A database designed for the Insurance Council of New Zealand to combat fraud, is to be offered to other industry groups.
The Insurance Claims Register (ICR) was originally developed 12 years ago, when five companies - Vero, Farmers Mutual Group, IAG, Lumley and Tower — committed money to set up the then mainframe-based database.
Two years ago, however, the members felt that they wanted to add some functions to the register that their provider was unable to do. After an RFP, the group chose Optimation to work on the project, which was completed last year.
The database alerts members to the possibility of fraudulent incidents like customers filing the same claims with different insurers. Or, it could indicate trends, like a spate of claims for lost iPhones on the eve of the release of a new model, or a big spike in insurance claims in a sector or area.
“It is a fraud indicator,” says Richard Godman, director of the council. “It is a system that will actually prompt the user to ask questions.”
“The public must know the insurance companies have this, absolutely,” says Godman who is also a manager at Vero.
“Any major insurance events will usually have some kind of influx of fraud activity, so whatever system we have in place must be robust enough to assist in those times as well as business as usual.”
The biggest change was in the shift from a green screen technology to a web-based Word tool. “When you have got someone who comes into the door and have never used our computer, but they know Word and Windows, they can go straight to it,” says Godman.
Godman says the past year has been spent fine-tuning and enhancing the usability of the system for its members. “We have given them ideas what fraud indicators are in insurance. Those fraud indicators will be built into a tool that lets them mine the data.”
The database contains 75 percent of personal insurance claims data and Godman expects this to reach 95 percent.
The system can now be licensed to other industry groups.
An example would be a group of telephone companies who will “clone the system as an empty database” and put the names and addresses of people who have their phones regularly cut off for failure to pay. When these people open up another account, the database can indicate that they have been repeat recidivist offenders.
“It cuts them off before the non-payment or the trouble starts,” says Godman. The potential revenue source from licensing of the database is just one of the “flow on effects” of the upgrade.
Godman says fraud is anecdotally estimated to equal 10 percent of premiums every year. Moreover, this cost increases everyone’s insurance premiums.
The project’s adherence to the Privacy Act is critical, as claimants have the right to access the information relating to their claim history and comments, and can have the information corrected if required.
The data is kept by the insurance companies and the information sent to Optimation is “purely basic”, mainly about the nature of the claim, the claimants and their address. “Other than that, nothing else gets shared,” says Godman.
A claims handler can check the database if a claimant has also filed a similar claim with another insurance company.
If yes, the handler can contact the other insurance company for additional information to determine whether there is a possibility of a double claim.
A customer, for instance, may file a claim for a stolen laptop. After conferring with the other insurance agency, the claims handler may find that the same customer has also filed a claim for stolen farm machine. “That overcomes the issue,” he says, because it meant that the owner took out a policy for contents with another company, and a policy for farm equipment with another company.
Fraud happens, he says, when the claims filed with two companies involve the same items or incidents. This can alert both insurance companies to send their own investigators and call the police.