Brazil has one of the world's worst payment fraud problems, which has come in handy for ClearSale as it tries to cut transaction decline rates for online payments globally. The Olympics also came in handy.
"The fraud rate was just so much higher in Brazil, so we grew in one of the most difficult markets in the world," said Whitney Hudson, regional sales director at ClearSale. "We have expanded globally since then and, knock on wood, the fraud problems we are facing here in the U.S., we have handled those previously in Brazil."
Brazil's security industry also had to grow up fast when the country was preparing to
"We have been providing input to law enforcement there, doing what we can to crack down on fraud," Hudson said. "There is a large amount of inequality in Brazil and it is primarily opportunists taking advantage."
ClearSale was founded about 20 years ago in Brazil, fighting against online fraud that grew so rapidly in the Latin America market after its
But ClearSale believes these early challenge has fed its machine learning technology and human analyst teams with the tools to combat payment fraud in other markets.
ClearSale deploys more than 700 analysts in its centers around the world who look over orders deemed suspicious after monitoring of the age, email address, billing address, the IP location and other facets that machines pick up on.
"It's about approving more transactions and giving them that extra look," Hudson said at the recent IRCE/RetailX conference in Chicago. ClearSale was debuting its new order approval rate calculator technology to help merchants balance approvals rates, false declines and chargebacks.
Up to 95% of transactions it monitors are automatically approved, while the other 5% are delayed for one to three hours while analysts examine them more closely, ClearSale estimates. "Fraudsters are sophisticated and there a lot of gray areas that a machine might not pick up on," Hudson said. "And if it were machine learning alone, there would likely be declines on good orders."
For example, a consumer purchasing a $5,000 handbag might decide a few minutes later to purchase a matching wallet for $2,000 from a high-end apparel merchant site. A machine would view that as two expensive transactions from the same card with the same retailer within minutes — a common fraudster trait — and decline it based on the general rules it is observing about what other types of purchases have been made on that card.
"A human analyst could determine that the consumer was simply looking to buy a matching set," Hudson said. "We have to make sure, from the customer side, that we are not turning down good orders."
E-commerce merchants need to conquer this problem as they continue to put up layers of card-not-present defenses from
Many merchants "are somewhat reticent" to fully outsource the transaction risk assessment to a vendor because they are worried about the potential for higher than acceptable false decline rates, said Julie Conroy, a research director at Aite Group.
"But ClearSale's blend of machine learning analytics combined with the human intuition that comes with manual review is designed to address this concern," Conroy said.
Joy Woodruff, marketing and online sales director for Sound of Tri-State, would agree with that assessment. Prior to hiring ClearSale to monitor transactions, Sound of Tri-State, a Claymont, Delaware-based car audio and video consumer electronics company, had its own employees Googling the customer address and calling them to confirm orders over a certain dollar amount, Woodruff said.
"Under a certain amount, we were taking on the risk ourselves," she added. "With ClearSale, our employees can concentrate on sales and customer service rather than spending time investigating whether a sale is legitimate or not."
ClearSale has already caught many fraudulent transactions that Sound of Tri-State can monitor through its ClearSale platform. "We feel pretty confident that Clearsale will employ data driven technology as well as AI and the human factor to help their business evolve with the times and with any new tactics that fraudsters employ in the future," Woodruff added.
It's allowing U.S. businesses to follow the lead provided in Brazil, where ClearSale reports 90% of merchants are using its technology, creating not just a knowledge base but also an impetus to expand.
"When you have that much of the market, you have to look beyond your borders to grow," Hudson said. Since those early days, ClearSale now has offices in Miami and Mexico City as well as Sao Paulo, Brazil.
ClearSale works with retailers of all sizes, with many of its clients falling into the high-risk categories such as automotive, high-end apparel and sneakers.
"Our main hurdle really has been for people to recognize who we are and about our roots," Hudson said. "It honestly really is a positive that we started in Brazil because it helped us solve problems for retailers in the past and, if they are not facing those specific fraud threats today, they will probably see them soon."
