The Ecology of E-Book Publishing

August 2000

You are listening to a session recording live at Digital Rights Management and Digital Distribution for Publishing held August 15th and 16th, 2000, at the Hotel Niko in San Francisco, California. This conference is presented by the International Quality and Productivity Center. Now to our session....

[The following text is edited somewhat for continuity and clarity.]

Welcome back from lunch. We're about to start the second half of the first day's agenda. This is really a treat to be able to introduce the next speaker. Tim O'Reilly is founder and president of O'Reilly & Associates, thought by many to be the best computer book publisher in the world. He's been a pioneer in the popularization of the Internet. The Global Network Navigator, which was sold to America Online in September 1995, was the first web portal and the first true commercial site on the World Wide Web. Tim has been an activist for Internet standards and for open source software. He, himself, has written numerous books on computer topics. He has conceived an award-winning series of travel books published by an O'Reilly affiliate, Traveler's Tales. He served on the Board of Trustees for both the Internet Society and Electronic Frontier Foundations. And in addition to that, Tim graduated from Harvard with a cum laude in classics. His thesis explored the tension between mysticism and logic in Plato's Dialogues, and I think that might fit what's going on right now. With the talk about publishing and digital rights, some mysticism and logic needs to apply. So, here, not talking about mysticism and logic, but the ecology, okay, of publishing on the Internet is Tim O'Reilly. Thank you.


The Ecology of E-Book Publishing

Tim O'Reilly: Thanks very much. Can everybody hear me all right? Okay, good. Despite being a high-tech person, I tend to speak without any high-tech aids. I'm a big believer in the conversational style of presentation.

Here we go. So, I want to have a conversation with you about my experience in two different realms of publishing --print and web publishing-- and draw some conclusions from those things about the brave new world of e-books that we're facing today.

The first thing I want to talk about is my evolution as a print book publisher, and I actually brought a couple of show-and-tell items. [Shows an early O'Reilly book, a pamphlet called Reading and Writing Termcap Entries.] Those of you who have been around a long time in publishing might remember these. This was one of my first books. I printed, I think, a hundred copies of this book at a local copy shop. Even though this is close to the beginning, this actually isn't a first edition. I know that because it has an index in it and the first edition didn't. These books didn't have ISBNs, so, again, I know this is not a first edition because it has an ISBN on the back.

My point is that I didn't know very much about publishing. Our first review lamented the fact that we didn't have any of the normal things that were expected in a book. We knew nothing about publishing. We were just a bunch of tech writers who thought, "Well, let's write down some things that we know about and see if we can sell these things." We sold these little pamphlets for five bucks and took them to a trade show in New York, UNIX Expo in 1985, and today we're one of the leading computer book publishers. We have a very well-known brand. I'll bring one of our contemporary books -- let me see. [Shows a copy of Programming Perl.]

Those of you who are out there in the computer book market know this brand -- these books with the animals on the covers. We're probably the number two or three computer book publisher in the country in terms of sales. We're very highly respected.

But, what was really interesting to me is the difference in the path that took me from that first book, the little pamphlet, to being a full-fledged computer book publisher as compared to a very different path that I experienced in the early 90s as a web publisher. I'm going to give you a little re-cap of that second part and then I'll pull the two back together.

We started out in the late 80s being interested in online publishing. We had developed a series of books on a technology called the X Window System, and these books were used by all of the large computer workstation vendors as documentation. They all came to us with a problem, which was that many of them were shipping our books as their documentation, and now many of these companies said, "You know, we'd like to do away with printed documentation. We want to ship only online documentation."

They came to us with what they thought was a wonderful value proposition. They said "Just put your books into fill-in the blank" -- it was IBM Info-Explorer, Sun Answer Book, HP LaserROM -- each one of them had their own proprietary system. And we said, "Oh, this doesn't sound like a very good business to us. We'll find ourselves always chasing all these different formats." So, we said, "No, we have a better idea. We want to come up with a common format for technical books. You guys all learn to read it." So, we started working with SGML and came up with a DTD for technical manuals called Docbook, which some of you may know a little bit about.

But, that wasn't the biggest part of the problem. I mean, it certainly solved some problems. We were able to represent our books online, even very complicated materials that are still too difficult for many of the e-book offerings today. But, that wasn't enough. We took a side-trip as we started working with Docbook. We said "Gee, we want to have a free reader that people can play around with," so you don't have to go buy one of these electronic book products to see this online documentation.

We got this cool tool called Viola, which was developed by a guy at UC Berkeley. It was an SGML browser. But, it also happened to be a browser for something else, namely the World Wide Web, which at that point had a couple of hundred sites. And we thought this was just the coolest thing ever. So, we took a big side-trip. We said "The Web is really interesting, so let's put this electronic book stuff aside. Let's build a Web site."

