In November, the New York Times announced that it would begin publishing e-book bestseller lists (fiction and nonfiction) in early 2011. “We wanted to be able to tell our readers which titles were selling and how they fit together with print sales,” said Janet Elder, NYT Editor of News Surveys and Election Analysis. The tracking system the paper developed was two years in the making. The Times is working on the lists with RoyaltyShare, and PT spoke to CEO Bob Kohn about the process.
Kohn said RoyaltyShare is helping the Times to validate the data it receives from e-book vendors by comparing the data against sales data those same vendors send to book publishers in the course of their regular reporting. “We have the cooperation and permission of the book publishers to do this, and the Times can trust that our validation is accurate, because we receive the publishers’ sales data directly from the e-book vendors and can assure the Times that the publishers cannot modify the data,” Kohn says. “It’s a win-win for book publishers, because it’s a hands-free way for them to make sure that e-books that would otherwise qualify as bestsellers don’t fail to make the list because of some data error that’s not in their control.”
Third-party validation of the data was important to the Times early on, Kohn says. He says the challenges of working with the Times are similar to those of aggregating and processing publisher clients’ e-book transactions: “The increasing volume and complexity of the transactions raise the risks of data errors caused by inaccuracies in the transaction data feeds, metadata errors, scalability challenges, incompatible reporting formats, data corruption, transmission error, and human error. . . . As book publishers start working with new e-book services worldwide and start adopting new business models, such as subscription and advertising based services, they will have no choice but to adopt an automated system to handle the transactions.” (The process behind compiling the Times printed book bestseller list, meanwhile, remains somewhat murky. This list is also created by the news surveys department, and is completely computerized. In a 2007 Public Editor column, Clark Hoyt admitted that the list “is not a completely accurate barometer of what the reading public is buying, and it has generated controversy from time to time, most recently . . . when [Elie Wiesel’s] Night . . . was summarily dropped because the editor of the best-seller list decided the book was an ‘evergreen’ that the Times would no longer track.” In the same column, Hoyt wrote that “some companies dump all of their book sales to the Times, while others fill out an online form based on the previous week’s best sellers and including space for unlisted books that have sold well.” It is unclear whether the e-book bestseller list will leave out “evergreen” titles, and we don’t know whether the printed bestseller list also has a third-party company like RoyaltyShare validating its data.)
USA Today added e-book bestsellers to its list in July 2009. The USA Today list differs from the Times list in that it is a single list including hardcover and paperback, e-books, fiction and nonfiction, and all genres in one list. (It also does not exclude “evergreen” titles.”) “The aim of the list is to tell our readers what the ranking of titles was in a particular week,” says Anthony DeBarros, USA Today’s Senior Database Editor. If a title is released in hardcover and is published six months later in paperback and e-formats, “we take the sales of all those formats and put them together so that a book’s rank is determined by a combination of sales for e-books and whatever print format it’s selling in. For some titles, we track several different ISBNs and put those all together.” The list does note which format was the bestselling for that week.
Although Amazon does not release its sales data publicly, the company does provide the data to USA Today, under a nondisclosure agreement. USA Today receives weekly sales data for printed books and Kindle editions. The paper also receives e-book sales figures from Barnes & Noble’s Nook and from the Sony Reader. The reaction to adding e-books to the list, DeBarros says, has been “overwhelmingly positive.” And in October, USA Today created a book list API, including (published) book list data, rankings, and archives, and opened it up to developers to create applications with the data. The open API is limited and data is provided on a one-week delay. On his blog, DeBarros mentioned “potential commercial uses” of the data as one reason access is currently limited.
E-book sales data is also becoming a revenue stream for LibreDigital. The company is partnering with media-tracking provider MediaMorph to create LibreDigital Business Intelligence, a tool allowing publishers to track and analyze their e-book sales and test out various pricing and business models. The companies will analyze data provided directly from companies like Amazon and Apple. The company does not plan to publish any e-book bestseller lists; rather, it will use e-book bestseller data to help determine e-book pricing.
For now, with e-book retailers like Amazon keeping sales data private, it’s difficult to find out how many copies an e-book has sold without a royalty statement. There is currently no E-BookScan, though Jonathan Nowell, President of Nielsen Book Group, recently said an e-book chart will be launched in “a matter of months rather than years.”
Currently, though, says Philip Stone, Charts Editor at The Bookseller, “a couple of major players in the UK e-books market [still aren’t] willing to share their sales data with BookScan, or anyone else for that matter.” The Bookseller was only getting “a handful of download stats on a handful of books by a handful of publishers.” But after looking at the data available, they settled on a way to provide some information while working around tight-lipped vendors. The publication decided to begin publishing its own weekly e-book bestseller list that “without doubt is [currently] the best thing out there.”
At the same time each week, Stone’s team pulls chart positions and selling prices from the online bestseller lists of the UK’s major e-book retailers. Each title is given a points score, based on estimated retailer market share and sales ratios between each book in that week’s printed book bestseller list (to take into account the fact that the #1 book in the chart doesn’t usually outsell the #2 by a ratio of 2-to-1). The information is then used to create an overall bestseller list. “What results is not as robust as we would like, but the proof that we’re doing something right is the fact that what we end up with on our e-books chart correlates nicely with our printed-book bestseller lists in [The Bookseller],” says Stone. “This is something we’d expect—that popular printed books are popular in e-format too.” He looks forward to publishing “a proper, bona fide e-book bestseller list as soon as possible,” hopefully in 2011.
Publishing Trends attempted to use a variation on The Bookseller’s method to compile a rudimentary e-book bestseller list. The bestseller lists we looked at (from Kindle, Nook, Apple’s iBookstore, Sony Reader, and Kobo) were quite varied, and estimating market share was difficult. For now, though, we’re proud to say we beat the New York Times in publishing an e-book bestseller list. Click to see it here.