The author tries in a book of over 200 pages to explain the ins and outs of a concept that everyone speaks today but whose contours remain unclear. To explain the concept of "Big Data," Davenport casts a wide net by addressing the exploitation of conventional databases or other more modern collected through Web 2.0 for example. The processing of these data takes place in various fields such as management, marketing and finance. Concrete case studies of large American firms mostly specialized in various fields, seek to show how we can optimize the use of databases. The idea is good but this is very general and is not intended to explain the phenomenon of "Big Data" in itself. The author nonetheless defends the concept of Analytic 3.0 which allows to consider a smooth transition from traditional database schemas and those newer radically different in terms of volume and structure. This book also highlights the human side, primordial according to the author, for example, by drawing people portraits in such key data management chain "hacker", "scientist", "trusted advisor", "Quantitative Analyst "or" trade expert. " In short, this book has the advantage of showing that data collection is not an end in itself and that there is a multitude of ways to use them intelligently and especially lucrative. The goal is not to store information in shambles for fun but to know what one can do really. However, its too generalist approach certainly disappoint those wanting highly technical answers to deploy infrastructure to collect and manage large volumes of data, or on the ethical and legal limits encountered by the use of the information collected, a theme totally neglected here. You will understand, some questions remain unanswered after reading this book, but it still offers some interesting food for thought.