This site uses cookies to provide you with a more responsive and personalized service. By using this site you agree to our use of cookies. Please read our cookie notice for more information on the cookies we use and how to delete or block them.

Bookmark Email Print this page

Taking the Pain Out of Warranty Claims Management

OEM drives value throughout the claims process with predictive modeling



By replacing manual, rule-based warranty claim reviews with scientific, automated methods, an automaker significantly improved its claims adjudication process with the ability to preemptively identify potentially false or inappropriate claims.

The Challenge

The current warranty claims adjudication process relied heavily on manual reviews and static rules. The automaker needed a more scientific approach to adjudication that could help it identify potentially fraudulent or false activity in a more efficient, automated manner.

How We Helped

The Deloitte team began by helping the company analyze over 16 million claims and merge data from multiple sources both within the company and from external credit report records. Then, using advanced modeling techniques such as Decision Trees, Association Rules, Logistic Regression and Benford Analysis, Deloitte helped the company develop five sets of new rules capable of identifying potentially false or inappropriate claims. The OEM added the new rules to their existing process in order to improve both the accuracy of its warranty claim reviews and the return on investment (ROI).


The post analysis predictive benefits , based on what the OEM called “very conservative premises,” were more than three times larger than initial expectations. The OEM also received an additional unexpected, highly valuable benefit. Currently, every warranty claim goes through thousands of rules before being paid. For a given repair, the rules can identify inconsistencies between the labor operations performed and replacement parts used. Previously, the company did not have a set of pre-approved labor operations and parts for a given repair according to specific make, model and year, which would drastically reduce the need for reviewing and interpreting every claim. Although the predictive models were not designed to generate such a list, it was a byproduct of the models and has enormous value. According to the OEM, the list would have been extremely difficult to prepare and can revolutionize the way it does warranty business.

Share this page

Email this Send to LinkedIn Send to Facebook Tweet this More sharing options

Stay connected