Booking the future

Subjects:
About the author
Ransom Stephens is a technologist, physicist and writer based in Northern California. He is the author of "The God Patent"

Is the book dead? Can the Six Sisters of publishing rescue books? Will publishers find a new profit model? Can bookstores survive the internet? Can writers make a living? What about e-books? Is Kindle the beginning and end of the revolution? Will Google Books be literature's savior or executioner? Where does Scribd.com fit in?

 

Ransom Stephens is a San Francisco-based technologist and author. His novel, "The God Patent", is published on Scribd.
openDemocracy's Media and the Net coverage has chronicled the politics of the web's transformation of media since 2001.

Though the role of publishing has not changed - connect readers to writers - the revolution will not be led by an established publisher. To date, no established player has prospered through, much less led, the transition to the digitally-based economy. What's left of the recording industry is still pursuing the fascinating how-to-best-prosecute-our-customers business model. No one was better positioned to profit from the web-based economy than Sears, with its legendary catalog, but Amazon all but killed it. Even IBM barely survived the computer revolution.

For some reason, even when entrenched companies can see the iceberg they can't turn the ship. In 2000, at the height of the "Napster Crisis," Time-Warner/AOL's CEO, Richard Parsons said, "It's an assault on everything that constitutes cultural expression of our society... And the corporations won't be the only ones hurt. Artists will have no incentive to create. Worst-case scenario: the country will end up in a sort of Cultural Dark Age."

Have YouTube, Facebook, iTunes, Blogspot, et al reduced cultural expression? Here's a better example. In 1977, Ken Olson, President of Digital Equipment Corporation (DEC) which, at the time, built the best computing hardware, said, "There is no need for any individual to have a computer in their home." Time-Warner/AOL, Sears and IBM survived, but are swimming in the wake of Dell, Google, Amazon, etc.

Three mistakes will plague the six huge publishing conglomerates, a.k.a., the Six Sisters: first, their blockbuster profit model is unsustainable; second, they're not capable of marketing those titles that are in the market segment with the greatest profit potential; and, third, they've stopped nurturing the majority of their talent with the editorial and promotional nutrition necessary for them to blossom into bestselling authors, the so-called mid-list authors whose early efforts showed enough promise to be published, but didn't return a profit. It looks a lot like what sickened Time-Warner/AOL, Sears and IBM and killed DEC.

Publishers' role as the gatekeepers of quality has always been dubious. Do book buyers have brand loyalty? Do you check the publisher before buying a book? Once we jump the low hurdle of spelling, grammar and minimal storytelling skill, literary merit is nearly as subjective as your favorite color. In a world where musicians can sell their best songs on iTunes, the only thing maintaining publishing's quality-control role is the carefully manicured perception that self-publishing is anathema to aspiring professional authors. Publishing, through its marketing plans and budgets, today effectively controls who sees what book. But the grip of the industry's role of gatekeeper is about to go.

The publishing company that turns the corner, leaving the Six Sisters in the dust, will leave quality control to authors - even grammar and spelling.

The obvious candidates include Yahoo and Amazon, but I think they are already too big and stodgy to make the move; Google has everything necessary on its place, but might be too fragmented to make the move; the big self-publishing companies Lulu and iUniverse are well positioned but might be too burdened by the "vanity press" label to emerge. Right now, I think the smart money is on Scribd.com. Anyway, for the sake of argument, let's call the emerging company NetBoox.

The biggest change in the terrain is that NetBoox will be able to "stock" every title not just online, but in every bookstore. Currently, the biggest bookstores hold well under 100,000 titles. The basis for their profit-model is the assumption that the fraction of stocked titles covers 80% of demand. In other words, they've been operating under the assumption that 20% of available titles accounts for 80% of realizable (as opposed to realized) profits. The 80/20 rule embodied: 80% of the results come from 20% of the work.  

It turns out that the 80/20 rule is wrong. It's more of a 40/20 game. This is the lesson of long tail economics.

Long tail economics is straightforward: First, assemble every title, not just those in or out of print, but those that could be in print, every possible title; second, organize the titles in decreasing order of demand; and, third, plot the number of copies that could be sold if everyone was aware of them, i.e., the potential demand. The result is the graph shown in Figure 1. Gimme bestsellers, like Patterson and King titles, are on the far left, mid-list titles in the center-left and low-volume sellers, niche books, on the right.

