The Long View

by Tim O'Reilly
September 1993

I started out this month with several separate articles. Then I realised that three of the articles have a common thread, and thought that I'd start by explaining just what that thread might be.

Years ago, back when I was starting O'Reilly, I read a science fiction book called Rissa Kerguelen, by F.M. Busby, that influenced me deeply.

One key idea in the book was that of "the long view"--the idea that some developments take years, and that you cannot control them, but can set them in motion, gauge their direction, and meet up with them years later. (If you've read much science fiction, you know that one corollary of some of Einstein's theories is that if you were to achieve star travel at near-light speeds, years would elapse on the outside while months elapse for the travellers. Busby used this principle to create a civilization in which it was essential to take the long view.)

We're not travelling at near-light speeds (though it may sometimes feel that way), but the model of decision-making Busby described struck me very deeply at the time, and still shapes my thinking in major ways.

I thought I'd spend a couple of issues of the newsletter talking about the long view. Rather than starting with global statements of purpose, or reviewing the history of the company in light of this concept, though, I'm just going to try to share how various things that are going on right now appear to me from the "long view" perspective...starting with these three seemingly unconnected topics.

The Davenport Group and the docbook dtd

As you may recall, the Davenport Group is an informal consortium that Dale organized a year or two ago to help create a common distribution format for online documents. We'd been approached by numerous vendors wanting us to "port" our X books to their proprietary online documentation systems. Dale looked down that path, and saw a future of endlessly maintaining multiple parallel versions of our books.

He didn't like that vision, so he set out to create a common interchange format for electronic documents, together with a consensus among vendors and publishers that that was in fact the right approach to take. This effort has been rolling along, quietly gaining momentum.

The Davenport Group currently consists of the following group of core "sponsors":

Jon Bosak, Novell, San Jose, California
Dale Dougherty, O'Reilly & Associates, Sebastopol, California
Ralph Ferris, Fujitsu OSSI, Emeryville, California
Dave Hollander, Hewlett-Packard, Palo Alto, California
Eve Maler, Digital Equipment Corporation, Boston, Massachusetts
Murray Maloney, SCO, Santa Cruz, California
Conleth O'Connell, HaL Computer Systems, Campbell, California
Nancy Paisner, Hitachi Computer Products, Boston, Massachusetts
Mike Rogers, SunSoft, Mountain View, California
Jean Tappan, Unisys, San Jose, California

One outcome of the group's work has been the development of an SGML Data Type Definition (or DTD) called the Docbook DTD, which we developed with HaL Computers based on what would be required to put our own books online.

Terry has been doing a lot of the ongoing development of the DTD, along with Con O'Connell of HaL; the two of them have been designated the Davenport "design team." I concur heartily with Dale's comment: "I give Terry a lot of credit for rather quickly becoming an SGML expert. Also, he has done an excellent job of listening to feedback from others and modifying the DTD to balance the needs of new groups of users as well as maintaining a consistent approach to the design of the DTD."

While there's never been a big splash about Docbook, and even our own efforts to implement online versions of our X books using it have fallen a bit onto the back burner due to GNN, you should know that it's been making slow and steady progress as THE standard for online documentation. Both USL and Novell originally agreed to use it for their online documentation efforts; now Digital has joined in, and the UNIX International docsig has made a recommendation that UI formally adopt it as the documentation interchange standard between member companies. (OSF has a competing DTD, but no one has ever really bought into it, and the membership, except IBM, have deserted it.)

This long-term background project (How many of you have wondered "Just what is it that Terry does?") will eventually result in interoperable online documentation products. In fact, we're in discussions with Novell/USL about doing a joint "UNIX library" on CD ROM using this technology.

Jon Bosak from Novell wrote recently: "DocBook SGML has clearly become the leading interchange format for software documentation, and Davenport currently provides the forum for input to the standard."

What's more, the path Dale followed with the Davenport Group led him eventually to SGML, to the World Wide Web, and to the technology that underlies GNN and other exciting online projects that you've no doubt been hearing about.

This is exactly the kind of thing that I mean by "the long view." You don't know exactly where you're going. But you know what you're looking for, and you trust yourself to know when you find it. This approach takes a lot of courage, a lot of tolerance for ambiguity, a willingness to pass up short term opportunities that somehow don't seem quite right, an ability to change direction suddenly when you realize that one path you've headed down doesn't quite take you where you want to go.... I could go on and on.

The point I'm trying to make is that this effort has been entirely in the same spirit that launched our book publishing efforts in the first place. That spirit of exploration driven by "I know what I'm looking for" that's so foreign to most business planning is the very essence of what has made ORA successful.

I also want to give some much deserved praise to Dale and Terry for working so long and hard on something that for several years didn't have many tangible results, but only the sense that somehow they were getting closer to the goal. Now, of course, with GNN, the results are all too tangible, and they're probably half wishing they could go back to some quiet research! (Not really. I know they're actually incredibly excited about the turn things are taking.)

Are we still an editorially-driven company?

We've always prided ourselves on following interesting projects, and not worrying so much about short term return and market opportunities as making sure that we were doing worthwhile books that would eventually find their own market.

In the past couple of months, though, a couple of people have observed the growth in staffing, visibility, and success of our sales and marketing efforts, and have asked me if we aren't somehow at risk of changing the company culture, of becoming driven by marketing opportunities rather than by our editorial vision.