We actually started with a pilot program to put Internet kiosks into bookstores to try to help sell our Internet books. But what happened was interesting. The idea was people could type in things and try the Internet. But Pei Wei, the guy who'd written the Viola browser, set out to make this demo using hyperlinks, where you can just point and click.

We looked at that and we said, "That's not a demo, that's a product." And so out of that came this site called GNN, or the Global Network Navigator, which was a guide to what was on the Internet. At the time there wasn't very much. This was in 1992. We had published a book called The Whole Internet User's Guide and Catalog, which was a huge bestseller, and part of what was in that book was a catalog of various interesting Internet sites - it filled about 20 or 30 pages in the back of the book. We thought, "We can make a demo for how to connect to all of these sites using this new World Wide Web technology." And, that's really what GNN started out as, it started out as a demo.

GNN was really a prototype for a lot of things that you see today - we were the first online magazine, the first online catalog. Before sites like Hotwired or Salon, before Yahoo. So, all the sites that now are the staples of the Internet were really knockoffs of GNN. And it was also the first advertiser-supported site on the Internet of any kind. We got a special dispensation from the National Science Foundation to do advertising. So, we were part of that breaking down this idea that the Internet was only for academia.

So that's background, but what happened? We found that we had an enormous amount of heavy lifting to do. And that's my point.

In the book space, we were able to start knowing nothing, publish this little pamphlet, and we were sucked into an existing distribution system. We were taught by the system what the rules were. So, in book publishing, for instance, if the first editions didn't have any of the normal apparatus, people told us that and we knew just what to do. You know, we found out this was a system for registering books. We went, "Oh, good. We can do that." We got an account with Bowker and started filing our ISBNs and getting listed in the appropriate catalogs. And there's this big distribution system out there and eventually bookstores started calling up. Among the big guys, it was Borders who found us first. They were watching what was happening at an independent book seller called SoftPro -- and they saw somebody carry up a stack of our entire 10 volumes of X Window System books to the counter and they said "Who are these guys?" And so they called up and started buying books from us.

I think we were a $9 million dollar a year company before we sold books to Ingram [the largest book wholesaler in the country, and a prime sales target for every experienced publisher]. We didn't know who they were. We just didn't know anything about the system. But there was a rich ecology in place in the book industry, where there were wholesalers, there were retailers, there were various kinds of suppliers who came together and it was fairly easy to learn what the rules were and how to play by them.

Now we come over to my other story. Here we were with GNN. This thing on the Web turned out to be a very big and exciting business. But we had a problem. Nobody believed in it yet. We were very early, we were a small company, since it was at about that same time when we were just starting to really get traction in the book business.

I should add -- we started publishing in 1985. We really committed to book publishing in 1988 with this X Window System series of books. I know I'm rambling a little bit, but the whole thing was just really great, just how open and how easy it was to get into the business. We sold 10,000 copies of a two-volume set of X Window System books that we printed at copy shops on xerox machines. We took down some drafts to the second MIT X conference and people went nuts and so we went up to the copy shop and printed some more.

But anyway, so here we were with GNN and we're trying to persuade people that the Internet is going to be a big thing and, for example, that advertising might be a good business model.

And we had to commission the first market research study of the size of the Internet to try to persuade people that advertising might work. Nobody believed in it. We actually had to go out and start another business to support what we wanted to do with GNN.

The point is that the Web, in its early stages, was exceedingly difficult. We had to do an immense amount of evangelizing about the technology -- what would work, what wouldn't. It was experimentation on business models, experimentation on even how to execute on the business models that we imagined. So, for example, we now are very used to the idea of whether you have a banner ad and that ad clicks through to a different ad server. There are even companies out there that provide the service of rotating ads. Well, we had to do everything, even host the ads themselves, because the potential customers didn't have their own web servers yet-- we did it all.

We invested a great deal in GNN but eventually we had to sell it because we couldn't afford to keep it up even though it was potentially a terrific business.

Even now the web is not a fully developed ecosystem, but you can see that here, seven years after we started doing commercial Web sites, there's a rich ecology of players who help each other to succeed. There are companies that do analysis of your site traffic. There are companies that serve your ads. There are companies that understand to buy the ads. There are companies that follow the market and track who's got market share. There are people who resell ads on behalf of other people and ad networks. So, over time we developed in the Web space a mini analog to what we had earlier in the print space, which was lots of different companies working together in a kind of business ecology.

So, with all this as backdrop, I want to just talk a little bit about where we are right now with e-books. I don't know how long all of you have been working with e-books, but I'm now tracing my work in trying to get this puppy to fly, for 12 to 13 years. Actually, 14. Our first e-book was in 1986. And I see that the biggest problem is the lack of an ecology. And ecology is a really good metaphor for thinking about how marketplaces develop.

