This study pursues the following question: how does the transition of work from in-house computing to cloud computing happens? While seeking for answers, an empirical study was carried out over one year. Its research strategy and method is grounded theory, combined with ethnography as a means for gathering empirical material. The phenomenon analyzed was named going cloud, an expression that refers to the process of shifting the work from local computing to cloud computing. The phenomenon is described, explicated and exemplified from three categories, each one regarding one temporal stage: justifying (temporal stage 1: the implementation and communication of cloud computing results in rethinking how things get done); re-mediating (temporal stage 2: rethinking how things get done results in limiting what can be done in new ways); circumscribing (temporal stage 3: limiting what can be done in new ways results in routinely doing only the managerial work in new ways). The core find is that, at the end of the transition of work from in-house computing to cloud computing, only the managerial work becomes reorganized. The study offers two original contributions to the emerging literature on cloud computing: I) it introduces ‘going cloud’ as a new substantive domain; II) offers a grounded theory, which suggests how and why only the managerial work is reorganized while the organization goes cloud. The study also offers two original contributions to the well-established literature on technology and organizing: III) it advances the theoretical model of the imbrication of human and material agencies; IV) it advances the discussion about the adoption of agential or critical realism while studying sociomateriality. The conclusions indicate how to advance knowledge about the transition from local computing to cloud computing: I) studying how this transition can result in technical attributes performing tasks that do not require expert knowledge; II) studying how organizations and people are developing criteria for deciding what kind of information may or may not go to the cloud. The main implication is to develop longitudinal studies, in order to make comparisons between the work before and after the routine use of cloud computing.