Use Business Rules and Profiling to Guard Your E-commerce Transactions
By Ken Burke
With e-commerce fraud on the rise, e-tailers are eager to find ways to cut their losses. Fortunately, a common e-commerce technology can be used to create an effective system that can filter out bad transactions before they get processed.
Any e-commerce Web site can be set up with its own automated anti-fraud system. The goal is to have your Web site watch for transaction profiles that indicate an elevated risk of fraud, then hold those transactions for scrutiny by a trained customer service representative (CSR). More advanced versions can automatically rank each flagged order according to risk and send the customer an appropriately worded e-mail.
A Basic Profiling System
Profiling is all about spotting certain customer activities or transaction characteristics that indicate potential fraud. Profiles use business rules that examine every order immediately after the customer submits it but before it is processed. If any of the key data points meet the fraud profile, the order is set aside for further investigation by a CSR. Orders that do not match any of these points are processed as usual.
Certain conditions are known within the industry to signal an increased likelihood of fraud:
>Ship-to and bill-to addresses are different and the dollar amount is over a certain amount;
>Expedited shipping is requested and the dollar amount is over a certain amount;
>The total number of items exceeds a certain level;
>The quantity of any one item exceeds a certain level;
>The order originates from an IP address that statistically generates a large number of fraudulent transactions;
>The customer's e-mail address is from a domain that statistically generates a large number of fraudulent transactions;
>The customer's e-mail address or other information is on your blacklist of known fraud perpetrators.