June 2003

Subject: Print Numbers and Sales Expectations
From: Justin Johnson

Tim,

Thanks for the So You Want to Write a Book articles on your site. A shame other publishers don't take note and provide such useful information.

It would be helpful to know the average of your print run numbers. I'm aware that this will be different from book to book. It would be most helpful for people like myself who are seriously considering authoring and must take into account the personal business aspects of doing so.

Many thanks in advance,

Justin Johnson


Good question, Justin. Unfortunately, I don't have a simple answer. In the old days, the initial print run on a book was an indication of how a publisher thought it would do. And even today, you'll see trade publishers touting initial print runs as an indication of the book's bestseller expectations, before they've sold a single copy. (In fact, many of those announcements are trial balloons, and the actual print number may vary by quite a bit.)

In the case of O'Reilly, we've moved so much to "just in time" printing, because we get five day turnaround from our principal printing vendors, that "print runs" are not a very good indication of sales expectations. We often print as little as a six-week supply for any print run, versus the six to twelve months we used to target.

So perhaps the right question isn't what our print runs are, but what our sales expectations are. There too, there isn't a simple answer, but I can give you some interesting statistics.

The number I like best is lifetime sales--after all, when we publish a book, we like to think it's going to stick around for a long time, selling year in and year out.

Let me take a quick gander at the lifetime sales report on our intranet (which only goes back to 1996, when we put our current order-entry and reporting system in place). I see a list of about 600 books. That includes titles that were published before that date (and so have only partial lifetime sales), special SKUs like "BN Bay Carton #4" or "Smiley 5-Pack," and multiple editions of the same book treated as if they were separate titles. This list also includes the approximately 120 new books and new editions we've published this year, some of which have only a few months' sales. All numbers are net of returns.

OK. Here's what I see for this six-year period from 1996 to the end of 2002:

  • Median unit sales -- 15,078 copies
  • Average unit sales -- 25,953 copies
  • Largest single-edition unit sales -- 319,176 copies
  • Largest multiple edition unit sales -- 654,341 copies
  • Smallest unit sales for a book with its lifetime fully played out -- 3,548 copies

In general, we have very few "bombs" like the last one in that list. And most of those are books that we (and the author) went into with our eyes open. It might be a topic, for instance, that the author is passionate about, that we hope there will be a market for, but for which it's too early to tell. Sometimes, it's even a book that we think will never sell a lot of copies, but that's important to put into print. In cases like that, we need to set the price appropriately high, if we can, and if we can't, we sometimes just grin and bear it.

If you only go after the "sure things," you'll always be a "me too" publisher. Many of our most successful books started out with very small expectations, and grew along with the market. Programming Perl, which is one of our all-time bestsellers, started out selling 6-8,000 copies per year back in 1991, but has ended up selling well over 600,000 copies in its lifetime. Our very first published book, Learning the Unix Operating System, had a first print run of 100 copies, but has gone on to sell many hundreds of thousands of copies, and is still selling tens of thousands of copies a year, seventeen years after it was first published.

Still, in terms of calculating your expectations, the vast majority of our books sell between 10,000 and 30,000 copies in a given edition. At average prices, discounts, and royalty rates, that means that you can probably expect to earn somewhere between $15,000 and $60,000 by writing an O'Reilly book. We have more than a few authors, though, who have been both extraordinarily productive and good at picking the right topics, who consistently earn royalties in the mid-six figures, and have made millions of dollars over the lifetimes of their books. And there can be collateral benefits as well, as many authors have enhanced a training or consulting business with the reputation that's come from writing an O'Reilly book.

A few caveats regarding income as an author. . . . Writing a book is often a labor of love, following on the heels of many years of technological experience. Some people can write a book in a few months, but many take close to a year. $15,000 is good money if you make it writing a book in a few months of full time work, or while holding down a day job. (That's why most computer books that are "day and date" are written by professional authors who are able to write quickly, and most in-depth or professional titles are written by people for whom writing is a sideline and so take much longer to produce.) New and rapidly changing topics often require revisions every 12 or 18 months, so you should budget for that as well.

Furthermore, although O'Reilly offers advances against royalties (as do most publishers), you won't see all the potential royalties up front. Even if your book earns $60,000 over its lifetime, most of that isn't seen until a year or more after the manuscript has been completed (given production time, the need to recoup the initial advance, and the 90-179 day delay from the time a sale is made until the quarterly royalty check is received).

Many of our authors are financially successful, but the most successful ones are those who have multiple books in print and have devoted the time to become efficient at the craft of writing. You should approach book publishing with your eyes open. Most skilled programmers can make more money outside of book publishing, and so many make writing part of their overall career mix, leveraging the additional reputation gained from their writing to get consulting gigs or improve their job prospects. Write about topics that you care passionately about, and if it pays off financially, that is icing on the cake.

Paradoxically, your passion for a topic is often the best route to financial success as well. At O'Reilly, we publish books that our editors, marketing staff, and readers are enthused about and believe in. Proposals from passionate and involved authors are the most likely to be accepted and to ultimately succeed as books in the marketplace. If you wait for a topic to be a sure thing, you may miss the pent up demand that will be harvested by the author and publisher of the first good book on a new topic. And the way to be out front is to follow your passions and interests, and those of your peers, rather than trying to understand what is already selling.

One final note--the small print, you might say. As mutual funds are required to note in their prospectuses, past performance is no guarantee of future performance. Like the computer market as a whole, the computer book market is in a period of serious contraction right now. Initial buys by the large bookstore chains are at all-time lows, as is the restocking of backlist titles. At the same time, there are still many opportunities. David Pogue's Mac OS X: The Missing Manual is our best-selling book since The Whole Internet User's Guide and Catalog back in 1992, selling well over 200,000 copies last year.

Meanwhile, I believe we're entering a period of enormous change in how people use computers, and their pervasiveness in everyday life, the most significant change since the introduction of the personal computer. So I do believe there will be many more great publishing opportunities ahead.

One good example is our recent book, Google Hacks, which has topped the bestseller lists since it was published in February. Google has not only become a verb, and a killer application that almost everyone on the internet now relies on, but it's also on its way to becoming a platform, with people building its functionality into other applications and sites. As the web becomes ever more pervasive, there will be new and surprising applications that need to be documented.

New technologies like digital photography can also create new opportunities--or add momentum to old ones. Photoshop was a nice niche market until digital photography made it the strongest single area in computer book publishing during the past twelve months.

If you have questions about the potential of specific topics, send us a query. If we think the topic has promise, an editor will get in touch with you, and you can ask about what our specific sales expectations might be. We'll do our best to be honest with you. Publishing works best when both author and publisher have the same goals and expectations.

I hope to hear from you.

Tim

Got a question? Just ask!

Return to: Ask Tim