a reasonable quid pro quo for referring direct business. Which is another reason for publishers to do it.

Any publisher for whom the attractions of a policy which minimizes direct selling isn't sufficiently persuasive should consider the dangers in the alternatives. Any other policy, even one that is ostensibly laissez-faire, letting the consumer buy wherever the consumer wants, will often be perceived by the bookseller as the publisher poaching the bookseller's business. And the cost of that perception could be substantial since the lion's share of online business will always be done by retailers.

It is also important to remember that, Amazon, Bookstacks, and Bookserve and other exclusively cyber-retailers aside, all the publishers' existing bookstore customers are also building their Web presences as well. As we have observed once before in this same forum, retaliation for perceived cybersins may take place in old, established channels.

In the next year, I think we'll see that Amazon, Bookstacks, and Bookserve will have established a baseline minimum for real Internet bookselling competitiveness. Ingram and Baker & Taylor will have systems in place to assure that every store served by them has the capabilities these stores have today. That adds up to being able to satisfy the lion's share of demand from as many as 400,000 titles in a couple of days, virtually the universe (including out-of-print searches) if you'll wait a little longer.

The independent stores that have already worked the Web with a database of their own inventories will continue to offer the edge of even faster provision of titles actually in the store, and will also benefit by promoting the store to the cybercommunity as a destination. The independents will also find themselves at an advantage when they adopt the "we have everything" Amazon model, since resourcing the non-wholesaler titles is facilitated by having direct buying relationships with the publishers, which all these independents do.

This adds up to some severe shocks to the current networks of international distribution of English-language books. Every territory in the world will lose sales to the United States because US cyber-retailers, with Ingram and the rest of the infrastructure backing them up, will offer the widest selection and fastest service on many titles of all the world's online English-language booksellers.

Every Internet bookseller we talked to in both the US and he UK reports that 25-50% of their online sales are coming from abroad. Britain's Internet Bookshop told an executive of Vista that their offshore sales are as much as 80%, which may be an artefact of their former supply model, but is still an attention-grabbing number. Among all the booksellers, we heard stories of sales to practically every country on the globe. And these sales are not just single-copy sales to consumers. Our booksellers report sales of 100 and 200 copies to schools, businesses, and other organizations world-wide. We believe these sales might already be having a noticeable impact in particular places with particular books, but they can't fail to have a widespread impact within the next 12 months.

Indeed, the Internet will accomplish what years of work by the export arms of American companies, changes in open market rules, and the best efforts of wholesalers and export sales companies never did: making US-published books by far the most easily available in the world. With retail prices in the US already often lower than the exchange rate conversion to other English-speaking markets, the cost of buying at or near US retail and paying international shipping will often, perhaps usually, be competitive with current local pricing for English-language books. But availability will be the lure that will put the non-US English-language book market online.

Perhaps the most stunning example of what this can mean to UK-based publishers comes from IBS having adopted the Amazon supply model. IBS told us that they would, indeed, order a book from Ingram for a UK-based customer if the UK publisher were out of stock. IBS themselves pointed out that the Ingram-supplied book would almost always be cheaper. That begs a rather obvious question that I must admit didn't occur to me at the time: would IBS order the book from Ingram for a UK customer simply because it was cheaper?

Of course, their major competitor, Amazon dot com, would. And they do it every day.

While we didn't think to ask IBS whether they would buy around UK publishers simply to save a UK customer some money, we'll bet some UK publisher will.

Of course, this represents a potential bonanza for US-based publishers, even if there may be some disruption to their extended family of subsidiaries and distributors. Book purchasers the world over are going to find the easiest way to buy English-language books is from a Web seller. The Web sellers with the best selection, best prices, and fastest supply will be US-based, so the US editions of all books will gain the greatest benefits as this market expands.

US-based publishers should expect less sales from their foreign subsidiaries. They should also be prepared to accept one world price; a big differential between the US dollar price and various other local markets will be less and less possible to sustain. And, similar to what has occurred with backlist in the superstores, publishers will see the "lift" from these new sales falling unevenly across their list, based on what books have the most offshore demand, particularly among denizens of the Internet.

In time, we might expect UK publishers to start selling direct to Ingram, or for Ingram to find a way to resource UK-based books. After all, what we are experiencing here is what the Internet does to a jerry-rigged, inefficient supply chain. Books that are published in the UK and not available in the US are prime candidates to fall under that heading.

Lest we leave the impression that there is no point to any publisher doing anything to serve this cybermarket, because Ingram is going to "control the market", let me hasten to say that would be a misleading oversimplification. While we believe publishers are foolish to pursue direct sales as a strategy, there is a great deal publishers can do to influence cybersales. And no matter how good Amazon dot com or any other cyber-bookseller gets, the low cost of entry assures that booksellers everywhere will be making cybersales.

Although the virtual bookstore model demonstrates that all you need is a database and access to Ingram's warehouse to be competitive, all the cyberretailers have attempted from the beginning to compete in addition with service and information. The information and service are both driven by meta-information, information about the books themselves. This will become the ultimate differentiation between Web booksellers. The sites offer reviews, their own and from various other sources, rarely linked yet as a field in the database for the books. Before long, you will expect to be able to search for a subject or author, pull up a list, and click to reviews of the books. The sites are also offering alert services: you tell them who your favorite authors are, and they'll tell you when their new books come out.

And this is where the publishers' big opportunity lies to provide what no wholesaler can.

