This is an archive page !!!

Slaves of a New Machine:
Exploring the For-Free/For-Pay Conundrum

by Laura Fillmore

Part 1

Presented at the Fifth Conference on Organizational Computing, Coordination, and Collaboration: Making Money on the Internet
Austin, Texas
May 10, 1994

Copyright © 1994 by Laura Fillmore; written permission required to reprint.

I very much appreciate being here today and thank Andrew Whinston and the people at the IC2 Institute, RGK, Texas Internet Consulting, and the University of Texas for inviting me to speak at this conference, "Making Money on the Internet." I've picked up many good tips, which I hope to apply at the Online BookStore (OBS) soon, because at the Online BookStore I am *not* making money by publishing on the Internet. And I don't know anyone who is. Since 1992 our company, the Online BookStore, has been involved in Internet publishing, and we have found it an exercise riddled with paradox and the unexpected, frequent bouts of optimism and what-if idea sessions; a conundrum whose parameters keep shifting.

We have enjoyed commercial success from publishing *about,* talking and meeting and consulting *about* publishing on the Internet, but are we making money from the real live act of online publishing for a price? Not yet. As far as I can tell, publishing intelligently in the distributed Internet environment still *costs* money. I don't think I am alone in this realization. I am coming to suspect that there may be no such thing as publishing on the Internet or, rather, publishing according to the definition of publishing as we know it from our familiar paper reference points. This may not be the news you came to hear, but I am happy to share what insights I have with you.

The Online BookStore started out under the umbrella of Editorial Inc., a profitable publishing services business which I started in 1982, and which produced hundreds of books for publishers. In 1991, we had 19 employees, three offices, three shifts around the clock, and used desktop computers to produce such titles as _The Sports Illustrated Almanac_ for Time/Warner, Andy Warhol's biography for Bantam Books, and _Doing Business in Kuwait_ for Ernst and Young. We were part of a services industry trafficking in paper. Using computers, we were able to integrate the various publishing disciplines under one roof, calling on hundreds of freelancers to supplement our in-house working staff and to become a publisher's publisher of sorts, a virtual corporation.

We caught the desktop publishing wave when it was just a swell out at sea in the mid-eighties and, in the resulting shift in the typesetting business from large centralized composition companies to distributed PC-based typesetting platforms, we rode the wave and produced books using page composition systems such as PageMaker, Ventura Publisher, Scribe, TROFF, TEX, and Polaris PrintMerge. Polaris PrintMerge was my favorite, not only because it was the first and crudest PC-based typesetting platform, but because it introduced me to the notion of electronic slavery, our topic of the day. Turn the clock back ten years, when my company was five people strong, and I was salvaging a typesetting job someone had abandoned in frustration before the machine; it was a manual for hospital custodians, arranged in three columns, detailing how to keep a hospital clean and sanitary. The janitors, our future readers, were supposed to start on the upper lefthand column of each page with the instructions to don their uniforms, and then by the bottom right hand entry, they had to "clock out wearing uniform." Every page had the same layout for a different duty--mopping the floors, emptying the trash--all items in all columns had to align three across. Before the days of WYSIWYG, assessing one's success as a typesetter meant printing out again and again, at about five minutes per page. The deadline loomed. I would type and wait, type and wait, a period here, a comma there, a drone before the keyboard, caught up in the electromechanical semi-idiot production cycle. When I finally clocked out, called the Fed Ex man for the finished package, I vowed never again to wear the uniform of typesetter. I would hire typesetters.

This was my first personal experience with electronic servitude in the publishing context, though I didn't realize it at the time: I saw it as a business opportunity instead, which from a commercial standpoint it certainly was. I learned that users of Polaris PrintMerge, no matter how smart, would become the victims of badly designed software, would turn into drones, because that was their inevitable function vis-a-vis their task and the tools afforded them for completing the task; the humans workers functioned as the erring component, the wetware, charged with coaxing a right-or-wrong result out of a desktop computer.