When you look at, say, ecological reclamations, for example what happened around Mt. St. Helens when it exploded, you have this gradual resurgence of species. One species makes way for another. So primitive plants will break down the rock and gradually make some soil so that something else can take root. There really isn't time to grow mature forests on, say, the bed of just-cool lava. This is also the way ecosystems develop in business. You have to say, "Okay, you've got to break this thing down."

And we're still in that stage for e-books. I know this doesn't really give us a great deal of guidance about what works in the e-book space, but it does give us a lot of guidance about what will not work.

What I think will not work are approaches that try to go it alone. For example, when FatBrain launched MightyWords, I had mighty words with Chris McCaskill, the CEO, because I thought his model was fundamentally broken. Why was it broken? Because he failed to think through the driving forces that make ecologies necessary.

Let me give you an example, just thinking mathematically about publishing. There are by most accounts 50 or 60 thousand books published a year. If we look at the Amazon-sized databases, we know that several million books are in print. Now, we also know that there are some tens of millions of customers. So, if we have an assumption that the model is the direct sales of all books to all customers, we have a massive mathematical problem. One million books times ten million customers means something like ten trillion possible matchups.

Now, if you're Stephen King, you're visible enough that you can in fact build a direct business. But the reason why wholesalers and retailers exist in the print publishing world isn't some plot, isn't some kind of entrenched monopoly. We don't have this seemingly inefficient distribution system because of diabolical market forces where people have seized the high ground and won't let it go. You have to deal with simple mathematical reality. Wholesalers aggregate publishers for retailers. Retailers aggregate customers for publishers. They make this thing mathematically more manageable.

So, when you're starting to look at distribution models for any kind of online publishing, you have to think through the math. The fact that it's online doesn't change anything that has to do with the way you reach customers. At the end of the day, if you're in a direct business now, you're in a pretty good position because you know your customer. But if you're in a publishing business that reaches its customers indirectly, you need to wait for eBook distribution networks to develop.

I had this conversation once with a guy who published a high-tech newsletter -- this was maybe in 97 -- and he said, "Gee, should I get on the Web?" I said "How many of your customers do you have a direct relationship with?" He said "All of them." I said "It's a no-brainer, because all you're doing then is offering them a choice of what delivery vehicle they want for your information."

On the other hand, when you're looking at, as MightyWords was, a class of information that's typically sold to anonymous customers, people who you don't have a direct relationship with (and in fact aren't likely to because there is not really a lot of commonality between the content that you're offering), then, you think "Gee, I'm going to take on trying to build a direct business out of nothing."

Now, would you do that in print? Would you say, "Hey, I'm just going to try to build a direct catalog business that will list thousands of random books?" Probably not. It's a bad idea. And so my complaint with MightyWords was that they were not looking at the way that the digital book ecosystem would eventually need to look like the print book ecosystem if it were to succeed.

The thing that makes me believe that e-book publishing may be poised to take off is that I'm starting to see businesses that don't say "Oh, yeah, we're going to do it all ourselves." I see businesses that say "Here's how we're going to enable this class of partner. This is how we're going to enable that class of partner."

When Microsoft gave their pitch on their e-book reader, a lot of it was about how this would work with publishers, how it would work with wholesalers, how it would work with retailers, how it would work with direct customers. There was some serious thinking there about the fact that you can't just collapse the entire ecosystem. Whereas MightyWords, in their initial incarnation, set out and said "We're going to collapse the whole system from publisher up through retailer. We're going to be a one-stop shop. We'll collect the content, we'll publish the content, we'll be the only source for reselling the content."

Now, I believe that distribution systems exist for the same reason that we have alveoli in our lungs. They create surface area. You know, any of you who have been in publishing know that there are two classes of customers. There are the people who already know that they want your product, who can come to you directly, and then there's the people who are going to encounter your product by chance. For most of book publishing and certainly for most of trade book publishing, the people who are going to encounter your product by chance are far greater in number than the people who are going to seek it out.

Now, I'm in a fortunate end of the business where, for example, a book like Programming Perl is the only book by the author of a program that's very widely used and so people say, "Oh, there's a new book out by Larry Wall" and they look for it. There are tens of thousands to buy it and there's a ready-made direct audience. And, certainly, you have people like Stephen King, or Prince, who has done this in the music world, who have already built up an audience over time and who can say, "Hey, come to me directly."

But, for the most part, digital publishing and online publishing systems are going to have to re-create the kind of richness of distribution networks that we see in the print world. So, the systems that we provide have to allow for the kinds of behaviors that have supported our print marketplaces. So when you're evaluating an e-book distribution system, you have to ask yourself questions like "Does it have pricing and mechanisms that support pass-through by multiple layers of wholesalers or retailers or distributors?" because you can't assume that you will have a direct relationship with everyone who might want to sell your books. Is there a mechanism for someone to pick up part of the margin? What does pricing look like? If you haven't thought that through, we don't have a viable system, or we have a viable system that will support only direct to consumer sales.

