G r a d u a l l y , then SUDDENLY.

For the past five years in these VISTA conferences, starting in June of 1995, we have been saying that the most important single consequence of digital technology would be that book sales and marketing would move to the Internet.

Well, it has. We'll talk some US numbers here, but they'll reflect an equivalent situation in Britain three months from now, if it isn't equivalent already.

True numbers are really hard to get at because a lot of the Internet retailers' business is transacted by publishers through wholesalers, but many publishers trying to figure it out can already see that more than five percent of their business is through Amazon.com alone. That number is rising along with a lot more going to BN.com and the online arms of Borders and other chains, major retailers' and other independent booksellers' efforts to sell books online, and a proliferating network of "affiliate" Web sites attached to the major book distribution capabilities. Amazon and BN are both very actively promoting this proliferation. And the American Booksellers Association promises to start helping more independents go online with their BookSense effort sometime next year.

In fact, several sales directors of major US houses representing a very broad range of books privately estimated their online sales today are already in excess of 10 percent, and rising at 30-to-50 percent a year.

So almost everybody is probably doing at least 7-1/2 percent of their business online at the moment in the trade and professional sectors. That minimum will be 10 percent soon. And it will be a quarter of a lot of companies' business a year or two from now.

The one exception to this trend may be juveniles. As one sales director said, "online activity is still [driven by] 'need to know' or 'got to have'" purchases.

It is past time for every company to consciously formulate a sales strategy to compete in the channel that is already for many books the single most important, and the one which we know is only bound to grow for the foreseeable future.

We have at the same time and in the same way, gradually, then suddenly, reached a similar sort of critical mass of Internet importance on the marketing side. It is safe to say that virtually every major reviewer, journalist, writer, producer, and broadcaster carries on most of his or her correspondence through email. Many, if not most, get the lion's share of their information from the Internet, much of it from brick and mortar brands. In other words, the buzz machinery for every conceivable kind of book is now online.

Indeed, a recurring strategic question every company is going to face is "can we even distinguish between 'sales' and 'marketing' on the Web?". That is a very complicated question, so fraught with political and practical distinctions that it can only be considered company by company and even task by task.

Before we proceed to consider five sales and/or marketing functions which are the building blocks to utilize the online channel and which must form the core of a book publisher's Internet strategy, I want to take time out to consider two fundamentals that should guide our strategic thinking: one a basic truth about business and technology, the other an observation about how books, almost ALL books, sell.

There is an observed phenomenon, previously remarked upon at many VISTA conferences, that describes the economic consequences of new technology or infrastructure adoptions. It is called the "S curve". It plots the "Effort" required to implement a new technology against the "Rewards" for using it and posits that the graph they create is shaped like an S.

That is, when a new technology is introduced, or an infrastructure is being built, there is, at first, very little reward. You may have an automobile, but you can't really use it if you don't have roads and gas stations. You may have a Personal Computer but business can't use it to its full potential until the right programs are written, the people hired are comfortable using the technology, and business processes adjust to incorporate the new capabilities.

Then you get a period, often a very long period, where you get a LOT of reward with very little additional effort and relatively less infrastructure construction. And then, as a technology matures or is completely installed, you flatten out again at a higher level. Increasing the effort or investment won't give you much incremental benefit. Think traffic jams. Or air travel.

All of the strategic suggestions we will consider today consist of infrastructure-building within a context of the even larger infrastructure-building of the capability of the Internet. But by investing in this technology, or, more precisely, by investing in utilizing this technology, before too long a company will have assets that will enable a great multiplication of future efforts.

Now a moment's consideration about how books sell. For these purposes, I am going to accept the possible exclusion of books whose authors start with a massive core audience so large and developed it is hard to grow, or those that get a huge event-driven boost: like an Oprah endorsement in the States or the death of a major public figure. Congratulations to you if you have one or more of these.

But the other 99 percent or more of the books all work this way. They start with a core audience of marketing megaphones and potential readers. For both, the true core group is both very small and very targetable. It is the author's family and professional contacts, those who care passionately about the subject of the book, usually specialist publications and journalists already on the story, particular librarians who passionately collect and catalog the subject this book is about, or corporations that use the technology or the marketing technique or the 401-K plan mentioned in the book.

And then that core group will pass judgement on the book and be effective to varying degrees at providing the impetus that gives the book real market penetration in its core communities and, in the best of circumstances, propels the book beyond the core communities.

Against that background, here's a template for a strategy to start building a sales and marketing effort that will pay off increasingly in the digital age. We're going to look at it as a list of functions heretofore pretty much unknown, but which need to be managed in the new digital environment. Imagine a person in charge of each of these functions immediately if you don't already have one. Each will have its own department pretty soon.

1. Call on the territory: serve your intermediaries.

If your company maintains the illusion which was widespread a year or two ago that you will win on the Internet by selling direct, fattening your margin, and cutting out your intermediaries, I would urge you to reconsider immediately.

