March 3rd may have been World Book Day, but things are looking dire for the book as we know it. This past weekend in Bristol, somebody tried to give me a title by Margaret Atwood to celebrate the day, but it was more than just a title, it was an entire book and daunted by the prospect of carrying it around all day, I gave it a pass. Even free, a fistful of pages is undesirable in an age where information is weightless. Just as we have cast off the physical shackles of the compact disc, we consumers are apparently rushing to pick up e-readers everywhere and this has analysts quite excited. Today in Figure it Out, we look at how you could use Operational Research to predict e-reader uptake. Systems Thinking ‘Systems thinking’ is an approach for examining how the key resources of a system interact with one another. This can be used to analyse your workforce, the Afghan insurgency, the formation of sand dunes, people in restaurants, or in this case, how the key resource – consumers – transition from paperback to e-book. A tool in systems thinking is the influence diagram. The first task of building an influence diagram is simply to write down all of the key resources and variables of a system, draw arrows to indicate which are directly influenced by which, and finally indicate whether that is a positive or negative influence. While relatively simple, it is that simplicity that makes it so powerful, as we can dissect system complexity and present it for all to see. The adoption of a new technology is a classic application of Operational Research to consumer behaviour and I present an influence diagram of a simple model below:

You can see simple relationships like how people buying e-readers diminishes the number of paper book users, and more complex relationships on how the number of paper book users and e-book users affects the impact of word-of-mouth. That is the only tricky part of this model:  on any given day, an e-book user may interact with a handful of people, but as time goes on and more people are already e-book users, so he will find fewer opportunities to convince new people to buy. Systems Dynamics Influence diagrams and systems thinking are an excellent way to break down a problem into its essential parts, but we can go further using a continuous simulation methodology called systems dynamics. After establishing the influences in our system, we can build mathematical relationships between them and calculate how the system will evolve over time.  I picked up Sphinx SD Tools, a free open source program to see how it could tackle the problem. Building the model was relatively easy and looked something like this:

A full implementation of this approach would involve some market research, greater complexity, and greater segmentation. However, for the purposes of this article, we collected the following in order to seed the model:

  • Given that Amazon’s Kindle Store has a section for “Baby-3” we can safely say that e-readers are available for all books for all ages. Wikipedia says there are an estimated 62,041,708 people in the UK.
  • According to, in the UK about 6.5 million people have e-readers to start 2011. That yields 55,541,708 without.
  • The International Digital Publishing Forum shows exponential American e-book sales growth with Q1-Q3 2010 up 172% over Q1-Q3 2009
  • The Association of American Publishers estimates show that the 2009 sales of e-books were up 177% on 2008 which was up 168% on 2007

Using this, we could pin down the immediate growth expectations, but not the proportion of growth that comes from gadget lovers and the proportion that comes from regular people. Unfortunately this data just isn’t available. If we pick a reasonably middle-ground figure, then we might expect the next 60 months to play out like this:

We see a 5-year transition of the population paper book users to e-books as you would expect. Interestingly, though, we see that the number of people buying their very first e-reader peaks just after the two-year point and then goes into decline. Was this a property of our middle-ground choice of the mix of gadget lovers and regular people? We can do a quick sensitivity analysis to see the possible futures for e-book sales:

Each line in the chart above shows how monthly first-time e-reader sales will evolve over the next 60 months for a given population mix of gadget lovers and regular people. We can see that the lines with the steepest curvature peak between 1.5 and 3 years from now and those are reasonably likely results given the forecasts. Conclusion If you believe the excitement about e-readers and e-books then surely enough the market is booming. However, that sort of growth can’t last for long and the market becomes saturated. Will manufacturers produce incrementally improving, innovative readers to tempt us to buy new versions or will sales crash once everybody has one? Perhaps we will find out that this period of history was a gold rush for retailers to grab on to consumers in order to connect them to their marketplace and services for the future. Systems dynamics would be an excellent tool for analysing the competitive dynamics as e-readers flood the market, but that is a project for another day.