Don’t drown walking across the river
An aphorism that I picked up years ago that crosses my mind frequently in my professional life is that “a six-foot tall man drowns walking across a river that’s an average of three feet deep.”
The point is that aggregates and averages might mask important truths.
I thought of this when I read the news from the Forrester study of ebook take-up announced earlier this week. NB: Forrester is in partnership with my colleagues at Digital Book World and will present data — although not on this ebook study, but on another project — at the DBW conference in January. I haven’t been involved in those discussions and, like most of the readers of this blog, only know about the ebook study what I read in the releases and commentary.
Kat Meyer on O’Reilly Radar expressed her doubts about the Forrester data, and data just announced by Bain Consultants in France. Kat’s concerns go to methodology. Although I don’t know if we’re in a position to evaluate the methodology because I’m not aware that much has been revealed about it, I’d say her point is well-taken but it isn’t what really concerns me.
(My own first take on the headline numbers is 1) $1 billion in ebook sales now? If this trade only — and there are some indications enumerated below that it might not be — then it is out of $15 billion which seems reasonable. If the number includes non-trade, which is $30 billion, then it is shocking for being low, not high. 2) Forecasting growth to $3 billion by 2015 from either base seems very conservative and the reduction in the growth rate over the next five years over what it has been the last two years is the story. I don’t think I’ve seen any accounts of the report that have characterized it that way. Being alone with this analysis makes me wonder whether I am missing something and don’t know enough to comment yet. That’s why this paragraph is in parentheses. It makes me feel better.)
The Forrester presentation of an industry study is one of several rooted in serious research that we’re planning for the conference. Last year we had reports from Verso Media about book readers, tracking their switch from print to electronic. Guy Gonzalez and his Digital Book World team have taken over that study and will update it for us with Verso. We also had a presentation last year from Bowker and BISG, who were just starting their study of ebook readers. They have done four fieldings since and will also be able to give us an update.
In both these cases, as long as the methodology of the studies has remained consistent, we’ll get important trending information, whether or not the precise percentages reported for various behaviors are accurate or not.
We got an opportunity to do another study when the team at iModerate, which has an online “chat” methodology to personalize research, volunteered to demonstrate what they do for our audience. We got to choose the topic and we decided to study the ereading habits on portable multi-function devices (smartphones and tablets). We chose that topic for two reasons: it is a new and rapidly-growing group of ebook readers and the color touchscreens and connectivity of the devices makes enhanced ebooks that might be hobbled on the Kindle or first-generation Nook fully accessible.
We will also debut work Bowker has done on the children’s book market supported by several publishers and organized with the Association of Booksellers for Children.
The headlines from the Forrester reporting were that ebook sales are approaching $1 billion and they expect that number to triple in five years. Also eye-catching was the fact that, three years into the Kindle era and more than six months after the iPad introduction, more ebooks are read on full-function personal computers than any other way. I say that was eye-catching; it goes to the heart of my concern about the data. It’s about the six-foot tall man.
It is my strong hunch that the content that is read on PCs is qualitatively different than what is read on portable and mobile devices. I am fully aware of the dangers of generalizing from one’s own experience, but I have never met a person who reads trade books on a PC. I know people who read on Kindles, Nooks, smartphones, and iPads. I am aware from having talked to people in the romance ebook business that people in offices reputedly read romances on their office machines (at lunch, of course).
But my intuition tells me that big chunks of that PC reading is professional and informational, not recreational and that this is where PDF sales are most likely. If 30+% of ebook readers consume content on regular computers, I’ll bet the percentages for O’Reilly’s Safari (whether reading a chunk of an ebook from that service is counted here is a good methodology question, but my intuition about interpreting the device data tells me it must be) are much higher.
So ebook reading is the river that’s an average of three feet deep. But it is only a foot or two deep near the shore (where the trade ebooks are read) and it is 15 feet deep in the middle (where the professional ebooks are read.) And the important point is that publishers who do one or the other are not usefully enlightened by data that puts those two distinctly different markets and environments together as if they were one.
This is not to suggest that nothing can be learned from Forrester’s research nor that any other study has a firm grip on this granularity. I asked a data-driven colleague who’s done a bunch of work in this area whether he shared my hunch about who’s reading those PDFs on PCs. He went into his files and ultimately agreed that the market parsing I was looking for was not evident in the extensive research he had done.
Obviously, there are people who know this. Amazon and B&N and Kobo know what devices the books they sell are read on. O’Reilly knows what devices the books they sell and the ones used in their Safari library are read on. When I interviewed the publisher of Ellora’s Cave at Digital Book World last year, she was quite conscious of the fact that many of her books were still sold as PDFs, implying a computer reader. The fact that this data has not been made ubiquitously available and parsed suggests that it is seen as having proprietary value by the people who possess it.
Trying to understand a strand of the market that might be distinct was behind our thinking when we decided to have iModerate focus on portable multi-function devices. We figured that those readers might use and value enriched ebook features more than Kindle or Nook readers and we also see them as the market segment of ebook readers likely to grow fastest. So understanding that market segment in some more detail might help publishers lead the target a bit on product development.
We have written many times before that the book business is not one business. The professional ebooks read on a laptop by a programmer in the middle of an assignment don’t tell you much about what format you should publish a romance novel in. The big change in the ebook world that hasn’t really happened yet but will in the next couple of years is greater adaptation and consumption of illustrated books in digital form. Anything heavily illustrated now pretty much has to be delivered as PDF to a laptop; that won’t be true anymore at the outer edge of the current forecast window, which is 2015.
On the day I’m writing this, new ebook sales data was announced and Cader analyzed it in a post that is behind his paywall. He calculated that ebook sales comprise 9.5% of adult trade sales but only 1.7% of children’s. That’s really charting the river bottom in a useful way.
So we’d say give us data, let us try to understand its limitations and gain insight from it at the same time, and let’s remember that the world of digital change in publishing is simultaneously dynamic and diverse and that no single body of data is likely to give us the answers to what we should do next or what we should expect in the years to come.
It is precisely because data needs to be interpreted that we set the Verso-DBW, BISG-Bowker, and iModerate sessions at Digital Book World back-to-back-to-back and will follow their presentations with a panel discussion meant to shine some light on what we can conclude from what they say.