Today I had a project rejected by a publisher. It's the author's third book and the prior two books had been well received and have sold well. But according to this publisher, "the Bookscan sales of his two titles have been modest in comparison to the great praise and attention his work has received, and in this economy that’s a very difficult obstacle for us to overcome with our accounts and booksellers."
There's only one problem with this argument: The Bookscan numbers are wrong.
According to the royalty statements I've received, the author's first book has net sales of just under 14,000 copies. According to this editor, Bookscan shows sales just over 7,200 copies. That's nearly a 100% difference!
Questioned about this, the editor responded, "Bookscan is indeed an imperfect tool, but it’s also enormously helpful to us, and we do interpret the information with the proverbial 'grain of salt.' Our rule of thumb is that Bookscan captures about 70% of retail sales, give or take. In this case, Bookscan shows sales of just over 7,200 copies for the ___________________ book, so it seems that the accepted formula may be a bit low here. 14,000 copies is a strong performance for this title, and it’s great that this one has earned out for the publisher."
A bit low?! Bookscan is reporting half of what the author's own royalty statements show and, in the case of the author's second book, Bookscan is reporting approximately ten percent of what the statements show.
Are publishers really so dense that they haven't compared Bookscan's figures with their own sales figures? Surely if they have, then they would have stopped paying Bookscan for its clearly and outrageously wrong data and put it out of business for lack of subscribers. Because, you see, Bookscan is pretty much only of use to publishers and to, say, news organizations writing about publishing. Sure, it's supposed to be the Nielsen Ratings of books, but Nielsen Ratings for TV and radio have a purpose: they tell advertisers where it's worth spending money on what shows. But books have no ads, so what is the purpose of Bookscan? To prove that the NEW YORK TIMES best-seller list is wrong? That the number one book this week is not RELENTLESS, by Dean Koontz? That the number two book isn't THE PHYSICK BOOK OF DELIVERANCE DANE? (Really? That's the number two book?) Do we need some ultimate decider beyond the NYT or Barnes & Noble or Amazon? I don't think so.
I am one-hundred-percent sure that if Bookscan were reporting higher numbers than publishers, publishers would pull their subscriptions and put them out of business. Why? Because agents and authors would be hammering publishers and demanding to know why the publishers are reporting lower numbers and where the hell are our royalties?! But because Bookscan reports lower numbers, publishers happily use its data to crush authors and insist that they can't pay what the author or agent believes a book is worth because Bookscan says the author isn't selling as many copies as the author says he or she is. Even presented with actual royalty figures, publishers seem to still favor Bookscan, which I think puts a burden on publishers to ensure that Bookscan gets it right.
Thus, I propose that publishers should have to report their sales to Bookscan and Bookscan should have to reconcile the differences. After all, don't publishers want accurate reporting? Doesn't Bookscan? So if Wal-Mart or Amazon or Target or whomever that isn't reporting to Bookscan won't provide the figures, why shouldn't the publishers? Sure, the publishers would be reporting wholesale numbers and not point-of-sale figures, but certainly that's better than having no data at all?
C'mon, Nielsen, put your money where your mouth is. And publishers, too. If you want to make or break authors' careers and base your acquisitions on Bookscan data, then you need to have accurate data, which you don't. So, please, put up or shut up.