Contextual spatial models for indoor and outdoor structures are created from a plurality of devices using crowdsourcing. A method for creating a contextual spatial model for a premises includes receiving location information and contextual information from a plurality of devices. For each of the plurality of devices, the method receives location information of the device's location, and contextual information related to at least one of the device and a user associated with the device when the device is at the location. The method then determines a designation for each subject location within the premises by applying the received contextual information associated with the subject location to a location model (e.g., rules-based model or a machine learning model).