This was the much-touted cutting edge, offering profit without honor and the opportunity to hire others to work Polaris PrintMerge till a better program came along--which happened startlingly fast. I hired others to stay up all night staring into screens, printing out, cursing the widows and orphans, and printing out again. This first generation of servitude involved securing output from single, unconnected machines, getting desktop computers to emulate the work of the large dinosaur machines lumbering reluctantly off into typesetting antiquity; we were selling output from these PCs, trying to recreate type of such quality that it did not appear to be what it was--computer byproduct. So, busily formatting the output of computers gave us a way to use and begin to understand the machines but, of course, in hindsight, how could we have been anything else but servants to the machines? We did not apprehend the utility of our machines; using computers to wrestle with the static and formatted output of single machines was an error few perceived and many committed, are still committing. We do what we know and, with the gift of hindsight, can see that this act of manufacturing type on paper, produced by distributed computers sited in decentralized locations, constituted an intermediate step from static to kinetic publishing.

As so often happens with computers, the cutting edge of desktop publishing became a swamp and then a backwater. It took about five years for publishers to discover that there are cheaper ways to typeset than paying New England style wages--particularly when they can demand from desktop publishers something they never got or even knew they wanted from the large type houses--ownership of the typesetting files and the macros that created them. Savvy publishers took advantage of the distributed environment coupled with the disconnectedness of it all to put the servitude of function together with the servitude of finance. People type fast when they are hungry, and a large part of the typesetting business, the keyboarding at any rate, after the advent of desktop publishing, migrated to romantic "offshore" locations--doublekeying in Taiwan or Jamaica or the Philippines proves more accurate and far less costly than paying a Massachusetts resident a living wage. So, while beginning to explore new avenues of employment in the areas of CD's and hypertext publishing, we continued to compete in the typesetting field, and got good prices for a while in Utah and the Southwest. We even hired freelancers third-hand in Singapore and Haiti. The publisher hired me; I hired someone stateside to hire someone in-country to hire the keyboarder and, still, the publisher ended up paying maybe half what the job would have cost him at $15 per hour. Our topic today is slavery.

In 1895, before Polaris PrintMerge was ever invented, Oscar Wilde wrote in _The Soul of Man Under Socialism,_ that "Unless there are slaves to do the ugly, horrible, uninteresting work, culture and contemplation become almost impossible. Human slavery is wrong, insecure, and demoralizing. On mechanical slavery, on the slavery of the machine, the future of the world depends." Somehow, in the arena of desktop publishing typesetting, human servitude to the exacting output demands of the machine is more the norm. The focus is output, and people become blind inputters of accurate information in the same way that minimum wage workers in supermarkets do when running barcodes over optical scanners.

I'm painting a grim picture, but the tide is about to turn. In the late '80s, we began to produce and typeset books about computer networking, books focusing on the structure and function of globally interlinked computers, which seemed to elicit life from people when used as a communications medium, rather than demeaning them when used as an output medium. One such title, probably the first book on computer networks worldwide, was John Quarterman's book, _The Matrix,_published by Digital Press which concerns computer networks and conferencing systems worldwide. In 1989, when we were working on the production of this book, the author introduced me to the then alien concept of electronic mail. My assistant would pick up mail from my lone correspondent, the author, print it out, put it in my in box, and I would handwrite responses which she would input and send back in due time. It sounds quaint, but it seemed to make sense to me at the time--in the same way computerized typesetting on distributed though unconnected PCs made sense. We do what we know.

Some of the messages he would send had nothing to do with the text of the book itself, however; messages posted to mailing lists from students in Tianamen Square during the uprising that spring, messages from Alaskans offering first-person accounts of what the oil company wasn't telling us about the Valdez disaster. Fresh and unmediated communication about things that mattered from far corners of the world--news just hours old, unsanitized by the media. Here was information, digitally recorded voices, coming out of the new machine, which itself is a vast collection of interconnected machines being used as conduits for human thought. Where the Haitian freelance typist was hidden and voiceless behind four middlemen and had no hope of a phone, no less an Internet connection, the students in Beijing and citizens of Alaska could talk electronically, and there were millions around the world who could and did listen immediately, electronically, and no one stopped them. The Internet is an open network, distributed, not contained, not owned by anyone.

I don't know if any Chinese students or Alaskan citizens profited in a commercial sense from their posting or "publishing" on the Internet--for, after all, what is publishing but writing for public consumption, regardless of the means of distribution or, in the case of Internet publishing, access--but they profited in other perhaps more valuable ways by making their voices heard as witnesses to events of their time. Clearly, in this case, the new and networked machine did not function simply as an output facilitator, a means of replication for familiar words on a paper page. It functioned as a kind of worldwide broadcasting medium.

