Are rising used book sales hurting new book sales – and if so – by how much?
Lots of people would like the answer to that question, from publishers and authors to paper manufacturers and new and used booksellers.
It’s fairly easy to answer the first part of the question: anecdotally, and intuitively, there’s probably no question that rising used book sales are eating into new books sales. If you have any doubt, look up theDa Vinci Code online. A recent search brought up 140 used copies on Amazon, 729 copies on Abebooks and 377 on eBay.
It’s when you try to answer the “by how much?” part of the question that the picture get muddier.
For starters, there are a host of very basic data collection issues unique to the used book market that affect any statistical analysis of total used book sales and the number of used books sold.
What used books are being included in the sales figure? Remainders? Antiquarian (a.k.a. old, rare, or collectible) books? Textbooks?
Sales from what selling venues are being counted? Online? Store sales? Catalogs? Library sales?
Sales to what customers are being counted? Consumers only? Dealer to dealer sales? Sales to libraries?
What is the source of the sales data? Consumer buying patterns or dealer sales? If dealer sales, what dealers?
Then there’s the overriding question that gets to the heart of the issue: Of all the used books sold in a given year, what percentage were in-print versus out-of-print titles? Clearly, the “new” book market is not concerned with sales of out-of-print titles.
It’s this last question that is the most difficult to answer – if, indeed, it can be answered.
Theoretically, a computer program could be developed that could easily match up ISBNs (International Standard Book Number) in seller sales records with the Bowker database that shows whether a title is in-print or out-of-print. Obviously, any book without an ISBN is out-of-print, although a later edition with an ISBN could be in-print.
The one caveat to this methodology, though, would be the assumption that most publishers, large and small, have, over the years, been diligent in notifying Bowker when their titles go out-of-print. Alas, there’s no quick and easy way to test this assumption for millions of titles.
Notwithstanding the above caveat, the ISBN match methodology does has one serious drawback: As only a portion of used book sales are computerized, any matching that could be done would, from the start, be incomplete. How incomplete? That’s hard to know.
According to the recent Book Hunter Press study of the U.S. used book market (highlights of the report, A Portrait of the U.S. Used Book Market, are available online at www.bookhunterpress.com), in 2003, 54% of used book sales were made online. For these sales, ISBNs, if they existed, would most likely be available and the in versus out-of-print match could be made.
But what about the remaining 46% of used book sales that were made offline?
Many of these sales were most likely recorded on individual sales slips, only some of which were then entered into a computerized accounting or inventory program which may or may not have included an ISBN in a separate searchable data field.
This means that an in versus out-of-print match using dealer sales records may only be possible for just over half of all used book sales.
Where does that leave us? Are online and offline inventories similar enough so that the online ISBN match statistics for 54% of the sales match can reliably be extrapolated to the 46% of sales that were made offline?
As an aside, the Book Hunter Press study also found that only 49% of total dealer inventory was listed online and that only 18% of dealers had 100% of their inventories online compared with 46% who had less than one third online. These numbers raise the interesting question of what portion of the remaining 51% of dealer inventory is computerized, the first step in any possible ISBN match program.
What about using sales figures from Amazon or Abebooks? Assuming that the companies wanted to cooperate with such a study (some would consider this highly unlikely), would an ISBN comparison be easier or more fruitful if these larger databases were used? Yes and no.
While Abebooks could generate statistics on the number of books with and without ISBNs, when it comes to sales statistics Abebooks only knows about sales made through its own shopping cart. The company has no way of knowing how many sales of what books were made directly by the dealer.
Would Amazon sales data be more helpful? Again, the answer is yes and no.
Yes, because, unlike Abebooks, all used book sales originating from the site must go through Amazon’s shopping cart. But, no, because not all books sold as used on Amazon are, in reality, used. Look at some of the Condition descriptions for books being sold in the Used and Collectible categories: brand new, direct from distributor or publisher, still shrink wrapped, etc. Are these used books? If they’re remainders, then that gets back to one of the earlier question of whether remainders are or should be considered used books when tallying up used book sales. And, if the books are “brand new,” then where are they coming from if not from the publisher?
eBay isn’t much help either as its listings do not include ISBNs in a searchable field.
The bottom line: The next time you read a report or see a press release that says used book sales have caused a ##% decline in new book sales, stop for a minute and ask yourself where the numbers are coming from and what they represent.
As discussed in greater detail in our study report, the used book market does not lend itself to easy statistical analysis.
Book Hunter Press publishes the Used Book Lover’s Guides, a series of seven regional guides to used book dealers throughout the United States and Canada. The company has been tracking the used book market since 1992. In 1999, the company published it’s first comprehensive report on the U.S. used book market, The Quiet Revolution: The Expansion of the Used Book Market. In 2003, the company published A Portrait of the U.S. Used Book Market. The report, based on a survey of 827 used book dealers, looks at the used book market in terms of dealer size and inventory as well as changes in selling venues and sales between 2001-2003. The report also ranks the major online selling sites in terms of dealer usage and sales generation. More information is available at www.bookhunterpress.com