There is a growing demand for more and better real-time information to make evidence-based decisions. This compels us to reflect on how official statistics are produced and, more importantly, on the range of information sources available in a world generating an immense and growing volume of data enabled by new technologies. These new data sets go beyond traditional data ecosystems and call for innovation in processes, mechanisms, and tools to generate knowledge from a chaotic ocean of unstructured data, which in many cases exceeds the capabilities of current official statistical information systems.