Well, a quick history. The Nielsen company enjoyed considerable success throughout the 1990s with Nielsen SoundScan, a service that aggregated and reported POS (point-of-sale, i.e. scanned through the register) data from a wide variety of retailers. For the first time, it was possible for the industry to really see how it was doing; sales from thousands of record labels could be tracked on a weekly basis at dozens of major stores. So, in 2001, Nielsen figured they'd try their hand at books.
There's a .pdf you can download from the BookScan site to tell you exactly which retailers provide POS data to Nielsen, but for your convenience, I've listed them below:
B. Dalton • Barbara's Bookstores • Barnes & Noble • Books-A-Million • Borders • Deseret Book Co • Follett Stores • Hastings • Hudson Group • Independents • Walden Books (owned by Borders)
Discount & Others
Amazon.com • B&N.com • BooksAMillion.com • Borders.com • Buy.com • CatholicCompany.com • CNI • CEORead.com • Cornerstone • Costco • Fatbrain.com • K-Mart • NBC.com • One World Enterprises • Powells.com • Stretch the Skies • Target
Non-Traditional (New for 2008)
H.E.B. • Kroger • Stop & Shop • Shopko • Toys "R" Us • Babies "R" Us • Starbucks
What's as important as the retailers listed here, however, is the retailers that aren't; specifically, Wal-Mart, Sam's Club, and BJ's Wholesale Club. Because they (and several other, much smaller) outlets don't report their sales figures to BookScan, BookScan's numbers aren't complete. As I've said before, on a good day, they probably capture only about 75% of the overall marketplace.
For the most part, BookScan generates an accurate picture of any given title's overall sales. Moreover, it's the only tool the publisher has to analyze sales of books that aren't theirs, so even if it weren't as accurate as it is, it's still all they've got. (Publishing houses do solicit POS data from their accounts, so they have separate, often more accurate sales figures for their own titles.) If Chester A. Author published his first book with House A and is now at House B for his second, House B will use BookScan to analyze sales for his first book. These numbers, combined with comp title information, will factor significantly into the initial buys for his second book. The theory goes that the aggregate POS data will provide an accurate picture of previous sales, thus providing the foundation for reasonable expectations on future sales.
You'll notice I said "for the most part." Since major retailers like Wal-Mart don't report to BookScan, a title that rocked out at said major retailers but failed to perform on-line or in the national chains (it's rare, but it happens) will have artificially depressed sales figures. (If you're the author of this unfortunate book, you'll be depressed as well. Ba-dum chh.) Seriously, though, if your book sold 10,000 copies, but 5,000 were at Wal-Mart, Sam's Club, and BJ's, your BookScan numbers will be off by a whopping 100%.
As agent Andrew Zack notes in his own post about BookScan, there are some publishers who, despite knowing full well that BookScan only provides a partial picture of overall POS transactions, treat it as if it were the be-all and end-all of book sales. Not only that, but in one case, BookScan reported numbers as low as ten percent of a book's actual sales, and editors at the (undisclosed) house had to be convinced of this by the author's actual royalty statements. (Zack doesn't say what said publisher actually decided concerning his client's new book, although he does follow up on his initial post here and here.) And, in case you were wondering, BookScan never changes its numbers—ever—so you can't complain to the Men Upstairs to adjust your reported sales if you think they're wrong.
Which brings us to the question: What can you do if you think you're getting B.S.ed by B.S.?
• First, do your homework. As noted in Alan Rinzler's excellent primer on BookScan, you can subscribe to Nielsen's service for $5,000 per year (pricey!) or just $85 per year if you only want to track a single ISBN. Know thy sales. Like Alan says in his post, an author's BookScan numbers are like a credit rating, and are just as (if not more) important.
• If you think there's a problem, approach your agent. Have him or her go over your royalty statements with you. This can help you figure out whether there's a major discrepancy between the publisher's numbers (which will be more accurate) and BookScan's.
• If you and your agent believe there's a serious problem, contact your publisher. Let them know you're afraid artificially low sales reports from BookScan may impact sales on your future books. If possible, ask them to review their sales numbers and let you know if there's a problem. (The publishing house I work with maintains huge Excel files of market share data, which is beginning to include data from previously reticent retailers like Wal-Mart.)
Of course, this all only applies to non-debut authors; for those of you whose next book will be your first, your numbers may depend in large part on the BookScan numbers for your comp title, which you can't really control. If that's the case, just keep on keepin' on—promote yourself, promote your book, and do whatever you can to get your name and your story out there.