Prior vehicle damage is known to be correlated with increased future risk, and claims data capture the majority of subject- and vehicle-related incidents. However, for those events where a claim is not filed, other data sources (e.g., vehicle history and police reports) should be consulted for additional indications of existing damage. With that increased visibility to other events comes challenges, though. One of the challenges is identifying which records across sources are incremental and which are duplicate. This requires advanced matching techniques, but it is critical for avoiding double-counting of events. Another is detecting and accounting for change of ownership, which is essential for accurately calculating predictive vehicle history variables as well as attributing the data returned to the current or prior owners. A third is compiling and interpreting data from sources like police reports. The lack of a consistent format makes automated parsing of report data from different agencies quite difficult. Therefore, having a reliable, standardized source for these data is essential. Challenges notwithstanding, the more complete picture of damage history provided by multiple sources offers a clear benefit for accurate risk assessment.
Learning Objectives:
Understand ways to find potential lift from new sources of accident and damage data and find additional segmentation.
Attribute comprehensive vehicle and accident data (Vehicle History, Police Reports, CLUE Auto) to current and prior owners.
Leverage such a solution in your underwriting workflow with an implementation playbook.
