This book questions the weight and use of data in the ways of governing. Data, online or offline, is becoming a key resource in governance and as such represents a strong political stake. Work on data has always existed, but what is changing is the massification of this data - transcribed by the term big data - made possible by digital technology. Especially since we leave many traces online without necessarily realizing it. Passively, our data is recorded. Data is no longer only used to quantify society and observe it, as was the case with statistics or surveys, but also to drive it. The algorithms make the data speak and would then make it possible to "predict" behaviors in order to better govern them. But these are not neutral devices and it is therefore a question of studying them in context. This is the contribution of this book which explores different fields of activity to understand how the algorithm is set up by its designers and used by actors, caught up in various professional ecosystems and usual ways of doing things that resist the change in algorithmic "governmentability". This approach makes it possible to “defetish” the algorithm and to shed new light on its functioning with regard to the beliefs that surround it, its concrete uses and the power struggles outside the tool itself, but having effects on it. its applications. The strength of the social sciences is indeed their critical reach, in the sense that empirical analysis makes it possible to deconstruct immediate knowledge, prenotions, fantasies. And on the big data side, myths and beliefs, reinforced by those who sell them, are numerous.