No publisher, even the great brands identified with clear product offerings, like Abrams or Scholastic or Dummies in the States, can possibly win the Internet battle for eyeballs in the tidal wave of page views to come. Every company in the world, large and small, will be serving its public and private needs on the Internet, becoming the destination URL for untold entertainment and information quests no publisher could hope to compete with.

Amazon originally, and now BN dot com and ultimately, in some manner or means, all of the full-service Internet booksellers, offer "affiliate" status to interest-specific Web sites that want to offer books to their own viewers. The relationship is achieved with simple linking, and can result cumulatively in lots of sales for Amazon and BN although the individual revenue streams back to the hundreds of thousands, or perhaps already more than a million, affiliated Web sites may individually be rather modest.

It has frequently been observed that infomediaries, Web sites that accrue information on a subject for a specialized audience, will be proliferating rapidly and often include book and information sales as part of their revenue and service model. The affiliate programs give them an easy way to do it.

There is important opportunity here for publishers. Infomediary Web sites need content, which the Web retailers can't really provide. The publishers can. I would not suggest poaching the retailers' affiliates by offering affiliates a better percentage to sell your books and link to you rather than selling everybody's books and linking to a bookstore. But I would suggest that you offer help with content that will get your books get featured and get more of them sold.

If you don't know how to do that yet, figuring it out can be on the starter list for whoever in your organization takes this new territory. You have to cover the existing and potential Web intermediaries the way you have covered the ones that are bricks and mortar: trying to figure out how to help them sell more of your books and participate more fully in your marketing efforts. And you have to include in your measurement of the success of this function how rapidly the number of accounts called on goes up.

2. Sell fragments. You can't just sell books anymore.

Almost all publishers have many books that will have a market in fragments as digital distribution and Print-on-Demand capabilities become more widespread. It may not be in use very much yet, but the infrastructure is in place for a professor, a corporate team leader, or a 5th grade teacher to take a chapter from this and a chapter from that to assemble material to be delivered in paperback books, durable enough for multiple use, at a manageable price. Will your chapter on 1970s diplomacy or benchmarking the opposition or the scene-setting opening from a major new novelist be left out because you don't have a way to make the deal?

If it is hard sometimes to distinguish between sales and marketing, this aspect of the digital revolution necessarily involves the editorial department as well. All publishers are already familiar with this sort of opportunity on behalf of big books, where the author gets an opportunity to write an Op-Ed piece for the New York Times, or even an article for a magazine, that is not a straight serial "lift" from the book but which will certainly help promote it. In the past these have been limited to pub date opportunities for the biggest books.

Not anymore. The drive for fragment identification and creation will come from both push and pull, as the consultants like to say. The push will come as Internet-savvy marketing and sales departments build their relationships with Web sites desperate for content that the publisher's author base and editorial expertise can provide, tied into the books the publisher already has on the list. They will push chunks of the publishers' content to these sites, for promotion and for direct revenue or for both. The pull will come from the teachers and professors and corporate trainers and many others who become aware of their power to assemble customized content, and even have it bound into books. They will become ubiquitous customers for rights and permissions, pulling material from the publishers for uses the publishers could never have imagined or created.

Let me give you one current example that I happen to know about. There is a new Web site called HungryMinds.com, which its founder Stuart Skorman described to me as "the place on the Web for anybody who wants to teach anything or anybody who wants to learn anything." Initially, HungryMinds links to existing distance learning programs at major universities and other existing tutorial situations that can be adapted to the Web.

But in the long run, Stuart does mean anybody, and he will be developing tools to help develop curriculum from sources of knowledge. This kind of site will be an important revenue stream for the creators of books, as well as a way to sell the books themselves. Some publishers will learn how to use this better on behalf of their authors than others. So will some agents.

All of this suggests that ultimately there is a whole new business that most publishers will find themselves getting into, promoting the sale of customized content, which today is the domain of a very small number of special sales departments scattered throughout the industry. You'll get to the new business in due course, but you can start with somebody taking responsibility for generating content revenues from the Internet, including going back to your authors when necessary to provide this new market what it really needs.

3. Diversify the offering of enriched metadata.

Basic metadata is what we used to call bibliographic information: title, author, price, trim size, and so forth. Enriched metadata is a deeper collection of material on the book: a GIF file of the cover, reviews, descriptive copy.

We have observed the value of book fragments to Web sites hungry for content. The first set of fragments a publisher has to offer Web sites is the voluminous information about its books that all publishers create and can assemble. Publishers routinely have proposal copy, catalog copy, flap copy, title-information-sheet copy, pitch letter copy, and copy prepared for sundry special promotions and purposes, although they don't -- yet -- routinely have easy access to it.

In addition, publishers could likely get the right to circulate reviews, or large sections of reviews. Permission to use review material this way could potentially become a condition for distribution of a printed book, which we'll talk more about later.

Our metadata techniques, as an industry, are now in such a primitive state that each Web re-seller is creating its own gif files to display book covers, an exercise that should surely be managed by the publishers who want the best images of their books shown to their potential customers.

