false positives spiked banner

Online travel bookings present unique challenges. Travel reservations are highly fluid, with customers often making multiple changes right up to departure time, leaving little time to identify and prevent fraud. Online bookings through agency services are the hardest hit. Lost bookings plus the cost of staff to manually review suspicious transactions and handle disputes can amount to an estimated one to two percent of total revenue in an industry with already thin profit margins. And, as the online retail channel continues to grow, so too does the risk of fraud from “card not present” purchases.

An Online Booking Service client initially deployed an automated anti-fraud solution, but soon abandoned its use due to a rise in false positives. Their decision to operate without protection proved disastrous when word got out among fraudsters that it was easy to place orders.

Manual fraud-fighting only increased the problem

Frustrated by their inability to identify the sources of fraudulent transactions, the merchant began applying strict filters that only proved to catch all questionable transactions, not just those that were actually fraudulent. False positives spiked, risking customer satisfaction and loyalty in addition to loss of revenue. Despite the new strict rules, within weeks chargebacks increased to more than 3 percent of transactions. Consequently, the merchant was placed on the Visa Chargeback Monitoring program, resulting in penalty fees and increased manual overhead. Within a few months, the retailer was reviewing over 20 percent of all orders, requiring the addition of 15 employees dedicated to the task. When order declines reached a high-water mark of 19 percent per month, seriously eroding revenues, the agent turned to J.P. Morgan Merchant Services for help.

False positives: A multi-billion dollar problem

False positives are legitimate card-not-present transactions that are declined due to overly sensitive fraud detection parameters. While actual fraud loses represent an estimated 7 percent of the total cost of fraud, false positive losses amount to 19 percent.

The solution: Safetech Fraud Tools

Safetech Fraud Tools® were designed to meet this merchant’s challenge, retaining legitimate sales while preventing fraudulent transactions.

Our Safetech Professional Services Group moved quickly to activate and train employees on a comprehensive Safetech-based risk strategy that included:

  • Identifying and aggressively addressing the sources of the fraudulent booking activity
  • Significantly reducing the number of good orders that are held for review or declined
  • Utilizing PersonaTM Order Linking to analyze hundreds of variables and construct a definitive link to online purchase behavior — either directly or indirectly and help reveal fraudulent activity
  • Applying Multi-layer Device Fingerprinting to identify devices in real time, whether fixed or mobile, and analyze various key attributes that can indicate fraud patterns

The results

By leveraging transaction data we were able to assist the merchant in writing rules for high-risk booking types and origins, such as select hotels, short check-in dates or even certain countries. And, since Safetech technology allows merchants to send any data they receive from customers through to the platform, additions such as “passport,” “country,” “Hotel User ID,” and “City ID” contributed to a highly effective set of nearly 30 explicit user-defined fields. This collaboration and diligence quickly led to a reduction of 86 percent in the order decline rate. Safetech tools also helped steadily reduce chargeback rates to as low as 0.05 percent.

As online purchasing grows, merchants need ever more sophisticated fraud protection
The rapid growth of e-commerce offers organizations of all types the unprecedented opportunity to expand their customer base. But of course, fraudsters follow the money. Along with opportunity comes increased risk and increased costs to fight fraud.

Unfortunately, many merchants rely heavily on manual review, having staffers check orders before acceptance. In a recent survey of risk and fraud managers, “excessive manual review orders” was cited as the top challenge at e-commerce businesses that use automated flagging to identify suspicious orders. And that use is growing, undoubtedly due to the rash of data breaches and the transition to EMV (chip) cards, which is pushing more criminal elements into online fraud.

As effective as human risk analysis may be, applying that practice alone invariably results in a rise of false positives, sapping revenue growth while it increases operating expense. Industry sources agree on a dangerously upward trend in false positives, with as much as 35 percent of rejected orders turning out to be good, up from 25 percent year-over-year.

False positives: A multi-billion dollar problem
The Safetech Fraud Tools approach combines J.P. Morgan’s transaction processing expertise and scale. The solution represents a powerful Software-as-a-Service (SaaS) based application, including multi-layer device fingerprinting, proxy piercing, dynamic order linking, dynamic risk scoring, custom rules management and auto-decisioning.

These applications provide a unique solution that can dramatically enhance a merchant’s risk management performance.


Need a turnkey fraud solution? Safetech Fraud Tools delivers.

Using artificial intelligence and machine learning, our powerful fraud mitigation solution features multi-layer device fingerprinting, proxy piercing, dynamic order linking, dynamic risk scoring, custom rules management and auto-decisioning.

Safetech’s Identity Trust Global Network and machine learning algorithms can immediately alert merchants of signs of card testing attacks. Best of all, Safetech can be implemented quickly and rapidly provide e-commerce fraud protection.

To learn more, please contact your J.P. Morgan representative.