fbpx Skip to content

Leading Credit Card Issuer Finds Systematic Fraud, Saving $900k

For a bank or credit card issuer, detecting transactional fraud is relatively straight forward, and there are many systems in place to help identify this easily detectable type of fraud. Systemic fraud, however, is more elaborate and harder to detect than transactional fraud. Systemic fraud is masked as legitimate credit card transactions across 100s or 1000s of cards at one time.

The massive effort required to pull off a systemic fraud incident is why it’s typically only pursued by organized crime rings who want bigger payouts. Systematic fraud, which can take days or weeks to detect, leaves an expensive opening for criminals to mask their fraudulent transactions.

A leading credit card issuer finds instances of systemic fraud using Outlier’s automated business analysis platform. One Outlier story showed the credit card issuer that a fraud score increased 36% over the expected range for one country. When investigators looked at Outlier’s Root Cause Analysis, immediately they saw the merchant category code of Sporting Goods Stores was the reason. This Outlier story saved investigators time and at least $300k in financial losses.

 

As a credit card issuer, we need to find and stomp out systemic fraud quickly. We leverage AI and Outlier’s automated business analysis platform to help us shut down systemic fraud sooner, saving days or even weeks. In the past few months, Outlier found a few instances that have saved us over $900K.”

Project Manager, Fraud Analysis Team

A Leading Credit Card Issuer

 

Before Outlier, finding this incident could have taken weeks

Before the Fortune 100 Financial Institution deployed Outlier, they had a number of ways to detect systematic fraud. They needed a more nuanced view that could help them catch a more obscure systemic event as it was developing.

The above three Outlier stories show a real incident as it played out historically where Outlier found unusual behavior. On May 25, a fraud score for a category of merchants spiked 14% over the expected range. The next day the fraud score spiked even higher, to 25% over the expected range. By May 28, the behavior had developed into an upward trend and the fraud score was 143% above the expected model. Together, these insights tell an alarming story of fraud brewing. Had Outlier been in place, the first story would have been enough for the analysts to quickly act to shut down the fraud.

By deploying Outlier to identify fraud, this company decreases the amount of time fraud looms undetected and its cost. Outlier also helps with other use cases:

  • Alerts when transaction amounts spike significantly for various merchants, merchant countries, merchant category codes, and point of sale, which are used in fraud detection.
  • Finds drops in overall data volume to surface any general data quality problems.

Outlier helps to quickly find unexpected patterns in this company’s data to ensure fraud is found and resolved early. On average this Financial Institution considers each systemic fraud incident costs them $300K, so speeding up the detection and catching fraud faster will save the company millions of dollars per quarter.

Quick Facts

  • $900k
    Savings in a few quarters
  • 1000s
    Number of cards one systematic fraud incident could impact

Request a Demo

Discover what you’ve been missing.

Request a Demo