We provide household-level estimates of the expected probability of risk from fire, flood, medical and collision events at full postcode resolution across England. These insights support risk selection and underwriting decisions within the insurance industry.
Our predictive postcode models are based on publicly available data from Fire and Rescue Services and draw insights from over 2M commercial and residential building-related incidents recorded between 2010 and 2025 reported at LSOA level. This enables us to identify relative risk zones across England at a granular full postcode level with 100% coverage.
Although non-catastrophe Fire and Flood claims are relatively infrequent, they often result in high-severity losses. Even the largest insurance providers struggle to detect meaningful trends due to limited data availability.
An expansion of the models to include Scotland, Wales and Northern Ireland is currently underway.
predictive25.com was founded to deliver predictive modelling, data science insights, and advanced analytics tailored to the UK insurance industry. We bring a unique combination of data science expertise and deep insurance sector experience.
We'd love to partner with your company on a proof of concept. Our postcode datasets are available at a fixed cost per model.