When I appeared on The Write Lines last week I mentioned that the average self-published book sells between forty and two hundred copies, depending on which set of figures you consult. Compared to mainstream publishing, where sales of three thousand copies for commercial fiction are considered by some to be disappointing, these figures are terribly low and the reaction from the other studio guests (a literary agent and three successful mainstream-published writers) was obvious: if you listen to the recording you can clearly hear them gasp.
After the broadcast I caught up with the reaction on Twitter, and found that a few writers were discussing the figures I quoted and reaching some rather unsound conclusions. While I'm not going to quote anyone here (it’s just not appropriate to single anyone out), I do think it’s important that I respond to their points. But first, a very quick primer on how sales figures are usually gathered in the book trade.
Nielsen Bookscan collects the sales figures of various online and physical retailers, then collates those figures and reports them to the book trade (it’s Nielsen which produces the best-seller charts we're all so envious of). However, as relatively few copies of self-published books are sold through bookshops, and quite a few self-published titles don’t even have the ISBNs which are essential for books to be tracked by Nielsen, the majority of self-published sales aren't included in Nielsen’s sales reports: therefore, if you rely on Nielsen to provide sales information about self-published books, you’re likely to be way out of whack with the real picture.
The Tweeters seemed to assume that I was relying on Nielsen, and that therefore my figures had to be way off.
They would have had a very valid point if I had relied on Nielsen’s reports for my statistics, but I used a far more generous source for the figures I quoted on air: the publishers themselves (there’s an obvious difficulty here: my figures came from companies which style themselves as self-publishing service providers, which many consider to be vanity presses: but for the purposes of this discussion I’ll ignore that issue, which is a little off-topic here. I shall return to it at another time, have no fear).
As most publishing service providers of this type offer only print-on-demand services, copies of the books that they publish are only ever printed in direct response to an order; a copy printed is a copy sold, no matter who buys it. So long as a book is printed and sent out from their premises they consider it a sale—when these companies report sales what they’re really reporting is the number of copies printed. You can see that there is not going to be a hidden stash of sales which fail to get included in the sales statistics: if anything, these companies are likely to over-report, rather than under-report, their sales.
As the sales figures I quoted came directly from POD-based self-publishing service providers, not only do those figures include all copies sold in bookshops or by Amazon, etc.; they also include all copies subsequently returned by bookshops to their authors (because with self-publishing the author is the publisher, and so they have to credit the bookseller’s account for those returned books even if they’re no longer in a sellable condition); they also include as sales every single copy that the authors bought and then sent out, for free, to reviewers, or gave away to family and friends; and every single copy which all those hopeful authors ordered, only to have them left mouldering away in their garages when they found they couldn't sell them—of which there are far too many.
Which means that it’s impossible for the sales figures I mentioned to be under-reported: unlike Nielsen's figures, they are going to be higher than the sales which really count in an author's career: the sales made to interested readers who considered the books potentially good enough to pay their hard-earned money for.