“We believe your card has been compromised and would like to verify the last couple of transactions?” That call from the bank can be a cause for fear or frustration to credit card holders. It could be worse: The card holder is at the ATM to withdraw cash or in a retailer poised to make a purchase, and the transaction is turned down. Now the customer has to make an embarrassing call to the bank or credit card company to learn the reason for the denial – knowing all along that the account should be fine. Sound familiar?
Unfortunately, this is what many banks and their customers are facing as banks try to safeguard their customers from fraud. Unfortunately, most fraud monitoring systems are based on rules – possibly even models – that try to predict if a transaction is outside of a person’s normal activity. The issue is that most of these rules and models are based on outdated information, so they can’t analyze the up-to-the minute behavior changes that many customers make.
For example, I may usually take out a $100 in cash each week for personal expenses. Also, I travel to various places for work about once or twice a month. (During those times, my transaction behavior changes.) I use a work credit card for most charges, but if I’m buying something personal while away – like a souvenir or toiletries – I use my personal card. During those times, a fraud business rule might be triggered if my travel pattern is significantly out of my normal behavior. You see, I might have used my personal card at home in the morning and abroad a mere six hours later.
This is often referred to as analytical behavior analysis: The ability to understand how the individual behaves and how the behavior may or may not change on a regular basis. Most of the systems used today can’t keep up with the “profile” or “customer signature” of individuals by looking at their entire account OR joint account behavior in real-time. Because of this limitation, most systems create too many false positives.
There lies the delicate balance of avoiding losses due to fraud while keeping the customer from being annoyed and possibly leaving the bank altogether. Customer behavior analytics combined with real time transaction monitoring of all transactions allow for greater accuracy than the flagged or stopped fraudulent transaction. In these cases, the customer call that says they have found suspicious activity can actually make the difference between an annoyed customer and a loyal customer.
Read more about reducing your customers’ frustration while still keeping them safe from fraud in this white paper about applying a multilayered approach to fraud management.