The operator of GiftCardMall.com wanted to fulfill orders faster, but without compromising its ability to weed out criminals.
When parent Blackhawk Networks, a third-party provider of retail gift cards, started using fraud detection software vendor Trustev in mid-2014, fraud fell dramatically. The decrease was the result of blocking troublemakers from buying cards on its main site, GiftCardMall.com, and of identifying legitimate customers, says Erik Brazzo, Blackhawk’s vice president of risk management.
Using Trustev to complement Blackhawk Networks’ internal anti-fraud efforts lets online shoppers immediately buy gift cards rather than wait while their purchases and trustworthiness are verified manually, Brazzo says.
Trustev is growing at a rate of 50% a month as retailers increasingly recognize how much business they are losing to e-commerce fraud, Trustev CEO Pat Phelan says. The company doesn’t disclose revenue, but reports that among its clients are online marketplace Jet.com, European shoe brand Camper and U.K. mobile telecom company O2, a commercial brand of Telefónica UK Ltd.
Phelan says Trustev, a 4-year-old company based in Cork, Ireland, with U.S. offices in New York, uses a hybrid of machine learning and human intelligence to help retailers keep customers happy, both in their expectations for receiving goods quickly and by not flagging honest customers who could get insulted if the retailer asks for a copy of a passport or other evidence that they’re not thieves. The software also can cut manual reviews that retailers often must use to weed through suspect transactions, he says.
The company’s software code is dropped into each page of a retailer’s website, rather than just at the payment point, and it analyzes behavior such as whether the customer is hiding his or her computer or device IP address; acting like an ordinary shopper browsing the site or quickly rushing in and out; or using a familiar rather than a new email address. The system, which takes a half-second to pinpoint potential fraud in many cases, can check the customer against social media sites, find voided telephone numbers and check applications in real time, Phelan says.
The system takes four weeks to set up, including a two-week listening period in which Trustev creates a database to compare incoming IP addresses against those known as legitimate. From 2% to 10% of orders draw attention for potential fraud, Phelan says.
“We’ve tuned our system so well, 90% of our customers don’t have manual review after 60 days” of using the software, Phelan says. “Across our customer base, the trend is to eliminate manual reviews altogether for both smaller retailers who don’t have a payments fraud team and larger companies whose current review rate is totally unmanageable.”
That’s particularly important because many retailers have amassed large staffs, including one department store who Phelan declined to identify that employs 300, to manually review customer orders that raise red flags for fraud potential. These retailers increasingly find themselves in a bind because they have set up a two- to three-day window to see if alerts pop up on a customer’s credit or debit card, but customers increasingly demand next-day delivery, he says. And demanding same-day delivery is a common red flag of possible fraud.
Trustev charges a flat fee per transaction and cites a 2%-16% jump in its clients’ revenue within 90 days of installation, he says The company doesn’t try to “rip and replace” a retailer’s existing systems, but is able to phase out a customer relationship management system as unnecessary within 12 months, he says.
Though experts predict that online fraud will double in the United States because of a new in-store card payment system, Phelan says the fallout will take a while to see because the chip card system is being phased in rather than implemented all at once. Trustev has raised $7.8 million in funding from investors TransUnion, Greycroft Partners, Mangrove Capital Partners and Wayra. Its client base of 97 includes more than 70 headquartered in the United States.
Favorite