By Eric Tran-le

Success has a dark side: More electronic payments bring increased exposure to fraudulent transactions. According to The Nilson Report, losses from global credit card fraud are on track to exceed $31 billion by 2020. This year, 7.2 cents of every $100 will be lost to fraud. And the challenge is compounded by the fact that fraud is not only getting more sophisticated, it’s spreading to new payment channels and methods, such as prepaid cards and mobile.

For partners serving industries like retail and financial services, cutting that number by identifying fraud-related behaviors will be one of the most important services you can offer. Companies must learn to efficiently spot fraudulent behaviors throughout the process of issuing, acquiring and reconciling payments.

To stay ahead of the pace of fraud, you must help customers build intelligence strategies based on three pillars:

Quickly detecting fraud involves obtaining the most comprehensive view and heavily leveraging cloud computing so that data can be integrated and shared across the business, and often between third-party suppliers and other external sources. This requires not only gathering intelligence from every channel within the organization, but also pulling in a wide range of external data sources and performing correlations so that patterns can emerge.

Fortunately, many industries are forming consortiums to gather and publish various forms of fraud-related data for general use. For example, the Mortgage Fraud and Valuation Consortium’s database tracks more than 200 million records that can be correlated with a customer’s own data. Similar consortia in retail and other industries provide access to a wealth of outside data.

To help customers minimize losses through fraud, follow these five guidelines:

1. Automate, automate, automate: The volume and pace of fraud has clearly outgrown the ability of human beings to prevent it. Experian, for example, noted in its recent report that 2016 fraud levels were 15 percent above 2015 levels. You don’t need to be a mathematician to understand that combatting such growth demands automation — customers can’t raise their anti-fraud staff headcounts by 15 percent each year and stay profitable.

That means they need to begin automating every fraud detection-related item that can conceivably be automated, and they need help getting that done quickly. Consider, for example, the task of identifying patterns in streaming data sets that indicate fraudulent activity. While these patterns may be discernable to a specialized and costly data scientist, the only way that an average company can combat burgeoning fraud is to employ statistical pattern recognition and other technologies that automate the task of identifying threatening actions at massive scale. Likewise, companies need to employ systems of auto-alerting that, when such patterns are identified, route the information to the correct department or individual so that fraudulent activity can be quickly shut down.

2. Deliver easily digestible information.

The volume of fraud-related information makes presenting intelligence through spreadsheets untenable — the picture is too vast and complex to be digestible even for large teams of analysts. Think of it this way: For every spreadsheet a customer spends an hour poring through, half a dozen fraudsters have each discovered a new way to siphon money from the organization. One speaker at the Black Hat security conference had it right when he stated that “data visualization is the only approach that scales to the ever-changing threat landscape and infrastructure configurations.”

3. Be adaptable.

Nowhere is Heraclitus’ observation that the only constant is change more true than in fraud analytics, so possibly the most important consideration for visualization tools is that they be highly configurable, with the ability to provide a multitude of different styles of charts and graphs that enable customers to look at data from any angle.

Of related and equal importance is the ability to cross-reference every kind of data set — in particular, companies need to be able to very quickly correlate past data, which may have been gathered over the course of years, with data that may have been generated a second ago.

4. Use the cloud.

Extracting maximum value from intelligence requires making it available in real time to a wide range of teams that can analyze it from different perspectives. Sharing data promotes a wider understanding of where the next attack might come from and how to prevent it. 

While cloud services have been frowned on in some sectors due to the sensitivity of the data under analysis, attitudes are rapidly changing. In fact, new research from Evolve IP shows that health-care industry IT departments may now view the cloud as more secure than on-premises data analytics solutions.

For many customers, the cloud will be the only viable way to make data available to a wide range of users simultaneously, ensure that various parties are looking at the same data sets, and that they are kept continually up to date. Demonstrating the sheer scale that can be achieved, GoNet FPI, for instance, leverages cloud-based storage, processing, analytics and visualization to monitor more than 300 million online payments and investigate 3,000 cybercrimes per month.

5. Permeate all areas of fraud prevention.

With the right pieces in place, fraud intelligence can be applied to a variety of sensitive areas, such as online payments, digital transactions and social media monitoring — all ultimately aimed at brand protection and creating a more positive online experience for consumers. According to 451 Research, companies are really just getting started in applying the kind of analytics we’re talking about, with about 45 percent saying that they’re using machine data analytics for fraud detection. By helping customers leverage a wide range of data sets, heavy analytical automation, advanced visualization and cloud computing, you can ensure that more sales means a growth in revenue — not in online fraud.

Eric Tran-le is the global chief marketing officer at Logtrust Inc., a real-time big-data-in-motion as-a-service solution provider for fast, big data analytics, where he leads real-time security analytics initiatives. Eric’s more than 20-year career includes extensive work at both Microsoft and Oracle, spanning product management and operational system engineering for large-scale private and native cloud applications.