Figure 1: Publishing's long tail. Demand for titles organized from highest to lowest. Bestsellers are on the far left, like King or Patterson titles, mid-list titles in the center-left, and niche titles, like my grandpa's unpublished memoir on the right.

To figure out how NetBoox will operate, consider the extremes. The Six Sisters concentrate their marketing prowess on squeezing as much money as possible from bestsellers, the darkened 20% of titles/40% of demand on the left - the so-called blockbuster strategy.

Stephen King can write a book, punt it over to Lulu, put a note on his web page and it will sell. Bookstores will order it, people will buy it in both electronic and bound form, and Stephen King will earn about twice the royalties he would from a conventional publisher. The loss of their golden-egg laying geese is the first huge profit-problem established publishers will face.

What about the cover art, distribution, and the promotional prowess required to propel a title onto the bestseller list? Established bestselling authors don't need a publisher to be promoted they need publicists; the retail e-industry has solved the distribution problem; and cover art and bindery will improve the instant Stephen King tosses his text to iUniverse or Kinkos.

Now consider the extreme right, the long tail of the demand distribution. If my grandpa wrote a bad memoir about the day we first rode horses together, the total market might be 100 copies. Or maybe three: mom, me and, well, surely someone else would like a copy. What crazy publisher would carry Papa's memoir? NetBoox would carry that title and make a cozy profit on those 3-100 copies.

Papa writes his bad memoir, uploads it to NetBoox and clicks the usual "I agree" button next to the usual infinite scroll of unintelligible text. Had he read the legalese, he'd have seen that he will only collect royalties if the book sells more than 1000 copies. Or maybe 10,000 or 100,000. It doesn't matter because Papa doesn't care; he just wants his book out there.

Mom and me and that other guy order it and, as we click, the text is either automatically submitted to a queue for printing, binding, and shipping, that is, it is Published on Demand (PoD), or we download the appropriate format for our preferred electronic book-readers. NetBoox' only overhead is storage of the text file. A novel takes less than a megabyte; storage of 100 books costs less than two cents per year. While the unit profit is small, it's still thousands of times the overhead. In other words, if NetBoox holds the rights to half the titles that the six sisters would never consider carrying, half the unshaded long tail in Figure 1, they'll make twice the money that the six sisters make from their caches of high-overhead bestsellers - bestsellers who don't need them and are going to leave.

After NetBoox acquires rights to all the manuscripts that established publishers have rejected - the slush of the slush, the unwashed huddled manuscript masses - they need to find people who will pay for them, and they need to do so at very little cost. Based on this problem, Harvard economist Anita Elberse argues that the blockbuster strategy will dominate long tail economics because, she says, "It is extremely difficult to forecast the demand for a new title." But she neglects the massive potential of sophisticated targeted marketing techniques that are only now starting to emerge.

With every click of your mouse, NetBoox' model of your preferences becomes more accurate. Composing a preference algorithm isn't rocket science, but rocket scientists are composing them at places like AudienceScience.com. By preference model, I don't mean a trivial guessing algorithm like Amazon's "people who purchased A also liked B." Netflix has a better example, they build a model of your preferences every time you rate a movie and the recommendations are pretty accurate. In fact, if you can come up with a better algorithm, Netflix will pay you a million bucks for it (Netflixprize.com).

Preference models work by building confidence levels from large samples of preference data. Along with your book choices, every click of your mouse can be correlated to your preferences. The complexity runs from "people who liked title A also liked title B" to "people who bought tickets to this event, liked that title" to "people who read this blog, drink that beer, wear those clothes, listen to that music, and ... liked title C." Correlations are everywhere.

Remember the third guy, besides mom and me, who would buy Papa's bad memoir? This is how it works: NetBoox has a rack of servers plodding along, calculating the confidence level for the potential of every person to buy every title. This guy surfs the web and, at some point, NetBoox' algorithm rings up a 75% confidence level that he would buy Papa's book. With his next click, he'll see an ad for Papa's bad memoir. It might be the only ad ever placed for this book. The guy likes the book, buys it in either electronic or printed form and NetBoox collects the dough.