We don't really know yet what the most useful collection of enriched metadata will be. So in addition to offering what we have as a by-product of our existing marketing processes, we would be wise to respond to Web site demand for a while, finding out what works. The relative value of the information about the books, the information within the books, and additional information that can be provided under the authority of the books, will vary in this new environment.

The person in your organization who takes this responsibility is most concerned with distributing information so that it will sell more books. This function connects very closely to the attempt to generate revenue from Web sites, and will occasionally conflict, as happens right now with publicity and sub rights.

4. Re-engineer your bound galley and review copy process. Now.

Microsoft's new ClearType initiative and the new print-on-demand capabilities create an opportunity for publishers to completely re-engineer their galley and review copy distribution processes to cut waste, save money, offer better service to the book review community, and, as a result, achieve a higher level of awareness and review attention for their books.

The time-honored process publishers have followed to gain review attention for their books is first distribute "bound galleys", makeshift softcover books created from an early and often uncorrected iteration of the typescript. The publisher orders a specific number of the bound galleys in advance, which is always more or fewer than the ideal number turns out to be. This either means that expensive bound galleys are thrown out, or that some opportunities to get an early reading are passed up.

After the books are delivered, the house once again delivers samples, this time of the finished books. Here the greatest expense is often the shipping; the incremental cost of a book as part of a run is seldom more than 10 percent of the retail price.

Certain components of a book's marketing effort, most notably magazines entertaining serial rights and reviewers within the trade who serve booksellers and librarians, require a reviewable copy of a book long enough before publication date that the book wouldn't be printed yet. But technology is offering the opportunity to eliminate the extremely costly bound galley.

Instead, publishers can offer a digital, e-book- or PC-readable copy of their book, with no printing or delivery cost in some cases, complemented with copies created through a POD system where that is required. It would even save publishers some money if they had to give a RocketBook or SoftBook e-book reader as a free gift to regular reviewer-recipients of their bound galleys.

This is not a simple suggestion to implement, and it might even be a few months ahead of its time. Certainly, some reviewers would adjust to this change more readily than others, which is why the digital file delivery needs to be tied to a POD solution, so printed and bound copies continue to be available alongside the digital distribution for quite some time (in what is actually a precursor to what will happen in the consumer market over the next ten or twenty years.)

Bound galley and review copy distribution is a significant per-title marketing expense that, because it drives up the investment that must be recovered on every title, drives out the midlist. In other contexts, we have discussed the challenge facing all publishers to figure out how to publish smaller-potential titles profitably. Using new technology to eliminate most of the cost of getting basic review attention could certainly be helpful in that regard. And somebody in charge of looking at these functions against the capabilities becoming available is likely to earn back their salary very quickly.

5. Manage your cyber LISTS!

The more direct-marketing-oriented you were before the Internet revolution began a few years ago, the more aware you would inevitably be of its power as a direct response medium that requires no printing cost and no mailing cost. But this direct marketing power of the Internet does not exist purely at the level of the ultimate consumer; it is also an essential tool to manage all the potential intermediaries and marketing megaphones that can help a publisher sell books.

It will pay off in the future to capture and categorize every email address and every Web site you encounter during your book marketing. The email addresses can, over time, give you the tool to announce new publications very cheaply to a core audience, perhaps to capture direct business but more likely to drive customers to your intermediaries and get the buzz going for a book.

Capturing and categorizing the Web site information from your current activities can be even more important. The most recent figure I've heard is that there are eight hundred MILLION sites on the Web. We are all aware that the biggest stumbling block to efficient use of the Web is the difficulty of targeting a search. Too many subjects one would search under yield many thousands of sites to examine and, of course, most of them don't really apply to the matter at hand.

But publishers who tend to work the same areas over and over again can work their way through those lists, weeding out the sites that are irrelevant or amateurish or specialized in some esoteric way that makes them useless. Cyberlist management that links the results of work done for many individual books will compound the benefits of earlier efforts.

Cyberlist management will require some elements familiar to old-style mailing list management. Lists will be acquired and have to be integrated, so there will be a need for merging and purging. Old sites will become defunct and old email recipients will die or change addresses, so list "cleaning" will be a constant activity. In time, lists that have proven effective may yield some commercial value. Controlling list "rental" so that lists aren't actually stolen will require dummy names that enable detection of the list's use.

If you don't have somebody managing your cyberlists, you won't even know what opportunities you are passing up. But they will be substantial.

So we have five brand new functions, none of which are identified as functions in any sales department I know of, that will rapidly become essential to every publisher competing in the digital world. Calling on the digital territory, placing content fragments, making enriched metadata readily available to Internet booksellers, delivering review copies through digital techniques, and managing cyberlists are all S-curve activities that will require a low-rewards investment in infrastructure creation before they pay off.

But they will pay off. Gradually, and then suddenly. The S-curve is one curve no publisher with a 10-year plan for survival can afford to fall behind.