I don't want to beat on MightyWords too much because they have, in fact, been evolving their business model, but it was pretty obvious to me from the front end that they had some serious problems. For example, they went out saying "We will be able to give this great royalty to the author" but there was no margin in there for anyone else. So, out of the box, that was a winner-take-all game plan. It's one that says "Okay, we're going to get all the customers for ourselves."

In the technical book space, there have actually been a number of eBook "aggregators" who came up with business models that defied mathematical reality.

For example, there is somebody called came to us with what they felt was a great offer. They said, "Gee, we'll make you rich and famous. We're going to put your books on the Web." And so we said "Okay, tell us the deal." They said "Well, we're going to allocate 20 percent of the revenues to the publishers. We're going to charge customers $299 a year -- though it later fell down to $99 a year -- for access to these thousands of books."

And the thousands number stuck in my head because Ingram puts out a listing called the top one thousand computer books. I said "Let's assume you have a thousand books on Let's just assume we get the same percentage of those thousand books that we get out of the Ingram top one thousand, which was easily 15 percent of that list in our books. So, let me assume that I'm going to get 15 percent of your 20 percent." I did the math for them out loud and said, "So, ultimately, what you're offering publishers is 6 cents per book per customer." And they went, "Yeah, I guess you're right." I said "So, now you tell me how you think I'm going to make a business on that."

So once again, you see a flawed business model that doesn't have any room for an ecology in which multiple parties all make money.

At that point it's like a come-on offer with no follow through. I guess what I've seen in the early days of this market are a lot of ideas like that that just don't have staying power. And I said "Gee, maybe your business plan is this thing that will create buzz and you'll make money in the stock market and maybe you should be paying the publishers and authors in stock if that's the business plan, but don't tell us that this is a good deal when, if you do the math, it isn't."

I see a lot of very immature thinking in this space that somehow, because it's the computer world, that all the rules are different. You know, mathematics still apply. Business still applies.

One of the things that also bothers me about some of the digital rights software that's out there is that it makes it hard for you as a user to share information that you have. Now, there's a lot of fear about piracy, but there are some real dangers in the other direction as well. When I have a printed book, when I'm done with it, I can resell it, I can give it away to a friend. I very often will loan books to people who will then go buy their own copy. One book I found recently in a used bookstore I then went and ordered 15 copies of, simply because I thought it was a great book and I wanted to pass along. So, if we cut off things like the first sale right, used books, the ability to loan a book to a friend, in our approach to digital publishing, that pass-along ability, we cripple the ecosystem in a small way.

And I'm not suggesting that I know the answers for every aspect of digital publishing. I don't by any means. But, I do know some things that are signs that somebody's going awry. And those signs are trying to swallow too much of the market at once. They're not figuring out how many other people they can enable. They're not thinking through all of the people who may currently benefit or participate in the market.

This cuts so many different ways. It has a lot to do with how you think about royalties, for example. If the idea is that, as many authors would like to have it, digital books are a rights sale and, therefore, the authors should get 50 percent, then what they're saying is "Well, this is a short distribution chain. There's not many players in this."

That's an implicit judgment that people are making. And what I'm telling you is that markets with short distribution chains are generally small markets. If you want a big market, you have to enable lots of players. So, thinking through the math of everything from royalties to wholesalers to how we build a rich infrastructure for e-book publishing is a lot of what I want you all to think about.

So, the message that I really want you to just keep thinking about is that while things start simple, they grow complex. This is good. This is not bad. When the e-book marketplace is mature, we will see lots of intermediaries.

I can't emphasize how much the rhetoric of the early Web reminds me of some of the rhetoric of e-book publishing. What we heard was that the Web was going to disintermediate everyone. That was the big buzz word, disintermediation. We hear this again in e-book publishing. The authors are saying, "Wow, I'll be able to go out directly and sell e-books to my customers." Again, some ebook vendors, like MightyWords, are pandering to that: "You can just publish directly and the world will come to your door." And this was a mistaken impression. What really happened in the Web was that there were some illusions early in the market. There were skews that were caused by the fact that there weren't that many players yet. You were able to say, "Oh, yeah, everybody's equal on the Web. Everybody can have their own Web site." And that worked when there were only a few thousand Web sites. It doesn't work when there are millions, because of all a sudden the user says, "Well, I can't sort through all that stuff."

You guys have all worked with search engines and you never go past maybe page two of search results. So, we start to say, "Well, gee, who's going to help me find my way?"

Until we have the development of a series of intermediaries who are the equivalent of the Books in Print or the Reader's Guide to Periodical Literature, search engines, Yahoo, people like that, we won't have an eBook industry that works.

Anyway, I'm getting the five-minute warning so I should probably cut off and just open this up for questions.