The same promotion model applies for every NetBoox title, but, unlike hobbyist writers in the long tail, writers of mid-list titles are willing to do whatever it takes to become profitable authors. Mid-listers receive tiny advances from established publishers and must self-finance their book promotion. Once the bestsellers start evacuating the inventory of established publishers, the vanity stigma of self-publishing will evaporate and the mid-listers will flock to NetBoox. NetBoox won't pay an advance, but will grant royalties that surpass those of the Six Sisters by the smallest margin necessary to maintain the exodus.

This will be the Sisters' moment of reckoning. They will invest more heavily in the literary-lottery, choosing a few titles that their instincts tell them have bestseller potential and reduce investment in the development of new authors. With centuries of publishing experience, the Sisters will hit on enough bestsellers to survive in the same way that Time/AOL, IBM, and Sears did before them. But with decreasing numbers of titles and formidable NetBoox competition, they will continue to whither until they are either acquired by NetBoox, find themselves in the DEC-memorial hospice, or... wakeup and rediscover confidence in the skills they've developed over centuries, skills so ingrained in their corporate culture that they might as well be instincts.

NetBoox needs data for preference models to predict a given customer's desire. Since Google has the greatest cache of necessary data, it must be considered a formidable contender. If Google Books musters the focus, it could dispel companies like Scribd and Amazon in a few months.


On the other hand, the established publishing industry has an instinct for identifying salable authors. To survive, they must fully capitalize on that instinct as they adapt to the terrain as it shifts under NetBoox weight. Rather than cast aside authors whose debut work faded into the disappointing mid-list part of the demand curve, the surviving Sisters will remember why they published that debut in the first place, sign the author to multiple book deals with reasonable promotion budgets, performance bonuses, and accelerating advances. The model is like free agency in professional sports. You try to hang onto the big stars and pile up wins, but are aware that it is the ascending special teamer, the guy tearing up the minor leagues, the sixth man on the bench who guarantees your future. Your only hope to build a dynasty is to sign the stars to multiple book contracts before they know they're stars.

In this way the Sisters can hang on to the middle of the distribution as NetBoox redefines the market.

And redefine the market they will:
NetBoox will provide a free e-bookreader in exchange for a subscription "commitment" - the same sort of long-term-with-penalty commitment that gets you a free cell phone. Your subscription level will allow so many e-book downloads per month and/or so many printed and bound titles per month with pricing structures based on a licensing model. If you buy the hardback, you'll get the audio version and the e-book for free. If the book has been made into a movie, you might get that too. Plus, short stories, novellas, even individual chapters will be profitable - you'll be able to assemble your own anthologies.

While lack of copyright protection has been the bane of the recording industry and, if they weren't so good at pricing DVDs, would be more than a threat to the movie industry, it's not a threat to books. There's a big difference between literature and music or movies; as bibliophiles, literature-lovers, literati-snobs, we are loathe to admit it: No one steals literature. Authors and publishers want their titles in libraries, fercrissakes. Even when blockbusters are downloaded to pdf files, there's no evidence that their proliferation decreases sales; on the contrary, giving free books seems to generate buzz and increase profits. Except for the case of textbooks - but that's a different article (the answer: textbooks in printed form will truly, conclusively die).

E-books will have an effect, but won't be disruptive.

The mass-market, pulp paperback will probably die. Its value as a medium for storytelling is easily replaced by text on a bookreader and its value as a printed record, gift, collectible and decoration is dwarfed by cloth-bound hardbacks, plus, it's not searchable, the font size is fixed, and there's no backlighting. The pulp paperback will only hang on as long as people are uncomfortable taking their bookreaders to the beach or bathtub. On the other hand, as the simplest PoD output, the trade paperback will survive.

Hardbacks are a different story. They will survive for a long time. However, NetBoox isn't going to release many first editions on hardback, it's too expensive. NetBoox will not do a first release in the most costly form without verified market demand. Instead, titles will have to demonstrate wings in electronic form before being released to bookstores in printed and bound form. But don't worry, since NetBoox' processes will be so much faster than publishing as we know it, the casual reader won't notice the difference. Bookstores will notice because they'll have far fewer runway-bound hardbacks to return.

