Big Data doesn’t “speak for themselves”. The role of models in the dissemination of analytics for management accounting
Big Data and analytics have the potential to significantly influence decision-making processes and all management accounting activities, from reporting to strategic planning and control. However, the diffusion of Big Data applications in companies still remains low, for a variety of reasons, relating to people, technology and cultural factors, which the literature has documented.
This article intends to propose a reflection on the different “philosophical” views that managers assume, more or less consciously, on Big Data. These views affect expectations on the way Big Data should generate knowledge and their effective use in practice. Those who adopt the representational view of data are led to believe that Big Data can help generate knowledge quickly and automatically, only by using advanced statistical-computational techniques. Conversely, according to the relational view, the central role in the knowledge generation process is played by models, understood as ways of organizing data, which act as a “bridge” between Big Data and their applications to decision-making processes. The article compares these two opposing views and highlights, through theoretical and practical considerations, how relying excessively on the automatic analysis of Big Data can fuel unrealistic expectations, which are then disregarded. The emphasis is placed on the function of the models, built on the decision-maker’s knowledge of the application domain, which determine what counts as “data” and guide the technical choices in the elaboration phase and in the interpretation of results, favoring an effective use of Big Data for the improvement of business processes.
Keywords: Big Data, Analytics, Models, Management accounting
