Online firms and their banking partners have both the means and motives to stop fraudsters from accessing personal information, but today’s connected economy encompasses markets and customer service channels around the world. Thwarting fraud in one area is not enough to prevent it from cropping up elsewhere, and businesses are essentially playing whack-a-mole as fraudsters’ attacks move across channels. These bad actors are leveraging numerous methods — including false mobile apps, loyalty and reward schemes, ATOs, phishing and phone scams — to find and exploit security flaws.
Digital fraud rose 30 percent in Q3 2019, with one out of every five new accounts revealed to be illegitimate. A data breach in one channel can have far-reaching consequences in others, too, making omnichannel fraud dangerous for both businesses and their end customers. Personal details stolen via one false account can be utilized to break into other platforms, thus perpetuating an endless cycle of fraudulent transactions.
There are tools that can protect businesses from these growing threats, however. Fraud decisioning technologies that rely on AI, ML and other advanced learning tools can categorize and weed out fraud faster than human analysts, but businesses must recognize that fraudsters possess similarly sophisticated tools. Fraud decisioning technologies, therefore, need to be applied in the appropriate scenarios to effectively combat cross-channel fraud.
Omnichannel commerce needs omnichannel security
Online businesses and their partners must fight to lock out fraudsters, even as the latter find an increasing number of ways to access sensitive data. Consumers are now comfortable interacting and transacting across various channels, including mobile app, phone, online and in person. They appreciate access to multiple channels, though many may prefer one over the other, meaning businesses cannot go back to providing fewer ways to communicate or accept payments.
No anti-fraud strategy is completely effective across channels, however, and therein lies the issue. Businesses need to devise fraud protection methods for each market and channel in which they operate. The dizzying number of vulnerabilities companies must monitor can leave them feeling overwhelmed, with 65 percent of retailers stating that they do not have the fraud protection tools necessary to adequately manage the omnichannel environments in which they operate.
Fraud decisioning tools bolstered by advanced learning technologies can alleviate these issues, but there is no one-size-fits-all solution. Effective fraud detection software is a necessity, but having a skilled team that can determine when, where and how to apply the insights that it provides is even more critical. Fifty-four percent of retailers believe they do not have such teams in place, meaning they are due for security strategy upgrades.
AI can be used to protect against loyalty and reward system fraud, but it must be carefully calibrated to do so. The technology can spot hackers who have created false loyalty accounts, made phony reservations or used personal information from other sites to attempt to game hotels, for example, but it must have access to all data points that surround the booking and payments processes to adequately distinguish between fraudulent and legitimate customers.
The same rules apply to the fraud decisioning tools that retailers employ. These solutions often fail to track customers’ behaviors in more than one channel, allowing fraudsters to exploit the fact that merchants and their security providers are not matching users’ transactions across channels in real time. The good news is that businesses can move their fraud strategies and decisioning tools forward in several ways, and technologies such as behavioral biometrics can provide additional security to separate legitimate customers from fake ones.
Businesses and banks are already using fraud decisioning tools to protect customers, but they must alter their use of these solutions. Companies need to remember that omnichannel commerce means omnichannel fraud is a given, and that bad actors have access to the same technologies. Fraud decisioning methods are only as good as the data that powers them, after all, and they must have access to as much information as possible.