Re. The Future

by Tim O'Reilly
September 1998

I wanted to respond to Joseph Sinclair's anecdote about the guy selling 5000 copies of the hundred dollar book online.

As I've noted before on the list, I worked my way up from self-publisher of a few books (albeit pre-web) to one of the five or six largest computer book publishers in America, so I have some relevant experience. Some thoughts:

1. Single data points do not make a curve. This is exactly the kind of thinking that leads hundreds of writers to try to produce the next bestseller under the current system. This anecdote has exactly the same value to the average writer as the following: "I know a guy who wrote a book aimed at dummies that sold several million copies." Or "I know a guy who sells 20,000 copies/year of his book from the back of talks that he gives." Or "I know a guy who got drunk one night, came up with a list of crazy ideas for bestsellers, and ended up selling one of them to Random House and going on to become a bestselling suspense-book author." I know all of these guys.

2. The key point is NOT that whoever this is sold his book online, it's that he somehow found some set of customers that thought his book was unique and valuable. This could be related to some clever marketing he did, or some unique content, or the fact that he correctly targeted a specific audience. But it likely has very little to do with the fact that he sold it online. If the book is good enough to sell 5000 copies online, I'd be very surprised if it wouldn't sell a lot more than that in bookstores.

   If you look at the history of self-published books, really successful ones are almost always taken over by commercial publishers, where they go on to sell orders of magnitude more copies. Mutant Message Down Under was a recent big example.

   Or take my own story: I started selling targetted books via direct mail, small ads in the backs of magazines, and by direct corporate sales. I had a $7 million/year business before I discovered bookstores, or they discovered me. But once I did, sales accelerated hugely, with many books that had sold 3-4000 copies a year suddenly jumping to 15-20K, and our top books jumping even higher, with one (the Whole Internet User's Guide and Catalog) becoming a million-copy bestseller.

   Similarly, if anyone does *really* well selling a book on the web, I'd be extremely surprised if they don't either end up subsequently selling the book to another publisher, or following the trajectory I did, and discovering all those other channels that at first were ignored. After all, if you can sell 5000 copies on the web, why not sell another 20,000 in bookstores? I guarantee you that a book that has sold 5000 copies/year on the web will get the attention of the bookstore buyers, just as mine did. Bookstores found me; I didn't find them (at least in the beginning), starting with specialty bookshops like Computer Literacy, Quantum and SoftPro. Then in 1992, Carla at Borders discovered us; we were her secret for a while, but then Ingram and Barnes and Noble caught on. At that point, we started to hire some retail sales people, and do more of what other publishers did.

   We've tried to keep doing what we did originally as well, with lots of emphasis on direct and corporate sales. But we would have been silly to ignore the channel that now represents about 60% of our business.

3. All that being said, if we'd tried at the outset to sell our books through retail, we would never have gotten a foot in the door. So selling on the web (which is a form of selling direct to consumer, as is the direct-mail and direct-response advertising that O'Reilly originally did) is a great starting point for a would-be self-publisher. It allows you to target a market that publishers and bookstores are still not aware of, or that is sufficiently specific that customers don't think to look in the bookstore for a title. But you still have to target the market.

4. As a good source of additional data points, I know a number of authors who have set up web sites to sell books they've written that are still in print, but largely ignored by their publishers. Some sell the books directly themselves, others through associates programs with Amazon, Computer Literacy or other online bookstores. Those of you using these Associates programs can ask yourselves how many books you're selling, and if you could live on that. Some people (Edward Tufte comes to mind) are doing hugely well by that channel; Amazon has led to a huge resurgence in sales of his books. But I'd guess that the average author who's trying to sell his or her existing books online is selling a few dozen a month. So if you want to experiment, you don't have to do it with your next new book: try figuring out how to sell and market one of your existing books online, and see how you do.

5. Early experiences are dangerous to judge by. Even if it's possible to sell a book today, it doesn't mean you will be able to do so tomorrow. As marketplaces become more crowded, the dynamics change radically. When the web first took off, it was easy to make a splash. At O'Reilly, we created GNN (Global Network Navigator), the first commercial web site; it was easy to make our mark on a shoestring. As the web was discovered, the ante went WAY up, and we decided to sell out (rather than put the company in play by going public), because we could no longer afford to play at the current stakes. All of a sudden, it cost tens of millions of dollars a year to get and maintain visibility.

   Success tends to breed imitation. Once it does, McGilton's law (which I've quoted before on this list) kicks back in: "First, Fabulous, or F***ed." (Actually, in a really hot market, it's probably more like "First, Fabulous, and very deep-pocketed, or F***ed.")

In short, I think self-publishing and selling via the web are really worth exploring for would-be authors. But don't think that some lousy book that you crank out in a hurry is going to make your fortune. Even more than with books put through normal publishing channels, you need to have the right product at the right time, and stick with it until you find your customers.

So it's the same advice that you'll always hear from me: write books that serve a real audience and meet real needs--and that means writing books on subjects that aren't already exhaustively covered, or about to be by the time your book comes out--and you have a unique opportunity to succeed. To actually capitalize on that opportunity, of course, you have to know your audience and how to reach them, and keep at it until they start telling each other about your product. (That's the classic definition of a market: a group of customers who reference *each other* when making buying decisions. That's what you ultimately need to succeed as a publisher, not some magic bullet. Figure out what it will take to get people saying "This is the book you've got to have.") In the end, the web is just another channel. All the rest of what goes into a successful publishing business stays the same.