Today we announced the selection of Andrew Savikas as the new CEO of Safari Books Online. Andrew wrote a post about his new position where he wants to take the company, and I wanted to share that with you. I’m excited about Andrew’s vision, and I think he has the ability to take Safari to a new level of technical sophistication and user experience. But I also wanted to take a moment to give a few of my own perspectives on Safari’s history – and in fact O’Reilly’s long history with ebooks – and more importantly, its future and by analogy, the future of publishing.
In many ways, the road to Safari began with Dale Dougherty‘s work to get our books online. His very first effort was a version of Unix in a Nutshell for Hypercard published in 1987! It was a product far ahead of its time, but it led in two fruitful directions.
The first was the development, starting in 1991, of Docbook as a standard way to represent technical books in SGML, and later, XML. (Docbook was developed by a consortium called The Davenport Group, which Dale co-founded, and our X Window System books were the “test subjects.”) What you need to remember is that the web didn’t yet exist. There were many online, hypertext, multimedia systems for content, but all of them were proprietary and non-standard. Because our X books were adopted as documentation by all of the leading Unix workstation manufacturers, we were being asked to “port” the books into a variety of these online reading systems: Sun’s AnswerBook, IBM’s InfoExplorer, HP’s LaserROM, and many others whose names have faded into memory.
Influenced by the rise of open source software, and in particular by Bob Schiefler, the head of the X Consortium, who had a wonderful vision of how standards could raise the level of the floor on which companies could then innovate, we decided that the best way we could influence the future was to encourage standards for formatting content, with lots of innovation in the ways to read it.
Back in the late 80s and early 90s, we assumed that one or another of these proprietary reading systems would likely emerge as the winner. (After all, these were the days in which our expectations were shaped by the rise of Microsoft Windows, and the power of open source software and the open content standards of the web was not yet fully understood.) But we knew that reading would remain open as long as books were in a standard format, and that people had the option of a good free e-reader.
And that’s what led Dale in 1992 to Viola, which was the first graphical web browser. That was the second big outcome of his explorations, as we adopted Viola, and used it to build the first ad-supported catalog of the web — effectively the first web portal — GNN, or The Global Network Navigator. We sold GNN to AOL in 1995, and by then the web was off to the races.
But we continued to produce our books in Docbook, with the idea that eventually ebooks would become commonplace, and we’d need a set of tools that would let us produce print books and a variety of web and ebook formats from the same source files. That vision has come to fruition today, with a production pipeline at O’Reilly (and for our digital distribution clients like Microsoft Press) that allows us to easily create ebooks in multiple formats. (Our typical catalog offering on oreilly.com, for example, offers people the choice of a book in print, on the web via a Safari subscription, or in the user’s choice of any of five different DRM-free ebook formats: pdf, epub, mobi (for Kindle), apk (for Android), and Daisy (for software that supports vision-impaired readers.)) It’s relatively trivial for us to add additional formats as they become available.
Incidentally, it was Andrew Savikas who led the development of the modern version of that pipeline, in his role as the head of our publishing tools team at O’Reilly from 2003 to 2010.
Here’s the key takeaway: our original ebook vision was of a world in which ebooks would be published in standard formats and could be read on any device, and where dominance of a particular piece of software or a particular e-reading device would not lock people in. We wanted reading to remain as open as it did when printed books ruled the publishing world.
When the web took off, we thought that battle had largely been won. But the new world of dedicated e-reading devices with proprietary formats has brought back the specter of a single company having a monopoly on the tools for reading – and publishing – long form content.
It’s important for publishers to take a stand for reading portability and open access. Offer books in multiple open formats, DRM-free, and from multiple points of distribution, or the world of reading will eventually come to resemble the world of desktop software in the age of Microsoft’s dominance.
The road to Safari’s subscription business model grew alongside our conviction that ebooks would eventually become the dominant way to consume book-length content, and that those ebooks needed to be in formats that were open and portable.
As early as 1994, I’d given a talk and in 1995 written a paper, Publishing Models for Internet Commerce, about how the web would eventually need to support all the same kinds of business models we saw online, including paid content, subscriptions, and free/ad supported. (I have the dubious distinction of being the person who first introduced the ad-supported model to the web, in early 1993.)
I was also convinced that moving online would change the nature of the book. “A book is a user interface to information,” I wrote. And it’s not necessarily the best interface. Some kinds of information were books by convenience only – an atlas, say, or a dictionary, or a bird identification guide. Applications can do the same job far more easily.
One early insight I had was expressed with the line “An online book will either be much larger or much smaller than the printed book.” We were seeing how the web was allowing readers to consume content in much smaller bites, but at the same time we were seeing the way that information could be aggregated into huge searchable works. It was becoming clear that search was the killer app, and searching one book at a time wouldn’t cut it.
And that was where we decided to go with Safari: a searchable repository of all our books, with a subscription business model that would potentially give people access to everything we published. But the more we thought about it, the more we realized that important as O’Reilly books were to our customers, we didn’t cover every topic, and the service would be even more powerful if we brought in other publishers. That’s when we approached Pearson Technology Group, one of our biggest competitors, to join us in a joint venture, and invited other publishers to participate as well.
Safari Books Online now contains more than 17,000 books and videos from dozens of publishers, and is the richest source for professionally curated technical and business content on the web. And we’re somewhat amused by the disbelief that the “paywall” erected by sites like the New York Times appears to be working, since we (as well as a host of scientific and professional journal publishers) have been serving our customers with subscription-funded sites for the past decade or more. As I predicted back in 1994, the web has room for every business model imaginable, and smart publishers will learn to exploit all of them.
And that leads me to the future: users want choice. They want choice of device (which means choice of format), and they want choice of business model. Subscription is the right model for institutions and for heavy users; pay-per-view (i.e. standalone ebooks) are a better model for occasional users. Ad-supported appears to be the best way to fund fast-changing current content. And of course, some content is better rendered as an app than a book.
And that brings me back to where Andrew is taking Safari. It was originally designed as a just-in-time reference service, but as we’ve increasingly realized, it’s actually a just-in-time learning service. With the addition of video, we’ve taken it partway in that direction, but there’s lots more to do.
In addition, as people buy our books in multiple formats, we need to erase the boundaries between where and how they buy them. Andrew is focused on making Safari an “access anywhere” service that’s far better optimized for mobile and tablet consumption, as well as keeping up with the competitive challenge of making sure that reading remains an open platform.