Of course, publishers have had the better part of a century to develop an array of weapons to gain territory with retailers on the ground. We have an arsenal that includes end caps, floor displays, co-op advertising, reading copies to get attention from the sales staff, tricks and techniques to create a successful in-store event.

But none of these guns fire online. There are a whole different set of skills required to service the cyber-retailer, and our successful cyber-retailers ranked publishers "from competent to clueless", mostly the latter.

Indeed, one of the most striking comments that we got when we asked retailers to grade the publishers for their cyber-assistance was, "Macmillan is clearly the best; Prentice-Hall is clearly the worst." What made this comment so arresting, of course, is that both are Simon & Schuster companies.

A great deal of what a bookseller needs from a publisher is practically free for the publisher to provide: gif files of covers, flap copy and catalog copy and press release copy and sample chapters in digital form. Very basic stuff, but our booksellers report almost universally that publishers universally can't do it.

An early sign of a US trade publisher at least exploring these opportunities was apparently occurring while we were doing interviews in September. Wordsworth reported that they had just heard from Random House that text-based information on new releases would be made available. The easy availability of this kind of information will present a significant competitive edge in gaining share of the cyber-market in the short run. Before long, the failure to deliver it will constitute a serious competitive handicap.

Ultimately, the failure to maximize opportunities to sell online will affect sales on the ground as well. Certainly at an account like Wordsworth the 90% of the sales still made on the ground will be affected by the 10% made on the net. All of our conventional bookstores with cyber-presence told us that their shoppers often came in with information they'd pulled off the net, including checking to see if the store had the book they wanted before they came down to buy it. It is easy to imagine a store customer preparing to shop by pulling down some online catalog copy and reviews to take along to the store. It is already happening.

It is getting rapidly past the time when a publishing sales department can afford not to monitor bookselling activity online. Even at the rep level, it is necessary for a publisher's salesperson to see the Web sites for accounts he or she calls on. Or be at a disadvantage versus reps who do. It is the modern equivalent of not being able to see a store's newsletter, or their front window.

The hurdles most publishers need to leap to avoid falling behind require only the development of competency. The additional cost factors, in a world where every desk already has a PC and almost all have a modem, are trivial. Developing competency will require changes in procedure as well as education of personnel. The capability of somebody in the sales promotion department to create a first chapter email promotion can be frustrated if getting the text file from the production department is too difficult or cumbersome.

We'd expect, before long, to see publishing sales departments employ a cyber sales manager, whose principal business will be with existing major accounts. The capabilities that will be developed will end up with broader implications: to publicity, to sales in non-consumer segments, and to rights and author relations. Somehow, each sales department will have to be working with cyber-retailers to gain access to their mailing lists, the feature spots on their Web sites, to coordinate an online chat session with an author.

What the Superstores have proved in the United States is that when it comes to selling books, the biggest selection wins. The first corollary to that principal is that a big selection requires information and knowledge to navigate. Good bookstores and good sales forces have always appreciated the value of knowledgeable sales staff.

We're now in a world where many retailers can have access to a virtually limitless selection of books, and their new virtual sales staff is the publisher's information, combined with the retailer's ability to sort it, combine it with other material, and present it. So competitive publishers will have to be able to give it them.

Fortunately for everybody in this room, Simon Verdon of Open Communications, our Australian cyber-retailer who specializes in computer books, has summarized the requirements very well. You'll recall that he was in business long before Amazon dot com. Here is his assessment of what the "serious publisher" -- those are his words -- should have:

One: an FTP site with bulk data for downloading, grouped by topic area, updated regularly;

Two: data that is normalized to allow importation into other databases with minimal problems;

Three: Annotations that explain the books;

Four: Full title and author information;

Five: US dollar pricing;

Six: ISBNs without dashes;

Seven: a cumulative report of out-of-print ISBNs updated at least monthly;

What Verdon has spelled out really are the basics: the equivalent to a rep calling on a bookstore being told to "have an order form and a pencil when you show up". No publisher is close to fulfilling those basic requirements yet. Given the explosive growth of the Internet market for books, it just could be that the number of publishers who will have mastered these capabilities by 12 months from now will be next year's surprise.

It is very gratifying to be able to conclude these remarks by very briefly urging you to become involved in two initiatives involving our hosts, Vista Computer Services, which very much relate to the growth in online bookselling for which too few of publishers are adequately prepared.

One is the Electronic Book Aisle, an invention of OverDrive Technology in Cleveland, which we mentioned in June when it was a prototype. The Book Aisle permits online searching and previewing and then sells downloads at retail. It already contains the full text file and visual presentation of selected titles from many US publishers. The Book Aisle makes the transaction, moving the file and collecting the money, and remits half the receipts to the publisher. Incremental revenue, no incremental cost of goods.

The best use of the Book Aisle, and, we believe, the most efficient path to full Internet utility for book publishers, is enabled by employing OverDrive's technology in the book-creation process. Vista has joined forces with OverDrive to help publishers install OverDrive technology into their processes. The Book Aisle itself and getting Vista's help to make systematic use of it are two separate propositions that should each be carefully examined by every publisher.

The second initiative involves these annual research papers that Mark Bide so ably designs and supervises. Vista is particularly timely with an examination over the next several months of how networks, including the Web, will affect the publishing value chain. Partly thanks to Vista's imaginative use of the new affiliation with OverDrive, we will be including the entire publishing community in that effort through the Web.