12.09.2013

Nielsen BookScan Sales Data Can Be Misleading (Or Alternatively Understanding This Brave New World of Publishing a Little Better)

I think most everyone is guilty of preaching. I see lots of opinionated posts that pretty much state, ‘I know everything because I have published a book.” And I know what they mean, because that’s what a lot of people mean when they’re giving their opinion. Sometimes I’m one of those people too. I’d like to think my 20 years of experience as a pro author and 12 years of experience as an indie author get me pretty far in the expert department (but I’m also smart enough to know there are very few actual experts on anything). Publishing has been changing so fast over the past decade that many so-called experts are left scratching their heads as they try to understand what’s going on. One of those experts being left behind is Nielsen BookScan.

For those who don’t know, Nielsen BookScan is a company that collects retail sales information from bookshops. Many companies, including Amazon.com, use Nielsen BookScan sales data to track the sales history of books. Some bestseller lists also use Nielsen BookScan sales data to determine which books are the bestsellers in specific categories at any given time.

Much like bestseller lists, which I’ve written about in Bestseller Lists Can Lie, Nielsen BookScan sales data doesn’t actually show the total sales of any book in any particular category or at any particular time. More accurately, Nielsen BookScan sales data is reflective of a statistical sample of sales of particular books in particular categories or genres at a particular time. More plainly, Nielsen BookScan sales data captures a relatively small portion of book sales in particular markets and locations.

However, Nielsen BookScan sales data is often presented as if it shows a book’s total actual sales or equally as bad some imagined percentage of actual sales. Even Amazon.com which should absolutely know better as a digital online leader misrepresents to authors what Nielsen BookScan sales data actually is. BTW, your Author Central account on Amazon.com shows sales data for Nielsen BookScan US (which is itself one segment of the ten currently tracked markets).

On numerous occasions, I’ve had publishers and others in publishing tell me with a straight face that you can tell the total sales of a book simply by multiplying the Nielsen BookScan sales for a book by some imagined value. Usually, the X factor stated ranges from 1.5 to 2.5 depending on the type of book. Twenty years ago, when books released in the US were primarily sold only in the US in physical, tracked stores that may have been true. Such a thing hasn’t been true for a long, long time in the US or anywhere else.

Here’s what Nielsen BookScan tracks exactly: English Language sales statistics for the consumer book market from about 35,500 tracked retail stores in 10 countries: Australia, Brazil, India, Ireland, Italy, New Zealand, South Africa, Spain, UK, and US. Prior to approximately early 2011, the tracked stores included only specific chain bookshops, key independent bookshops, and key supermarket stores. After approximately early 2011, Nielsen BookScan also began monitoring sales from leading internet book sites.

What doesn’t Nielsen BookScan track? Everything else. Nielsen BookScan doesn’t track education (school, college, etc) or library sales. Nielsen BookScan doesn’t track sales from most department stores, even though many carry books. Nielsen BookScan doesn’t track non-English Language sales. Nielsen BookScan doesn’t track any bookshops that don’t participate in the program or that aren’t considered key or leading. Nielsen BookScan doesn’t track non-book sites that also sell books. Nielsen BookScan doesn’t track internet book sites that aren’t considered key or leading. Nielsen BookScan doesn’t track English Language sales in any country other than the 10 where it currently does business.

To understand Nielsen BookScan, you also need to understand that the data is segmented by market (country). Nielsen BookScan US data is reflective of English Language sales statistics for the consumer book market in the US and began operations in January 2001. Nielsen BookScan Australia data is reflective of English Language sales statistics for the consumer book market in Australia and began operations in December 2000.

Nielsen BookScan New Zealand data is reflective of English Language sales statistics for the consumer book market in New Zealand and has been operating since December 2008. Nielsen BookScan South Africa data is reflective of English Language sales statistics for the consumer book market in South Africa and began operations in December 2003. Nielsen BookScan India data is reflective of English Language sales statistics for the consumer book market in India and has been operating since October 2010. Nielsen BookScan International data is reflective of English Language sales statistics for the consumer book market in the UK and Ireland.

Something else you should know: From its inception to approximately early 2012, Nielsen BookScan was specifically for tracking titles that were physical products, meaning tracking of printed books and packaged audio. In approximately early 2012 Nielsen BookScan began working to track sales statistics for the consumer ebook market. Prior to this Nielsen BookScan did not track ebook sales. Nielsen BookScan also does not track digital audio sales in any market.

Nielsen says that BookScan in the US tracks approximately 16,000 locations or roughly 85% of the market but then states that it collects sales data for only 500,000 titles in a typical week. Visit Amazon.com and you’ll find about 12,000,000 titles. My best estimate is that about 35% of these titles are books actively available in the US, including printed books, digital books, packaged audio, digital audio, etc. 35% of 12 million titles is 4.2 million. 4.2 million - .5 million is 3.7 million. Where exactly is the sales data for these 3,700,000 titles? Well, it’s not tracked.

Within a market, such as the US, Nielsen BookScan also provides historical sales data by category, format, author, and publisher with weekly data aggregated into four and twelve week blocks. Titles are tracked by ISBN, the unique identifier for book products. Nielsen BookScan can track sales across formats. At present, this cross-format tracking is for alternate physical editions of books and the tracking is dependent on each related title being properly cross-linked by the publisher in the first place (generally by specifying a related ISBN).

The clear problem here is that modern books are often released in many formats, in many editions, and in many markets. As an example, in the English language there are over 600 William Robert Stanek titles at Amazon and over 900 William Robert Stanek titles in library distribution. You can check these numbers yourself by counting all William Stanek, William R. Stanek, William Robert Stanek, and Robert Stanek titles.

There are an additional 400 or so of my titles in the English language that don’t even have my name on them. They are sold under a brand name. For example, Bugville Learning and Ruin Mist Publications are two of my brands and titles released in these brands may or may not have my name on them even though I have of course written them.

I have hundreds of inactive and/or otherwise out-of-print titles as well. Additionally, over the past 20 years, my books have been translated into 34 languages. All of these translations were released in many formats and editions as well. In total, there are thousands of my titles published in the past 20 years.

How many of these titles has Nielsen BookScan US ever tracked? Well, Nielsen BookScan US didn’t track any sales prior to January 2001. Nielsen BookScan US didn’t track Internet sales until approximately early 2011. Nielsen BookScan US just started tracking ebooks sales and doesn’t track digital audio sales. Nielsen BookScan US doesn’t track education sales. Nielsen BookScan US doesn’t track library sales. Nielsen BookScan US doesn’t track any sales outside the US. I could go on, but I won’t. I think you get the point.

Finally, don’t confuse Nielsen ratings with Nielsen BookScan US. With Nielsen ratings, the television viewing habits of a relatively small cross-section of households are statistically extrapolated to be representative of millions of viewers across the US. With Nielsen BookScan US, the book buying habits of a subset of leading US retail locations are statistically extrapolated to be representative of the millions of books sold across the US. If you’re an author whose books are sold primarily in these retail locations, you’ll look like a superstar. If you’re an author whose books are sold primarily in other locations, you won’t look like a superstar.

Thanks for reading,

Robert Stanek

No comments:

Post a Comment