Feedzai is attempting to sell the U.S. on a multichannel approach to card data security in the wake of the recent data breaches afflicting major retailers.
"Bank robbers used to break into banks because that is where the cash is. Today, it's card payments and card data," says Loc Nguyen, chief marketing officer of Feedzai, a London-based payment security provider that has operations in Europe, Latin America and Africa.
While Feedzai's entry to the U.S. was not prompted by the recent incidents at Target, Neiman Marcus and Michaels, the events draw attention to how crooks can be adept at exploiting consumer and transaction data, Nguyen says.
"There is a data problem. It's not being used to act and react faster to security threats than what is happening on the criminal side," Nguyen says.
These data issues are becoming more complex as retailers enable "omni-channel" shopping, allowing consumers to begin the shopping process in one channel (such as a mobile app) and complete it in another (such as a store). However, many fraud engines focus on only one channel, says Julie Conroy, a research director at Aite Group.
"Many retailers' payment acceptance systems have been build out in silos, so having the ability to detect fraud across multiple channels is just as big a problem for them as it is for banks, if not more so," Conroy says.
Feedzai uses machine learning (a form of artificial intelligence that accumulates and analyzes data to improve a program's performance) to predict electronic payment losses based on behavior analysis of consumer actions in stores and digital venues. The software accumulates transaction information to build profiles of consumers, merchants, merchant locations and point of sale terminals.
"Crooks are stealing data from one place and using it in another, so you want to be able to create a profile that's more than one segment of a person's activity," Nguyen says.
The profiles include analysis of up to three years of data for consumers, with attention to the different ways consumers use each channel, Nguyen says.
"It's like how your spouse knows your spending habits, this artificial intelligence knows your behavior," Nguyen says. "Even if the data is breached, it will be used by the crook in a manner that deviates from the normal pattern."
The analysis can also reduce false positives, since purchases made in different channels can reveal a pattern of legitimate spending, Nguyen says. A large online airfare purchase, for example, can be used to clear a purchase made at the traveler's destination, an out-of-market purchase that may otherwise get flagged.
Feedzai did not name its customers in the U.S., but its international clients include Coca Cola, Logica, Vodafone, Ericsson, Payment Solutions and Servebase Credit Card Solutions. Feedzai's experience in Europe should prove helpful, particularly as the U.S. market migrates toward EMV-chip cards, which deter theft at the point of sale but may direct crooks' attention to card-not-present channels, Nguyen says.
"Those merchants with an e-commerce presence are actively looking for ways to fortify their defenses in a customer-friendly way, recognizing that fraud attempts will sharply intensify as the U.S. migrates to EMV," Conroy says.
"To that end, there is a benefit in having a vendor with deep expertise in the European market, where they have helped merchants live with the fraud shifts that take place, and have learned some of the hard lessons about what works and what doesn't from a customer usability perspective," she says.