In a blog post by Google software engineer Leonid Taycher, the processes which led to this figure were explained in great detail. Initially, it was necessary to define the concept of a book. The result was based on the idea of a book as represented by an ISBN (International Standard Book Number). However, a few modifications were required as this system only dates back to the 1960s.

Google also found that the definition of a tome was also a beneficial one. This description refers to a text which can exist in a single copy, or has millions of copies. However, this definition also has its drawbacks, insofar as hardback and paperback editions of the same book are counted twice.

The physical process of calculating the data relies on a complex algorithm which collects meta data, before attempting to weed out any duplicates or non-books such as pamphlets, maps, videos or even T-shirts (some of which also have ISBNs). As the algorithm is refined, and Google gains access to additional information, the numbers change invariably.

At the time of the Google blog post, there were 146 million instances of “books” which were “printed and bound”.

In the knowledge that there are almost 130 million books in the world, it urges the question of “how many of those books have I read?”. Or furthermore, how many of those books have never been read at all? If nothing else, this exercise in book counting has reinforced the benefits of the good work of marketeers. If it weren’t for marketing and advertising, many aspiring authors would never reach an audience larger than their extended family.