The holidays are a stressful period for online retailers, who get not only more customers but also a higher ratio of fraudsters. And when manual fraud reviews are added to the mix, the problem only gets worse.
According to a report by ACI Worldwide, online fraud attempts went up 31 percent this past holiday season compared to last year, while transactions went up only 16 percent. In fact, 1 out of every 97 transactions was a fraudulent attempt.
There are two main risks associated with manual reviews during peak shopping periods, said Michael Reitblat, CEO at payment security vendor Forter.
First, external fraudsters will try to take advantage of employees who are trying to handle a higher volume of transactions than at other times, and of under-trained temporary employees who were hired specifically to deal with the rush.
Typical frauds include criminals buying goods with stolen payment cards, or calling companies and trying to extract sensitive information that they can use for future frauds.
Second, fraudsters will also try to get jobs at retailers during these frantic hiring periods, to defraud companies directly, or to steal customers' financial information.
"We've seen several cases where, because people were not properly trained, criminals were able to extract information from them that they could use somewhere else, or they were fraudsters on their own and they have access to information about customers and then go and sell it on the black market," said Reitblat.
Temporary employees hired to review transactions for fraud get access to a great deal of sensitive information.
"They have access to all your systems that can look at your transaction history, account history and so on," he said.
According to Forter, the rates of incoming attacks went up 39 percent from the third quarter to the fourth quarter of last year.
"Not only is total fraud substantially higher during the holiday season, but as a proportion it was also higher," he said.
Employees hired to do manual reviews are also less consistent than automated systems, he added.
"When you compare two different agents that try to make a decision on one transaction, they won't always make the same decision," he said. "Different people have different biases when it comes to ethnicity, location or just based on their moods that day. And if you're pressured you won't be looking at transactions with the same level of detail."
When companies switch from manual to automated fraud reviews, charge backs go down by 30 to 70 percent. Plus, approvals increase by 5 to 10 percent.
"This indicates that automation is more effective at both stopping fraud and avoiding false positives, as well as speeding up the process and reducing delays since every decision is real-time," said Reitblat.