Cost effective reduction of electricity demand in residential sector is a significant problem worldwide. Feedback intervention is a hot area that possesses considerable potential for achieving electricity saving. However, how to make feedback intervention more effective deserves to be properly explored. In the smart grid case study described in this paper, 3666 greater Sydney region households are sampled. Among these sampled households 2814 residences were equipped with 3 different types of feedback technologies. The sampled households provide a suit of datasets including individual residence’s electricity demand, matched residence’s survey. Combining these with feedback interventions, this paper presents an exploratory study and demonstrates a method of analysing effectiveness of electricity reduction interventions on various household types. Result of the study shows some interesting phenomenon, such as high income households are much more sensitive to feedback technology than low income households; household living in units are statistically not affected by such intervention.