How FIs Adapt Fraud Detection During Times Of Behavioral Upheaval

The shuttering and reopening of states’ economies are causing seismic shifts in consumer behaviors. Tried and tested ways of catching fraudsters, such as comparing consumers’ activities against their 90-day transaction histories, are now resulting in false positives, says Carlos Mejia, chief digital executive at Pacific National Bank. In this month’s Next-Gen Debit Tracker, Mejia explains how creating short-term-only customer behavior profiles can help banks uncover bad actors.

Financial institutions can keep customers safe by accurately determining when attempted transactions are legitimate versus when they are being made by fraudsters who have stolen customers’ cards or taken over accounts. Banks’ security efforts depend on analyzing consumers’ transaction histories to gain firm understandings of their normal behaviors, enabling financial institutions (FIs) to spot unusual activities that could suggest fraudsters are involved. FIs generally look at 90 days of customer purchases when making their analyses as these wide time frames allow them to glean meaningful insights while still accounting for natural changes in spending patterns. 

The COVID-19 pandemic has undermined this practice, however, as stay-at-home orders and business closures prompted by the crisis have overturned what consumers’ normal purchasing looks like and rendered 90-day histories moot. Many consumers suddenly turned to eCommerce for items they previously purchased in stores, transacted in different frequencies and found new retailers. FIs confronting this new reality have worked hard to redesign their fraud-fighting approaches to suit the new commercial landscape. 

“You had a seismic shift in spending patterns, where a lot of card-present transactions all of a sudden moved to eCommerce,” Carlos Mejia, chief digital executive at Florida-based Pacific National Bank (PNB), said in a recent PYMNTS interview. “You’re really wanting to start with almost a brand new transaction history.” 

FIs have worked to accurately create new understandings of their customers by monitoring their behavioral changes during the pandemic and as businesses reopen, Mejia explained, and strong customer communication and machine learning (ML) tools have been key supports. 

Understanding Customers Amid The COVID-19 Pandemic

FIs that were well-versed in customers’ behavioral profiles before the pandemic have had to reestablish what consumers’ normal activities look like as they transact in vastly different ways, Mejia said. Banks have worked to build detailed customer profiles over the past several months as consumers adjust to their new financial situations and develop clearer online habits. PNB shifted to looking at customers’ past 30 days’ worth of transactions when stay-at-home orders went into effect to avoid relying on potentially inaccurate longer-range assessments, for example. 

“Much like you saw with card-present transactions, where [customers] consistently go to the same stores, you start to see a trend in the eCommerce sites that they use [and in their] spending trends,” he noted. 

Maintaining strong messaging with customers was critical to keeping them informed and trusting as they transitioned to eCommerce. The bank would send alerts about any potentially suspicious transactions, allowing cardholders to confirm legitimate payments or reject fraudulent ones. This helped the FI not only stop crime but also prevent fraudulent transactions from being incorporated into customers’ behavioral profiles. 

“Customers confirming transactions is the most effective fraud-mitigation tool in terms of building [an] acceptable history for that customer and being able to develop what their patterns will be,” Mejia said. 

ML is another powerful tool that can help FIs ensure their fraud-fighting models quickly adapt as consumers’ behaviors change, he stated. The technology can help detect developing patterns and compare specific consumer behaviors with those of the larger customer base. This analysis helps ML predict which transaction behaviors consumers are likely to exhibit in the future based on their current spending. 

Defending Agains Fraud During Periods Of Reopening 

States are now enabling some brick-and-mortar retailers to reopen, which is causing another major shift in buying habits. Banks must once again determine customers’ normal behaviors as they begin mixing in trips to physical stores with purchases made online, Mejia said. He recommended that FIs look at customers’ previous 30 days of transactions to establish new baselines for their purchasing behaviors. 

Reopenings could also lead to rises in certain kinds of fraud. Consumers who have spent months relying on eCommerce may not be as vigilant in following security recommendations geared toward in-person transactions, such as checking gas station pumps and ATMs for card skimmers. 

“[FIs need to be] watching out for complacency on the consumer’s end,” Mejia said. “Gas station transactions and a lot of the more traditional places where you have card compromise could see quite a [fraud] resurgence as places start to reopen.” 

Consumers’ behaviors and the overall retail landscape have undergone massive shifts, and additional changes are likely in store. FIs’ fraud responses must be fleet-footed as they work to adapt to changes and defend customers amid rapidly evolving circumstances.