Reusable, “standard” data models have long been an aim of the data modelling community, most recently with an emphasis on patterns. However, many standard pattern approaches call for a close modelling of the subject domain itself, and are not easily transferable to other domains. The approach we describe here, situated abstraction modelling (SAM) describes an approach to data modelling that defines a template for the functional “shape” of the solution, which is then instantiated for a particular set of circumstances. In effect, we have a template not for the data model, but for the situation to which it applies. We describe a set of six data model templates that can be instantiated over a wide range of application domains.