All-Star Fraud Fighters: Keeping Up with Fraud in the New Normal

2020 has seen a significant uptick in E-commerce sales as consumers follow social distancing recommendations and embrace online shopping and other digital transactions in record numbers. A recent analysis of the US Department of Commerce data by Digital Commerce 360 revealed that online spending accounted for 18.6% of all retail sales in the US in the first two quarters of 2020. Similar trends are playing out globally.

However, with the surge in digital commerce has come an unwelcome surge in fraud as well. In the latest virtual event in the DEFEND series entitled “All-Star Fraud Fighters: Keeping Up with Fraud in the New Normal,” DataVisor hosted a panel of professionals on the frontlines in the battle against fraud to learn what they are seeing in their own organizations and what strategies they are using to combat the increasingly sophisticated methods of fraudsters.

Fraud, Both Old and New, Is on the Rise

​​​​DataVisor’s own research indicates that fraud is increasing in both volume and velocity since COVID-19. There has been a 20% increase in account takeover attempts, a 40% increase in new account fraud, and transaction fraud has doubled.

When asked about what she is seeing in the field, panelist Carolina Rincon, Trust and Safety Leader for OLX Group, noted that digital fraud is both more prevalent and larger in scope since COVID-19. Mark Leung, Trust and Safety Officer for Letgo, agreed, stating that both volume and success rate for fraudsters is increasing since the pandemic began, largely because the playing field has expanded exponentially as more people moved their business completely online.

Expanding on Leung’s observation, Hao Li, Customer Success Team Lead for DataVisor, noted that as more traditional industries like banking have made the move to digital, fraudsters have upped their game to include the use of bots, machine learning, and social engineering. Using both old and new techniques and technologies, fraud has taken on a more “professional” tone as fraudsters make it their job and main means of income to defraud innocents.

Complex Fraud Patterns Call for Comprehensive Fraud Platforms

Fraud detection and prevention has never been an easy task, but today’s fraud landscape is particularly challenging. The sheer complexity of fraud today calls for more than simple point solutions. A comprehensive platform approach is needed. 

One of the tools that all panelists mentioned as a virtual necessity for organizations working on a large scale is machine learning. Tools that automate parts of the fraud detection process, like machine learning, help fraud practitioners make rapid decisions based on real-time data. Visualization tools like DataVisor’s Knowledge Graph also help fraud teams quickly identify and understand the relationships between accounts and detect suspicious patterns.

Other tools and strategies include rules engines, robust network security for blanket attacks, and human fraud practitioners, who add needed context and experience to the mix.

It Pays to Partner with the Right Vendor

Though machine learning has largely been touted as one of the best tools for fraud detection available, it is still a tool. To use it effectively, you must think about where to use machine learning in a comprehensive fraud strategy. Fraud practitioners must give thought to which areas of their fraud practice would most benefit from automation and leverage machine learning in those areas first.

In a frank discussion of the challenges of implementing machine learning solutions for their fraud practice, the panelists mentioned issues such as battling for the resources for implementation and ongoing maintenance and upkeep, educating team members who would be using the tools every day to understand how to use them effectively, and being realistic about how and where to rely on machine learning.

Partnering with a vendor that will offer support through the implementation process can mitigate many of these challenges. Choosing a vendor that offers training and education for your team members can alleviate much of the initial stress of an implementation and help your project start on the right footing.

Agility and a Multi-Layered Approach Win the Day

​At the end of the day, fraud practitioners know that the fight against fraudsters is not a skirmish, but a long and serious war with some of the smartest criminals on the planet.

​​​Defeating fraudsters requires an agile mindset and a willingness to iterate again and again as their methods of attack and technologies change. At the same time, fraud practitioners must ensure that their reporting and data are configured correctly so that they can act quickly on reliable data. 

Fraud detection and prevention requires a multi-layered approach that takes into consideration the fraud patterns that you recognize today and the ones that you will see for the first time tomorrow or years from now. A comprehensive fraud detection platform should be scalable, ready to grow as your needs grow.

Learn More with DataVisor

DataVisor’s comprehensive fraud detection platform uses multiple fraud detection tools including machine learning algorithms to help organizations stop fraud at the gates.