Rewards points and miles hold true cash value. However, with fewer monitoring and protection measures in place than traditional bank accounts, the unfortunate reality is that loyalty reward programs are seen as easy targets by fraudsters.
Loyalty fraud: The problem
In fact, as many as 72 percent of loyalty program managers have experienced an instance of program fraud. This type of breach can invite a variety of negative repercussions: increased costs to replace points or miles, loss of loyal members, even a lowered NPS after a bad customer experience.
Loyalty fraud: The solution
Fraudsters are finding new ways to steal identities to open fraudulent rewards accounts, even joining cybercrime rings to steal and sell miles and points online. Criminals are getting smarter. And so must the protections against them.
Pairing rule-based with artificial intelligence (AI) to combat fraud
- Rule-based approach: Traditional loyalty fraud prevention programs operate based on rules. A rule-based approach uses a series of static if/then statements to filter out good events from bad events. For example, if multiple fraudulent redemptions are made online from a certain IP address, a rule can be established to flag all such redemptions for review before fulfillment.
- AI modeling approach: Employing artificial intelligence for loyalty fraud detection and prevention utilizes a system that learns the difference between good and bad events at scale over time, without human intervention. Machine learning, for example, is a type of artificial intelligence used for fraud detection in which machines are given large amounts of data. The machine then “learns” to predict behaviors or events over time.
Pairing rule-based with AI allows loyalty programs to draw conclusions and uncover trends that would take a human much longer to identify. The pairing of these tools allows for a proactive, rather than a strictly reactive, approach to identifying fraud to protect loyal customers.
A customer-centric approach
This proactive approach also helps ensure reward members’ points are safe without interrupting or damaging the customer experience. The ability to predict future instances of fraud before they occur protects the customer by identifying risk and reducing the amount of human intervention needed to assess and identify fraudulent activity.
The result is less friction for the customer and the ability to ensure that rewards are received by loyal members in the timeframe they expect. Brands that pair rule-based with AI modeling improve protections and reduce the impact that fraud monitoring has on the overall customer experience.
As long as fraudsters continue to alter tactics and companies continue to launch new products and features, there remains a need to integrate rule-based and AI models.