Social Data, i.e. data produced and/or processed by people, is more and more the fulcrum of monitoring and decision making processes. The technological progresses achieved with mobile infrastructures, Web-scale systems, and ubiquitous sensing devices, fostered a cultural shift toward the pervasive use of Social Data as key source of knowledge about the world and its citizens. In this talk, I provide an overview of the research activities performed in the context of the Delft SociAl Data Science Lab at TuDelft. I discuss the main challenges characterising social data (e.g. scale, speed, sustainability, and semantics) in three diverse contexts: Knowledge Creation and Acceleration, using StackOverflow as use-case; Intelligent Cities, and the Social Glass project; and in the “Inclusive Enterprise”, a behind-the-firewall initiative with IBM Benelux. While heterogeneous, these use cases demonstrate the need for approaches able to cater for the intrinsic diversity of the people involved with the data production and processing.