Recent technology breakthroughs have enabled data collection of unprecedented scale, rate, variety and complexity that has led to an explosion in data management requirements. Existing theories and techniques are not adequate to fulfil these requirements. We endeavour to rethink the way data management research is being conducted and we propose to work towards modular data management that will allow for unification of the expression of data management problems and systematization of their solution. The core of such an approach is the novel notion of a datom, i.e. a data management atom, which encapsulates generic data management provision. The datom is the foundation for comparison, customization and re-usage of data management problems and solutions. The proposed approach can signal a revolution in data management research and a long anticipated evolution in data management engineering.