Space systems have to deal with massive spatio-temporal Earth and Space observation data collected by space-borne and ground-based sensors. Despite the data latency in communications, data is collected at enormous rates, and a sophisticated network of ground stations is set up to collect and archive telemetry data. The data that is received at the ground segment can be made available to the end-users. Beyond archiving data, the available data provides opportunities for data analytics that can support the decision-making process or provide new insight for the target requirements. Unfortunately, for practitioners, it is not easy to identify the potential and challenges for data analytics in the space domain. In this paper, we reflect on and synthesize the findings of existing literature and provide an integrated overview for setting up and applying data analytics in the space systems context. To this end, we first present the process as adopted in space systems, and describe the data science and machine learning processes. Finally, we identify the key questions that can be mapped to data analytics problems.