Fraud Screening - The Solution

The Requirement

Most insurers provide online insurance direct to the consumer. This is a highly exposed "customer not present" environment. Although the vast majority of online customers use insurance facilities for legitimate purposes, unfortunately, a very small minority take out policies where fraud is the intent.

Online fraud takes many forms. In its simplest form, it can involve payment for a policy with a credit card that is either stolen or is used without the owner's permission and, frequently, the policy is used to support another action such as taxing a motor vehicle. But, more importantly, it can also involve fraudulent or dishonest claims. The focus of these frauds is an intent to deceive and a desire to induce the insurance company to pay a fraudulent claim or to pay more than it would normally be liable to pay.

Fraud increases the loss ratio and can seriously damage the competitiveness of online insurance.

Screening for fraud risk using the data captured for the policy is a specialist process that can be used to decline applications when a fraud risk assessment threshold is exceeded.

The Solution

Discover IT has developed and deployed a comprehensive fraud screening system for one of its customers selling short-term automotive insurance. This has been operating successfully for over 3 years. Within a few months of introduction, the loss ratio for the product was improved from "unacceptable" to "highly competitive", and has made a major contribution to the success of the product.

The screening system uses a 3-stage process with "reactive" and "proactive" stages. The key inputs used include client name, date of birth, address postcode, email address and day and time of submission. The key stages are:

  • Reference to a fraud history database.
  • Application of custom rules.
  • Application of credit card rules, including "velocity" check.
    Exclusion of credit/debit cards from "exclusion" countries, based on a BIN lookup database.

The fraud screening process is hidden and works in the background. The rules are never exposed to the customer at any stage.

Fraud Screening System Model

Fraud Risk Detection Process

Reactive Stage – This stage can only return a Pass or Fail result. Performs a "fuzzy" lookup of prospective customer’s input data against the fraud history database.

  • Pass – Progresses to next stage.
  • Fail – The applicant is informed that a quote for the risk is not available. No reason is given.

Proactive Stage – Test prospective customer’s input data, including postcode, against custom rules. The combination "score" is returned that results in a Pass, Refer or Fail according to the thresholds set.

  • Pass Progresses to next stage.
  • Refer – A page displays informing the applicant that they should phone the call centre to progress the application. The quote record is updated with the stage fraud score. They cannot progress to the next stage; only the call centre can progress the application.
  • Fail – The applicant is informed that a quote for the risk is not available. No reason is given.

Final Stage (Credit/Debit Card) – Test prospective customer’s credit/debit card for excluded countries first, and then apply the normal payment clearance AVS, velocity and 3D-Secure rules. Results in an Accept or Decline.

Fraud Screening Administration

Fraud Screening Administration

An administration interface allows rules to be reviewed and adjusted based on experience gained. Previous fraud data can be used to test the effectiveness of the ruleset.

Also, individual fraud history database records can be added, edited and deleted. Bulk upload of fraud history database records from a spreadsheet is supported.

The whole process has been designed to enable the fraud stages to be switched on and off without disrupting the service.

In an ever changing, fast moving marketplace, there is a high cost of doing nothing. Fraud comes straight off the bottom line and higher costs equates to higher prices. If your competitor has accurate data that it can trust, and lower levels of fraud, then they are in a stronger position to offer more competitive products.

We're ready when you are. Let's talk.