Q: How accurate are bestseller lists?

A: Not very.

But that doesn't mean they aren't important. They are great for publicity and will probably help generate more sales. Many people peruse the various lists to help them determine what to pick up next. They are fabulous for an author's ego. Admit it, wouldn't you like to have the tag New York Times Bestselling Author under your name every time you published a book? All it takes is once!

Why aren't they accurate?

Book publishers don't use a upc code on the back of their books. Why? There is an ancient custom that retailers should be able to set their own prices. Instead most publishers use an ISBN number and code. Some use nothing at all. That means a whole new reporting system is needed to gather point-of-purchase data. The biggest collector of this data is Nielson's BookScan system - modelled after music industry's SoundScan.

But not all retailers feed their data to BookScan and not all bestseller lists use BookScan anyway. The New York Times has the most prestigious list, which is based on several large chains, a number of independent booksellers, and select mass market accounts. USA Today and the Wall Street Journal employ similar methods of sampling, including judicious use of BookScan. Ditto Publisher's Weekly. However, many large booksellers -like Sams and a number of other mass market retail chains, schools, Christian bookstores, rackjobbers, e-books, and high volume tabletop display marketers - don't provide their data to BookScan. The largest Christian retail chain doesn't even provide its data to the CBA (Christian Booksellers Association) bestseller list. One wouldn't, of course, expect the lists to account for other 'special markets' including direct sales nor organizational and author purchases. The good news is that Amazon sales are now included with BookScan - but several of the lists resisted using Amazon's sales until the last few years.

What percentage of book sales are reflected on bestseller lists? No one knows for sure based on all the above reasons. I've heard estimates ranging from 30% to 50%. Anecdotally, one author friend has now sold three million copies of a single book. Of those, 100 thousand have sold in traditional book selling settings and the other 2.9 million have sold direct to consumer, business to business, or through back-of-room sales when he speaks.

Because bestseller lists do create positive publicity and sales momentum there are more than a few occasions when authors and publishers have attempted to manipulate their book's placement on the lists. For example, back when it was harder to track single store sales, an author or agent might order the five or ten or twenty or thirty thousand copies of a new book needed for speaking engagements through a single bookstore to 'force' a book onto the list. I'm sure this has helped ongoing sales just for the fact that accounts would see the book show up on a list and order more store copies. But point-of-purchase data, at least within a chain, is now sophisticated enough to spot this as an anomaly, not a trend. The New York Times at least used to put a dagger symbol next to books that had large bulk orders. (Do they still do that?)

There is a publisher axiom that says you can get a book on any bestsellers list through marketing but in order for it to stay on the list it has to be a great book. Longevity of a book on various bestseller lists is almost always an indicator that the book has real 'legs'. Or, in the case of books that sell hundreds of thousands or even millions of units and never show up on a list, they either need to be great or the author needs to have a great platform for moving product.

So bestseller lists are important indicators of what's happening in major swatches of the book selling environment but they have information gaps in that environment and don't even attempt to measure what's happening in special markets, so they can't tell the whole story of which books sell most.