Analysis Objectives and Sample Chart

ACG fraud detection models are designed to review the claims submitted by the manufacturers’ dealers who are responsible for handling warranty repairs and replacements for which they are reimbursed by the manufacturer. Some of these claims have the potential of being fraudulent. These claim types include cases where a customer (in collusion with the dealer or is fictitious) receives hundreds or thousands of warranty replacements or where hundreds of “customers” names lead to the same address.

Our models use various data interrogation techniques such as query and reporting, data association, clustering, etc., to identify anomalies and suspicious activities from the data itself. These models enable the user to adapt to the changing patterns of fraudulent claims submissions as they evolve in real time.

The models are not dependent upon knowing the attributes of the last fraudulent claim to be effective. The models are built to detect suspicious patterns in the claims data itself, their association with other claims data, and dealers as they are compared against their peers. The anomalies and suspicious patterns are presented to an investigator with their reasons for final determination of fraudulent activity.

The following chart illustrates a report generated from a fraudulent claim project. The report details the relative importance among the three variables that were used in detecting auto warranty claim fraud:

Data Analyzed, Models Used and End Results Achieved

Historical warranty claim data are used to identify the indicators and patterns associated with claim fraud. Models are then built based on these indicators and patterns to recognize possible fraud in future claim submissions. The models produce a report detailing the identity of claim as well as reasons for investigation.

Based on our prior experience, the identification of following types of anomalies and suspicious activity are in our pre-built models:
• Phantom customers or bogus parts.
• Customers receiving dozens or even hundreds of warranty replacements.
• Claims submitted from many customers all leading to the same address.
• Claim volumes that don’t correspond to quantities of product actually manufactured or sold in specific geographic areas.
• Double billing on the replaced part.

Benefits Realized and Reference Case Study

• Reducing claims costs by 20% to 30% without denying any legitimate claims.
• Discourage submission of fraudulent claims in the first place.
• Increase revenues through the elimination of stolen parts being sold illegally to manufacturer’s customers.
• Reduce problem claims by encouraging submitters to validate and verify claims before submitting them.
• Detect earlier and increase the focus on true product problems by removing the “bad” claims from the mix.
• Focus field audits more effectively and improve analytical models from what is learned in these audits.

Click on the link entitled Reduce Warranty Claim Expense by Better Fraud Detection found in the Resource Center for the case write-up.