Perhaps the greatest failure of established publishing is time-of-delivery. It takes the Six Sisters 18 to 24 months from the time a title is acquired until it's launched - total disregard for the market window that renders nearly every title a "period piece." How could Julia Angwin's Stealing MySpace be issued over a year after FaceBook rendered MySpace obsolete? NetBoox will have titles out in days. Following the high tech licensing model, the beta release e-book will be followed by a tidier, better edited, v1.0 release - free or as an upgrade to anyone who paid for the beta version. Those books that can fly based on e-book and PoD sales will be released in hardback, online and at your local bookstore.


The Kindle, Sony Bookreader, iBooks on your iPhone - are prototypes for what the bookreader will look like. We can trust engineers and market evolution to produce a bookreader with incredible battery life, awesome graphics, a killer GUI and so forth. The bookreader will have i/o ports, so you can listen to the audio version or plug into your car stereo, and integrated visual-audio bookmarks so that you can listen in the car and read in bed without losing your place. Only the beta version audiobook will use clunky text-to-audio robotics. Stories should be told by authors or actors, please.

Bookstores will survive but must evolve. In fact, NetBoox will heal the epidemic of bookstore failures. People love bookstores and your presence in them is an advantage to everyone in the industry except Amazon. NetBoox will want you in the store because every time you go, you'll buy more of their products than you would online alone. To get people into bookstores, NetBoox will offer free shipping to the bookstore of your choice; if you don't want to go to the bookstore, you'll have to pay for shipping. Of course e-books don't need to be shipped, but bookstores will carry them too and readers will benefit from the usual "what's good?" banter with bookstore employees and other customers.

Bookstores will benefit by not having to stock books that don't fly - though NetBoox is not going to abide the age-old consignment model of the Six Sisters. When a bookstore stocks a bound copy, it's theirs for keeps, no more free returns. But NetBoox will nurture bookstores in other ways. With NetBoox' database of bookstore sales, they can derive preference models for individual stores and target-market at the bookstore scale. For the first time, bookstores will have sales predictions of known uncertainty. Known uncertainty allows accurate calculations of risk/reward probabilities. Right now, as in all old-school marketing, mathematical models are nothing but stodgy accounting estimates. No high tech company, and certainly not NetBoox, would permit such mathematical inelegance in the house. Bayesian statistics, Fisher discriminants, Hadamard matrices provided by NetBoox to the indie bookstore on the corner. Nice.

From the vision provided by Powell's, Book Passage, Books Inc, and the other great independents, we can see that bookstores will look a tiny bit less like libraries and more like cafes and speaking venues centered around author events, book clubs, and storytelling, not to mention cappuccino, Earl Grey, good beer, and, I pray, fine scotch. This is convergence we can live with.

Bookstores will also look a little bit like Kinkos. If it's not already in stock, why should you have to wait more than fifteen minutes for delivery of your book? What are we, cavemen? A PoD printer/binder is desktop equipment - the Espresso Book Machine, so-called "ATM for books." This is how NetBoox will allow every bookstore to "stock" every title.

Our concern should not be for the future of books. The timeline below demonstrates that, in the evolution of storytelling, the introduction of a disruptive technology rarely kills previously established technologies: written stories didn't kill spoken stories, though the book did kill the scroll, but movies didn't kill books, TV didn't kill movies, video games didn't kill TV, and virtual reality won't kill video games. I didn't include the electronic bookreader in the timeline because I don't think it's any more disruptive than the pulp paperback was. It's just another way to read books.

Figure 2: Storytelling technology timeline.

Our concern should be with NetBoox's potentially monopolistic stranglehold on content. In a world where one operating system company assumes inordinate market share, as does one online retailer (Amazon) and one online flea-market (E-Bay), what we need to worry about is the largesse of the provider. When you click the "I agree" button to get your free e-bookreader, you will grant NetBoox permission to acquire whatever data from your computer that they desire to build a model of your preferences. Preference models are built from all the data, not just your choice of books. They're powerful predictors of our choices and, if they can tell what books you'll buy with known uncertainty, they can tell who you'll vote for, too.