Call it epiphany thanks to insight from the above incident, or call it simply economic necessity, our business shifted in the direction of electronic publishing, and away from paper-based publishing. Another shift in the tide. The first major step in the new direction, which involved our creating work rather than producing it for publishers, was _The Internet Companion: A Beginner's Guide to Global Networking_ by Tracy LaQuey, which was the first popular trade book about the Internet back in 1992. This book, produced with lightning rapidity and penned by a very gifted and knowledgeable author, seemed to grow beyond itself even before it was born and soon became a bestseller. At a time when there was precious little current copyright information on the Internet, and Acceptable Use Policies stood in the way of for-profit publishing on the Net, we couldn't just put a book up there with a pricetag on it. It was a brave act for Tracy LaQuey to take the innovative leap, to take the words we both wanted people to pay for in the bookstores and give them away, in ASCII, on the Internet. That was the beginning of the Online BookStore in 1992. Many thousands came and grabbed those files; many wrote in asking for more. None of the users paid a dime.

However, a conundrum is a paradox of sorts, and counterintuitive as it may seem, giving the ASCII files away by anonymous FTP spurred the print sales of the book. Who wants to read hundreds of pages in ASCII, anyway? Even our publisher was supportive of our effort and happy with the resulting sales figures. Addison-Wesley is not alone. Prentice Hall publishes Brendan Kehoe's _Zen and the Art of the Internet,_ which is available for free on the Net. His book continues to sell very well. The same applies for MIT Press's publication, _The Hacker's Dictionary,_ which is available for free on the net and sells briskly in paper as well. This leads to conundrum number one: that giving something valuable away for free can make money. It points to a richness not found in the tangible world quite so readily: the more I give to you the more I have. Some call this a new kind of marketing, and this was a pleasing lesson to learn. But was this experience really online publishing, or was it the success of an early hybrid of online/paper publication?

The popularity of the online _Internet Companion_ ASCII files drew my attention further away from paper, and I was seduced by the prospect of the then 10 million people on the Internet--10 million literate people with disposable incomes--attached to the Net. Why not acquire lots of Internet rights to lots of books and put them online at the Online BookStore? Surely some percentage of those people would buy files of a popular author's books for a reasonable price. So to test the concept that people would pay for books online, we approached one of the bestselling authors on the planet, Stephen King, and acquired first serial rights to a story from his new book, _Nightmares and Dreamscapes._ The numbers were enticing: if only one percent of the ten million people paid $5 for Stephen King's story, available only at the OBS and only on the Internet, then that's half a million dollars!

We tried to make it as widely appealing and usable as possible: we formatted it as a Voyager Expanded Book, in plain ASCII for those with only email access, in Adobe Acrobat, in HTML format for Mosaic aficionados, acquired the German rights, did a dual language edition, and released in time for the 1993 Frankfurt Book Fair, the largest book fair in the world. The result: they all came, the radio, the TV, the print media, creating lots of smoke and a nice firm footprint in the sand of Internet history, but sales? The half a million dollars in per-copy sales? All the companies who participated in bringing this story into its Internet incarnation--the Internet Company, Texas Internet Consulting, Viking Penguin, Hodder Stoughton, EUnet Germany, Hoffman und Campe, Aldea Communications, Bunyip, and the Online BookStore--we didn't pull in enough in per-copy sales to pay the phone bills for setting up the deal. A vast amount of smoke, a tremendous marketing boost for the printed book again, lots of noise--and by extension, lots of profit for the publisher and for the author--but handfuls of per-copy sales. The per-copy sales model for a contained publication, a publication which is complete in and of itself and is not linked to anything else of significance on the distributed network, does not seem to work. The OBS is not the only online publishing site which has shown these results.

However, where per-copy faltered, site licensing proved a far better option, which resulted in some commercial satisfaction on all sides. We have sold site licenses to networks and organizations with good results. Site licensing offers exclusivity to the organizations and networks which optioned the work, while offering the author the reassurance of a having defined set of users, and a certain hedge of protection against rampant copying and posting for a profit of his work. One key element in site licensing seems to be timeliness; one publishes first online, before the information or ideas grow old and gather moss. Perhaps this site license model proves more lucrative than the per-copy sales model because it enables the licensor to give the information away for free (after paying for it), while achieving a defined benefit, a market advantage over its competitors, by giving away scarce information on an exclusive or semi-exclusive basis.

Copyright © 1994 by Laura Fillmore; written permission required to reprint.

OBS White Papers Part 2