Online fraud was up 30% last year. If left unchecked, what’s the doomsday scenario for fraud?
Card-not-present (CNP) fraud is big business, and business is booming. The doomsday scenario is an increase in false declines. More and more legitimate consumers will experience a false decline because the merchant is trying to protect their bottom line, but at the cost of a good sale. The damage false declines can do to a company’s long-term sales and growth cannot be minimized or overlooked. In some cases, a customer who experiences a false decline will make 3x fewer purchases with that merchant in the next 12 months – if they come back at all. Merchants need to be doing everything they can to drive approval rates as high as possible, maximizing not only present-day conversion, but future sales as well. Radial recently completed a data study that found customers who experienced a false decline ended up spending $220 less in the following year with that retailer than customers who did not experience a false decline.
What does the average suspicious order look like?
Let’s flip this question around: what does a good order look like? The answer is a good order can look like anything. It can be high dollar, and have a high distance between billing and shipping. The customer may be shopping with the retailer for the first time, and may be using a free email account because they don’t want spam in their main account. The list goes on and on of potential red flags, all of which can and are often found on both good orders and fraudulent ones.
What’s some of the smartest and dumbest fraudster behavior you’ve seen?
The smartest fraudster behavior is the same as smart business behavior. Fraudsters who embrace intelligent automation of their processes are the “scariest” to deal with, as you need to deal with both scale and sophistication. Opportunistic fraud tends to be pretty dumb. This is where the fraudster ships the product to their own house or uses an email address that links to online accounts that boast about all of the items they’ve stolen.
Who’s better at fighting fraud today: robots or humans?
I’m a firm believer in robots for heavy lifting/bulk processing, and humans for exception cases. The companies that will win in fighting fraud are those who combine the best of machine learning with human interaction/manual review. Radial leverages both to get our approval rate of 99.7 percent.
Is mobile payments riskier than traditional commerce?
Mobile payments are notoriously tight lipped when it comes to sharing information, even when it comes to fraud prevention. In these cases, fraud management needs to rely on other pieces of information that don’t come from the payment method. As more volume gets pushed to these payment methods on closed loop networks, those networks will have to either open up a little more data or cover the losses if they want merchants to keep driving volume to them.
Biography: Michael Graff is a Risk Analytics Manager at Radial, where he manages fraud risk for more than 100 eCommerce merchants, and works closely with Radial’s Payments and Fraud analytics teams to identify opportunities to improve approval rates and mitigate fraud losses. He and his team of analysts use analytics, data, and machine learning to fight the next generation of criminals while driving as many good sales as possible for retailers and brands. Michael has a M.S. from Northwestern University in Predictive Analytics.