Earlier this month, reports surfaced of an Amazon claim that a quarter of the top 100 best-selling titles on Kindle were self-published. Famously sparse in its public comments, the company said little more than that, but the press coverage was more than enough to set tongues wagging.
As it happens, the announcement was not really news, as anyone who follows Porter Anderson’s work can attest. In October Anderson, likely the most prolific and among the most probing people writing about publishing these days, moderated a self-publishing panel for CONTEC, a one-day event planned by Kat Meyer for the Frankfurt Book Fair. The growth of self-publishing was confirmed and reported then, even if The Guardian wasn’t watching at the time.
Throughout the fall, Anderson continued to explore the trends and issues related to the growth of self-publishing, aided in part by observations from once-independent author Hugh Howey. Though he is now signed up with one of the Big Five trade publishers, Howey maintains an independent’s sensibility, something very much on display in his comments about a survey that attempted to estimate how much self-published titles make.
Anderson’s written work, coupled with online conversations he has hosted, point to a potentially significant data gap in the growing world of self-published titles. Although the number of independent titles published with ISBNs has grown to 391,000 in 2012, a significant and unknown number are released without the traditional supply-chain tracking identifier. The largest online bookseller does not share its data with the industry, making the Bookstats database a projection as much as it is a compilation of industry sales.
Viewed in a larger context, this challenge isn’t surprising or even new. For the better part of a decade Nielsen, a company that tracks both what people watch and what people buy, was denied access to data for products sold by Walmart, the largest retailer in the United States. Like Amazon, Walmart saw the data as a source of competitive advantage, and Nielsen had to profile the market in a variety of ways that estimated the total even without Walmart’s participation.
Nielsen made that work by investing in consumer panels, analysis of manufacturing data and development and refinement of models that took into account purchasing trends. Consumption is dynamic, and Nielsen found a way to stay current, so much so that the models wound up being pretty close to the actual results that Walmart gave Nielsen when the retailer agreed in 2011 to once again share data.
Because Nielsen was able to show Walmart that its models gave everyone else access to estimates that were pretty close to what was actually happening, Walmart could see that its sales data was no longer a source of competitive advantage. Through Nielsen, manufacturers knew as much as much about the market as they might have gained with Walmart. Over time, access to data reduced the retailer’s power over its customers, as even Walmart represented only a portion of the overall market.
Nielsen was able to use data to model the marketplace because most products were tracked using standard identifiers. Even as the markets shifted to offer and sell new and different products, a known share of the transactions could be monitored and modeled.
There’s a significant lesson for publishing, whose primary identifier, the ISBN, has been shunned by a share of the self-published community. Some companies that advise independent authors have argued that proprietary identifiers like Amazon’s ASIN are just as good and cost the author nothing (a set of ten ISBNs currently sells for $250; lower per-ISBN prices are available at higher volumes).
Decisions like this weaken the data on what is actually happening in publishing. This is an industry concern, but it should also be something independent authors care about. Data reserved for use by one company only strengthens that company’s hold on the market as a whole.
Building on research conducted for BISG in 2012, I noted ten things the publishing industry could do to improve the effectiveness of metadata preparation and use. The last recommendation addressed the independent publishing community directly:
Engage new supply‐chain entrants and encourage them to understand and support the use of ONIX [the metadata standard that includes identifiers].
I’m not writing to blindly defend ISBNs or Bowker’s historical decisions in setting prices for them. I’m trying to make an industry argument that independent authors should consider carefully.
The Walmart example applies. For a decade, the retailer told its suppliers that the only data they would get about their sales would come from Walmart. Without access to Nielsen data, tens of thousands of manufacturers would have had to accept Walmart’s view of the world. By building a data alternative, Nielsen gave suppliers options to understand what was going on in the marketplace.
According to the Bookstats database, in 2012 48.2% of all retail book sales took place online. That same year, 21.1% of units sold were digital products; almost 90% of that was eBooks. This year, more than half of all retail book sales will likely take place online, the bulk of them on proprietary platforms whose management has shown little interest in sharing data.
Standard identifiers, coupled with effective modeling, can help reduce the power of these platforms. The more opaque the data, the more likely authors will find themselves at the mercy of a platform that serves its own interests. I started this week writing about “the danger and significance" of platforms that control access to markets. I’ll end with it, too.