Like tightrope walkers, the PULSE Fraud Operations team sets out to block as much fraud as possible with minimal impact to the cardholder experience. Lean too far toward blocking fraud, and cardholder experience suffers. Lean too far the other way, and cardholders are happy, but fraud losses can multiply.
“There’s always going to be a balance between stopping as much fraud as you can without causing collateral damage by blocking good customers,” said Andrew Fong, Senior Manager, Fraud Operations with PULSE.
These efforts have been complicated in recent years following the U.S. implementation of EMV chip cards. Because chip cards are more difficult to counterfeit and therefore have significantly reduced fraud at the point-of-sale, fraudsters have shifted their attention to transactions in which no physical card is needed – e-commerce, e-wallet and other digital transactions.
Muddling things for fraud fighters like Fong and his team is that the shift just happened to coincide with a boom in card-not-present (CNP) transaction volume. Findings from the Debit Issuer Study, commissioned by PULSE and conducted by Oliver Wyman, highlight this shift. Five years ago, when the EMV migration was getting under way, just 13% of transactions were CNP. In 2019, that number had ballooned to 27%, and CNP accounted for an outsized 70% share of gross fraud cases. It’s also notable that while overall CNP losses have increased due to higher transaction volumes and ticket sizes, the loss rate per gross dollar value went down (3.3 bps in 2019 vs. 3.7 bps in 2018).
The surge in CNP transactions in 2020, which has been accelerated by the COVID-19 pandemic, combined with the tireless attention of fraudsters has elevated the urgency of addressing CNP fraud. That’s where PULSE DebitProtect® comes in.
Enhancements to DebitProtect
PULSE DebitProtect monitors all transactions on the network, detects potential fraud and blocks fraud based on network wide rules. With custom functionality developed by PULSE Fraud Operations, DebitProtect applies artificial intelligence and machine learning for modeling cardholder behavior and transaction scoring during the authorization process.
To detect suspected CNP fraud better, PULSE Fraud Operations recently rolled out an enhancement to the DebitProtect fraud model. Such enhancements are fairly routine as PULSE and its partners continuously make investments to leverage the latest in artificial intelligence and machine learning, but this one represented a leap forward in the effort to make CNP fraud more difficult.
The new model, which guides the decisions of DebitProtect’s machine learning system, incorporates specific analytic technology to distinguish between legitimate CNP transactions and potential CNP fraud.
CNP is Different
“The fraud model enhancements take into account that CNP transactions are different than card-present transactions,” said Fong. “CNP transactions have distinct risks and deserve to be scored in a different way, based on an algorithm specifically tailored for CNP.”
The enhancements have quickly paid off, resulting in an increase in account detection rate without increasing false positives. In fact, PULSE Fraud Operations reports a 7% increase in account detection rate and a 40% reduction in false positives. Soon after the enhancement was deployed, DebitProtect was recording a 60% increase in fraud dollars blocked.
“PULSE’s objective is to constantly improve our network’s fraud-detection technologies to improve accuracy and reduce the number of legitimate transactions that are blocked,” said Fong. “What we find most exciting is that by increasing the account detection rate, we can improve customer experience for cardholders while decreasing fraud losses.”
Another likely outcome of making CNP fraud considerably more difficult for transactions on the PULSE network is that criminals will be motivated to move on to easier targets that have a greater likelihood of success. The DebitProtect model enhancement is one more step toward driving nefarious activity away from PULSE issuers.
If you are a PULSE participant and haven’t already registered to use DebitProtect, contact your Account Executive or visit www.pulsenetwork.com and click “Connect with Us,” at the bottom of the home page.