Measuring Success on Self-Published Titles

Over on his blog at Adobe, Bill McCoy comments on the numbers used in a Wired article to demonstrate "success" for a self-published book:

No insult intended to author Zeraus nee Suarez who "is planning to release a sequel". It may be great stuff. But [1,200 copies sold] are not stats to write home about, much less to hang a "who need Random House" thesis on. Per an established literary agent: "Less than 5000 actual sales, result: misery … A solid midlist novel would reap on the order of 3,500-7,000 hardcover sales and 10,000-25,000 paperbacks in the US."

Bill is absolutely right that as measured by the yardstick of trade publishing, 1,200 copies can’t rightly be called a success. But do self-publishing authors use that yardstick?  Let’s build a simple scenario and run some numbers. Of course every author is different, but I think it’s fair to say that for many authors, two big goals for writing a book are to (1) make money and (2) elevate their reputation.

I’m assuming that for #1, most authors interested in self publishing are modest enough to not expect to get rich from their first (and self-published) book. Instead, they’d like merely to earn the equivalent of a fairly standard advance on royalties. And for #2, let’s assume that the marketing outreach needed to sell that modest amount (through, for example, blogging or bootstrapped book tours) will effectively elevate their reputation enough to be counted as a success. That ties the two goals together so they can both be measured via sales numbers.

To make things easy, I’ll use my own book, Word Hacks, as an example. In order for me to earn out the advance O’Reilly paid me, the book needed to sell roughly 8,000 copies. Certainly not a blockbuster, but according to Nielsen BookScan data, only 2% of the 1.2 million books sold in 2004 (the year my book was published) sold more than 5,000 copies.

If I’d instead gone the self-publishing route, based on the cost structure of working with, how many books would I need to sell to earn the same amount as that advance? Roughly 500.

Number of copies needed to earn equivalent of publisher advance.





Granted, selling 8,000 copies and having the association with a well-respected brand like O’Reilly is certainly going to do more for reputation than something self-published. But the recognition I’d have had to build with bloggers and their audience to sell those 500 copies would arguably put me among the top choices if and when a publisher like O’Reilly goes looking for their own author.

If you accept my generalizations, the numbers above suggest that in terms of author income, selling one self-published book is roughly equal to selling 16 through a publisher. Getting back to the Zeraus nee Suarez book (that’s a pseudonym, btw), 16:1 is the conversion factor needed to make a (somewhat) more accurate comparison with trade sales. Self-publishing and selling 1,200 is therefore more like selling almost 20,000 through a traditional publisher. Not too shabby.

There are some flaws in a rough comparison like this, but my point is that if publishers use their standard sales measures to judge the performance of self-publishing authors, they are underestimating the "success" of those authors.

tags: , ,