Why Every Data Breach is Your Security Problem, and How to Fix It
By Bruno Farinelli, Fraud Analytics Manager, ClearSale
When you’re a merchant, it is scary to realize there’s an entire criminal industry built around stealing and reselling your customers’ data, but that’s exactly why there are so many data breaches of retailers, dating websites, government agencies, and healthcare providers. Data Breach Today recently reported that breach-notification service LeakedSource told the media to “assume every website has been hacked.” Fraud techniques are evolving in tandem with the type and quantity of data stolen during breaches, too. That means every data breach has the potential to affect merchants around the world. To protect themselves, merchants must have multiple layers of protection against different types of fraud.
Every merchant needs a security plan to protect against internal data breaches caused by insider theft, accidental data exposure, device theft, hacking through a third party, and phishing emails to employees. All merchants should also understand how data stolen in breaches could hit their bottom line. Here are some of the security challenges, from simple to complex, which merchants face when targeted by thieves using data from an external breach.
Simple credit card fraud
Let’s say a fraudster has access to the names, addresses, and credit cards numbers of merchant’s customers. With this information, he can target e-commerce stores that don’t have fraud controls by simply shopping with the stolen credit card real information.
Credit card fraud with location manipulation
Now, let`s assume our fraudster decides to target better-protected merchants who check customer addresses with the Address Verification System, which MasterCard, Visa, and American Express use in the US and other countries. This system catches attempted fraud by thieves who have credit card numbers but not correct customer shipping and billing addresses. In this case, though, because he has the customer’s addresses along with the card numbers, the fraudster’s purchases won’t trigger an AVS mismatch. In many cases, his purchases will be approved with no further screening or security.
However, the fraudster now has to reroute the package from the original customer address to a new delivery address so he can collect his stolen goods. Thieves usually do this by contacting the merchant’s customer service department or contacting the shipping company directly. To stop fraudulent orders before the thief gets the goods, merchants need another layer of protection. AVS screening is not enough, to filter out purchases that passed AVS screening, merchants who ship physical goods should put orders through another analysis, checking, for instance, if that order fits to customer’s expected behavior or profile. Reanalyze the order if the customer asks to change the delivery address after the transaction is approved is also recommendable.
What if the fraudster has all this stolen data and works for (or has a partner who works for) a shipping company? Now the stolen goods will pass AVS matches and avoid post-transaction delivery change screening, because he can intercept the package before it reaches its AVS-matched destination. Another option for fraudsters with access to detailed data is customer account takeover using stolen login credentials. Now the fraudster can change the delivery address on the customer account – another way to bypass AVS and delivery-change screenings.
These types of so-called sophisticated fraud and location manipulation fraud may seem farfetched to some merchants. However, they are very real problems, especially for merchants in the luxury goods and electronics categories, where ticket values and the resale value of stolen goods are higher than average. The PYMNTS Global Fraud Attack Index for Q2 2016 found that although sophisticated fraud accounted for a very small amount of overall fraud attacks (botnet fraud dominated), the average transaction amount per sophisticated fraud attack was US $173. That’s more than double the average amount per botnet fraud attack. The per-transaction figure for location manipulation was even higher: $294 per attack.
To counteract this costly and complex type of fraud, merchants need the most rigorous and comprehensive fraud screening capabilities possible. This most stringent layer of fraud protection should ideally include machine-learning tools that can screen transactions quickly according to behavioral analytics models and other input that evolves as fraud techniques change, such as consumer’s mouse behavior, time spent in each page, clusters of orders placed by a fraud scheme etc. It should also include human analysis to avoid raising the merchant’s costly false-decline rate and to ensure that the best possible data is fed into automated screening tools. By layering these tools over other transaction-screening methods, merchants face the best odds in the fight against fraud.
Bruno Farinelli is an expert in biometrics and browsing behavior, and heads up the Fraud Analytics department for the ClearSale US branch. Bruno holds a Bachelor’s degree in Statistics from top Brazilian University UNICAMP and an MBA in Business Intelligence from one of the most well-known Technology Institutes in Latin America FIAP. Follow on twitter @ClearSaleUS.