In its second-largest acquisition, Amazon buys the company for $970 million.
41st Parameter’s TrustInsight Network scores individual customers based on collective data.
A new service from online fraud prevention company 41st Parameter helps e-retailers cut by a third the number of good transactions they mistakenly block as fraud, in turn boosting their revenue by 1% on average, the company says. Called the TrustInsight Network, it supplements retailers’ existing fraud-detection technologies with an additional piece of information, a “TrustScore,” about each customer based on information about a customer’s shopping activity on multiple retail sites, the company says.
The service launched for select merchants on March 21 and will be available to all merchants on May 28, 41st Parameter says. So far the network includes two merchants and one e-commerce platform used by 120 merchants, all of which contribute to consumers’ TrustScores, the company says. It declines to name those clients.
The TrustScore is similar to a credit score, but instead of creditworthiness it gauges the likelihood of a transaction being fraudulent, says 41st Parameter founder and chief innovation officer Ori Eisen. “We want to prevent false positives and do it in a way that doesn’t require massive changes in the payment network,” he says.
Improving fraud detection typically means adding more checks, such as home address, e-mail or phone number, to verify a consumer’s identity, Eisen says. But, for a large retailer especially, updating existing fraud-screening systems can be time-consuming, he says. TrustInsight circumvents that process by not requiring merchants, banks or consumers to provide or collect any additional information, Eisen says.
Instead, the TrustInsight Network uses its broad view of consumer transactions on many web sites to determine whether a shopper is who she claims to be, Eisen says. It does this by comparing how she’s shopped in the past on the web sites in the TrustInsight Network to her current checkout activity. The TrustScore is compiled from such data as the shopper’s average order size, the type of device she’s using, credit card information, e-mail address and phone number, 41st Parameter says.
The more merchants that join the network, the more precise the aggregated customer scores will be, Eisen says. And because TrustInsight scores customers with collective information, it can help a retailer validate a first-time customer’s transaction if the customer has purchased from another site in the network, he says. For example, many shoppers buy plane tickets just once or twice a year, and may buy them from a different airline with a different credit card each time. If both airlines are in the TrustInsight network, though, the second one can be assured her identity is valid and clear her transaction even though it hasn’t yet collected any data about her itself, Eisen says.
Besides deterring fraud, Eisen says the system prevents the flagging of legitimate transactions, which can lead to frustrating delays for good customers. “Sales go up, customers are happier and the cost of investigating and handling suspect transactions is decreased dramatically,” he says, though he doesn’t share specific figures for those savings.
"TrustInsight is unlike most risk management services,” says Russ Jones, a consultant at payments research firm Glenbrook Partners. “Instead of helping merchants identify and decline likely fraudulent transactions, they help merchants identify most good transactions—and there are a lot more good transactions than bad transactions."
In effect, the service lets customers carry their good reputation between online destinations, he says.
To use TrustInsight, retailers set up a simple data feed known as an application programming interface, or API, that allows 41st Parameter to transmit a customer’s score to the retailer’s payment processor as soon as she hits the Buy button on an e-commerce site, Eisen says.
For a merchant processing tens of thousands of transactions per day, TrustInsight costs $36,000 to $50,000 per year, he says. Because the price depends on the number of transactions processed, the returns vary based on the average order values, he says. At least one client paying $50,000 per year has generated more than $2 million in sales since implementing TrustInsight, Eisen says. He did not identify the client.