More on the Publishing Industry

In addition to recently writing about the market for short fiction (which I have done here, here, here, and here), I have also been writing about the market for longer fiction and the publishing industry as a whole (here and here). Much of the debate in the latter has been about low incomes and sales for literary fiction, as well as about the relative role of eBook versus print and traditional versus independent publishers.

My summary of this debate was that, although the data are limited and hard to interpret, I thought that the fiction market was in decent shape driven by the expansion of independent publishers, but also supported by some strength (if not strong growth) in the sales of traditional publishers. I figure that the experiments that traditional publishers have been running with the pricing of eBooks probably explain some of their recent declining sales results and especially the decline in their share of eBooks. Much the same conclusion was reached by Mathew Ingram at Fortune in No, e-book sales are not falling, despite what publishers say.

These opinions have only been strengthened by the recent research that I have read. Or more accurately, by the older research that I have only recently read about the industry and the quality of its data.

Jim Milliot's post for the Independent Book Publishers Association Keeping Count: What Industry Statistics Do and Don’t Reveal does a good job reviewing the information available on the size of the publishing industry as a whole. In addition to the problems with the AAP numbers (data only from its 1200 members) and Nielsen BookScan (doesn't include eBooks), Milliot also reviews data from consumer surveys such as the renamed Nielsen Books and Consumers surveys of 60,000 book buyers. In summary: the data we have is limited and hard to interpret.

There are also good reasons to be skeptical over the quality of BookScan numbers beyond the fact that they do not include eBooks. On the accuracy of Nielsen BookScan data, especially for SFF titles, one useful analysis was provided by Suw Charman-Anderson Can Nielsen BookScan Stay Relevant In The Digital Age? in Forbes. The idea behind BookScan is that it requests data from retailers which it then aggregates and sells. It is limited to the extent that the retailers it requests data from are representative of the market as a whole.

This coverage has not always been great. For example, it was not until January 2013 that BookScan started capturing sales from Walmart stores. According to Charma-Anderson's article, it is believed that BookScan captures 80% of book sales in the US and 95% of UK sales, including all the major stores.

Nonetheless, there are some gaps in coverage. As Nielsen BookScan themselves note “library, professional, corporate, premium, export, and some specialty retail sales are not included in the BookScan physical panel.” This is a significant problem for SFF titles as the category "specialty retail sales" is likely to include your local SFF bookstore and almost all comic book stores.

As for information from the publishers themselves, as Dorie Clark points out in Harper Lee and Dr. Seuss Won’t Save Publishing for Harvard Business Review

If I were to self-publish on Amazon, I’d see thorough, up-to-the-minute sales data about how my book was performing. Publishing through a traditional house? Most of us get weekly Nielsen BookScan reports—courtesy of Amazon—and sales figures every six months from our publisher. It’s an 18th century level of opacity that seems shockingly out of date for authors trying to make smart marketing decisions about how and where to promote their books. How can you even know, if you get zero real-time feedback? (A hat-tip here to Penguin Random House, Portfolio’s parent company, which recently launched a comprehensive Author Portal that tracks weekly sales, putting them light years ahead of the competition when it comes to analytics.)

Although this may simply be the publishers strategy to keep their own authors in the dark, I suspect it reflects the fact that the publishers own data is not very good.

That is not to say that the story about flat eBook sales is necessarily entirely false, either. As pointed out on the Stratechery blog in Are eBooks Declining or Just the Publishers?, the Author Earning Reports snapshot shows flat revenues in total. This is inconsistent with what Amazon is reporting as a whole and may reflect the fact that randomness inherent in taking a single snapshot of the market.

In summary, the data are poor and it is hard to know exactly what is going on. More data please!