I don't think so at all, but I do think that the topic deserves a little discussion in the event than anyone other than the people who asked me have had similar thoughts.

To me, creation and marketing of our books are and have always been inseparable. From day one, the books we've created have been shaped by a keen sense of a gaping need out there in the world and the knowledge of how to fill it. That's the essence of both editorial vision and marketing savvy.

Our sales and marketing departments have grown a bit faster than the rest of the company for three reasons:

1. Sales and marketing doesn't just have to support the new books that the editors create, but all the old ones as well. Unless we want to fall into the "frontlist" mentality that traps most publishers, we need to develop marketing strategies and support programs for all those old books that the editors are not quite so interested in any more, and that have a tendency to fall off in sales.

2. We're still playing catch up as we enter new markets. In fact, we're investing in marketing in the same way that we invest in editorial and writing: to help create interesting new possibilities. So, we're investing heavily in the development of international sales, and trying to figure out new channels such as government, user groups, libraries, and so on. In addition, we still need to recognize that we still haven't reached as deeply into our existing markets as we can.

3. Precisely because we are driven by an editorial vision that finds interesting things off the beaten track, we are entering entirely new markets, which require proportionally far more attention. Books like Love Your Job, Outsider's Travel Guides [later renamed to Travelers Tales], and even Oracle Performance Tuning require specialized marketing efforts. For that matter, our new focus on the Internet requires us to reach out to new customer groups. We're past the point where we can look at all of our books at appealing to the same monolithic user block (UNIX hackers). We need to develop separate marketing programs for Internet books (which include PC and Mac users), UNIX programmers, UNIX users, UNIX system administrators, and so on. The more we diversify our product line, the greater the demands on marketing.

I also find it somewhat baffling that anyone could think that we've abandoned our old long-term view of editorial development in the face of speculative development projects like GNN, travel books, Love Your Job, etc. These are pure "field of dreams" products in the grand ORA tradition. Find something really interesting and just do it, trusting that you'll be able to find the market for it later.

The one area that I can see where some concerns might be justified are perhaps in a couple of instances where we've allowed ourselves to be drawn into "horse races" to get a book to market when we hear about a competing book coming out. I'd say the jury is still out in this area. We've probably become the number one competitive target in computer book publishing, and we're going to see a lot of people trying to one-up our titles. For the most part, we can probably remain above the fray, but there are other cases in which we may need to move aggressively to secure a market that we plan to be in before someone else makes us look like a "me too" player. As in all things, it's important to find a balance. But I certainly don't think that we're out of whack here, or have a "bestseller fever" that is making us abandon our more considered, creative approach to new book development.

To me, the real danger we face is in not beefing up our marketing and sales capabilities to match our editorial output. If that happens, we have "orphaned books"--books that were developed to fill a market need, but never have the support to reach that market.

The basic message that I take from the changes in our approach is that as circumstances change, the company has to change too. But this doesn't mean its goals are changing. In fact, it is more likely to mean that we've still got our eyes fixed firmly on the goal instead of getting hung up on the ways that we used to get there.

Our Foreign Rights Strategy

I just got out of a meeting trying to "prep" Cathy Record for some of her foreign rights meetings for the Frankfurt Book Fair. As I came out of the meeting, I found myself saying: "Yup...same message as in those two articles I was writing for the newsletter." So I decided to write another installment.

Cathy's got some great leads for foreign translations of our books (see her report elsewhere in this newsletter). As we were going over them, I was explaining why I found the ones involving Russia and South America more exciting than the ones involving more developed markets...and what she should be looking for in foreign publishing partners. Some of this has to do with the "romance" of these untapped markets. But part of it is this same "long view" thing. How is it that we're best going to reproduce overseas the kinds of things that have made us successful here in the U.S.?

I found myself saying: Look for people who really care about the books, and are excited about the need for the information, rather than those who have the best distribution or the best lock on the market.

Why? Most publishers simply look for the "best deal" when selling foreign rights. It's a source of incremental income, not to be ignored, but hardly something to spend a lot of effort on. By contrast, we want to use foreign rights assignments as a way of getting to know a market, so that long term, we can develop the same kind of business there that we engage in here.

We've followed this strategy very effectively in Japan (after some early mis-steps in Germany and Italy).

The key elements of our foreign rights strategy are:

1. To make sure that the translations are really top-notch, by requiring that they use technical reviewers approved by us. (When we haven't done this, the results have been quite uncertain.)

2. To work with people who can get excited about the books, and if possible are part of the market we're trying to reach. (In many cases, we're finding translators and would-be purchasers before we're finding publishers to do the local publishing. In some cases, we may find ourselves doing local publishing ourselves before we know it.)

3. To create opportunities for long term business opportunities in the target countries. (E.g., it could be years before Russia produces any significant revenue, but there are people there who really need our books, and could benefit from them, so why not. We'll meet up with the markets as they emerge.)

I'm always amused a bit by the way I see myself reacting to various business opportunities that come up. Almost always, the ones that are exciting are not the obvious ones that will make easy money, but the ones that seem to me to be a stretch, that are difficult because they are addressing a need that has been hard to fill.

Not so oddly, those are also the kind of opportunities that do pay off best in the long run.