This thesis outlines research completed in partnership between Merseyside Fire and Rescue Service and Liverpool John Moores University. The aim of the research was to investigate ways to develop and implement a bespoke Geographic Information System framework that could be used to identify risk of future anomalous accidental dwelling fires. This thesis outlines the techniques used to develop the framework and its application. In particular, the thesis presents an understanding of accidental dwelling fire causal factors and how data related to these can be incorporated into a model for identifying risk and targeting initiatives relative to the risk. The thesis also investigates two strands of customer insight developed for Merseyside Fire and Rescue Service. These are community profiles, based on a cluster analysis approach, to understand risks present within communities and the vulnerable person index, which identifies individuals most at risk from fire using data shared through information sharing agreements. Nationally recognised risk modelling toolkits, such as the Fire Service Emergency Cover toolkit do not utilise local information or have the ability to identify risk to an individual level. There is a need for this intelligence to be able to proactively target services, such as the Home Fire Safety Check. This paper also discusses some of the key operational and strategic areas that benefit from this information and presents some case studies related to the application of the research.