The other day I came across a blog that essentially accused the publishing industry of “corruption” and my own publisher, Tor, of “gaming” the system to produce artificial bestsellers.  While there is truth in the fact that Tor and indeed all publishers do their best to release titles in a way that will maximize the possibility that new releases make various “bestseller” lists, the idea [or at least this particular blogger’s idea] that the “big” publishers are somehow actively corrupting publishing is based on some serious misconceptions about bestseller lists, both in how they’re compiled and how well/poorly they reflect sales.

To begin with, the lists don’t represent total sales or anything close to that.  They represent sales in a given period, based on data from various – and often differing – sources.  Publishers Weekly now bases its lists on Nielsen BookScan numbers, which, in general but not always in specific cases, comprise about 60% to 80% of the major retail sales points, depending on which expert one talks to, but that leaves major gaps, such as Walmart, whose numbers aren’t included in BookScan.  The New York Times does not reveal its specific methodology, but has stated that its survey includes 4,000 bookstores as well as representative wholesalers. USA Today surveys some 3,000 retailers.  In the F&SF area, Locus magazine does its own survey of bookstores and publishes its own monthly F&SF listing, which, according to some editors in the field, may be more representative of the genre, depending on one’s definition of fantasy and science fiction.

But there are other factors to consider as well. While the BookScan sales numbers do represent actual sales, all “pre-orders” or sales made before the release date are tallied for bestseller numbers in the first week’s sales.  That means that a book that’s been listed at, say, 5,000 on the Amazon sales ranking for five months may well rack up higher numbers than a book that’s only been available for two months, and shows more pre-orders at any one time, but for a shorter period.  Then there’s the fact that books have differing readerships, and the audience buying patterns reflect such differences, so that a book that never makes the “bestseller” lists may actually outsell a one-time sensation over the course of a year… or twenty. Add to that the fact that the BookScan numbers don’t reflect sales at some large outlets and at a considerable number of independent bookstores.

There’s no doubt that publishers do try to game the system.  A publisher generally won’t schedule two possible blockbuster books from its own list in the same genre or with the same readership list in the same month or even the same season. They may set up special events just after midnight so that a book comes out in the first minutes of the next reporting week, maximizing possible sales for that one week.  Or they may try to generate huge-pre-orders, as it appears Tor did with the last book of the Wheel of Time.  Sometimes that works; sometimes it doesn’t; and sometimes, even the publisher gets surprised.  No one at Tor, for example, had any idea that Imager’s Battalion would make The New York Times bestseller list, especially since it was released something like two weeks after A Memory of Light [the final WOT book].

Upon occasion, however, individual authors have also gamed the system, like the authors of a business book who bought something like 10,000 copies of their own book in targeted areas across the USA in order to get on the bestseller lists – and did. I don’t have any idea if they made back the likely quarter of a million dollars that cost, but maybe they did.  And so did the author of a book called Leapfrogging, but doing so required getting others [i.e., his business clients] to pre-order some 3,000 books… and that’s not all that different from simply getting 3,000 people to buy your book.  As a very practical matter, major publishers don’t “game” the system this way.  They’ll persuade, talk to book buyers, advertise, create fan groups and buzz, etc., but it’s rather obvious if your purchasing department buys on the market 10,000 copies of a book you just published… not that for high-priority titles they won’t do quite a few other things.

Because Amazon does make weekly BookScan numbers available to authors – just of the sales of their own books – I can compare what BookScan reports to what is actually sold, albeit on a greatly delayed basis, since I get author royalty reports only twice yearly, and those numbers are for a 6 month period ending four months previously. I think it’s fair to say that, while royalty reports may not reflect all books being sold in a period, they’re far more accurate that any bestseller methodology, and they certainly don’t have inflated sales numbers because I can be absolutely certain that my publisher is not going to pay me royalties for books that were not sold.

Based on that comparison, the sales numbers for my books that I get from BookScan through Amazon [sales, not their hourly ranking] only account for roughly 50% of actual sales, suggesting that I have significant sales in smaller retail outlets – and if those outlets are ones surveyed by the Times or USA Today, I’ll rank higher on their lists than on the PW list, which is in fact exactly what has happened on those few occasions when I’ve made the lists [near the bottom, I will confess, but it’s still very nice to be on those lists].

As for reader ratings, they reflect nothing more and nothing less than reader preferences, and have little to do with sales numbers, except that books that sell more copies usually have more reader reviews, but such reviews usually comprise less than half of one percent of the number of copies sold – except in the cases where authors have made a huge campaign to get their readers to write reviews [and in the handful of cases where desperate authors used multiple pseudonyms to pad their reader reviews].

So… the bottom line, so to speak, is that bestseller lists do reflect comparative sales, but only roughly, not inclusively, and only for a defined period of time… and publishers do their best to maximize the appearance of their books on lists… sometimes… but the information is far too scattered and imprecise for the kind of  wholesale “insider corruption” that is periodically suggested, particularly since the bestseller lists compiled by different entities with differing sources and methodologies seem to share a great many titles, if in differently ranked positions.  And while the cost of “gaming” the system by buying your own books can work… it’s far too expensive, even with the “multiplier effect” of being on the list, for most people or publishers to find it cost-effective over the long run.

In short, from what I’ve seen, there’s a fair amount of “gaming” on the part of authors and publishers but not, if you will, list “manipulation” or “corruption” on the part of either publishers or those who publish the various best-seller lists.