October 22, 2020 - Todd Ludwig

4 E-Commerce Companies Waged War with Fraudsters: Here’s What Happened

For all the benefits and convenience that e-commerce offers consumers, it’s also created complexities for businesses. One of the chief concerns shared among all e-commerce merchants is the ability to differentiate between honest transactions and potential acts of fraud. Knowing this difference is critical, as fraud is an estimated $12 billion problem that comes directly out of merchants’ bottom lines. What’s more, it is estimated that every dollar stolen from a merchant by fraud will actually cost the merchant $2.40. 

Fraud is a major threat to profitability, but it can also cause serious damage to your reputation. Merchants that become victims of fraud demonstrate their vulnerabilities, and even a strong response system isn’t always enough to earn back consumer trust. 

In today’s landscape, fighting e-commerce fraud is an ongoing war. With the right weapons and strategy, merchants can position themselves to win bigger and lose smaller over time, even as fraud patterns and techniques evolve.

Effective Fraud Detection and Prevention: Real-Life Lessons

To be clear, most e-commerce companies have some form of fraud detection measures already in place. Traditionally, fraud detection relies on rules-based systems that have been taught what to look for, then will create an alert when certain behaviors are triggered. Today, however, rules-based fraud detection is no longer enough.

Rules-based systems only account for fraudulent activities that have already been discovered. These systems are continually retrained as new fraud emerges so that they can account for future occurrences of that same nature. However, fraud is evolving at lightning speed, so traditional systems cannot capture all occurrences of fraud, leaving merchants to continually play catch up.

Two Online Marketplaces Go Proactive to Stop Fraud Before It Happens

Shifting from a reactive rules-based approach to DataVisor’s proactive, machine learning approach allowed one global online marketplace with over 350 million monthly active users to capture 88% of fraudulent accounts before the first scam. This proactive approach resulted in over 20% detection accuracy improvement over that achieved by the organization’s prior fraud solutions.

Similarly, a large mobile customer-to-customer marketplace was having trouble maintaining a safe and trusted platform where consumers could trade goods and enjoy a seamless customer experience, largely due to fraud and abuse activities such as fake listings, spam, and account takeovers. By examining all activity holistically across all users instead of evaluating user events one-by-one as a traditional system would, DataVisor uncovered hidden patterns of fraud and abuse common to fraudsters. Having identified the patterns, DataVisor’s systems flagged malicious accounts earlier and blocked them before damage happened, resulting in a 10x increase in detection over existing rules-based fraud systems in place.

Lesson learned: Reactive rules-based methods are out; proactive methods are in.

Two E-commerce & Delivery Companies Fight Organized Fraud Holistically 

What’s more, fraud is no longer reserved for simple one-off activities committed by individuals for personal gain. Online crime is often conducted by complex, organized crime rings that can only be detected from a holistic point of view. On the surface, individual behaviors can easily bypass typical rules detection systems because activities look normal. 

This was the case for a food delivery company that was experiencing a high refund rate due to order cancellation after the order had been fulfilled. When the company started using DataVisor’s machine learning solutions, they found that these requests were being made mostly via mobile devices and pertained mostly to cash orders. Other details also signaled a sophisticated crime ring. Once DataVisor’s fraud detection technology identified the issue, the company was able to quickly make bulk decisions and prevent fraudulent transactions from being completed, saving $6 million in fraud losses.

A top delivery services company that processes over 6 billion packages per year was experiencing high rates of shipping fraud and mass registration of fraudulent accounts. Leveraging DataVisor’s comprehensive fraud detection solution, this company was able to preserve trust and safety on its platform, prevent fraudsters from hijacking its services for criminal purposes, and save over $4 million in the process.

Lesson learned: A holistic approach is needed to spot organized crime patterns.

Learn More about Winning the War Against Fraud

Waging war on fraud is no longer optional for companies that want to protect their bottom lines and reputations. Download the ebook Digital Businesses Say No to More Fraud to read more about how the four e-commerce companies mentioned above are leveraging DataVisor’s comprehensive fraud solution to reduce fraud costs and risks and deliver an optimal customer experience.

about Todd Ludwig
Todd Ludwig is an accomplished Cyber Security service professional with 20+ years of experience in the industry. Starting in the early days with a focus on education and training on information security best practices and advanced training for government agencies and Fortune 500 organizations. For the last ten years, I have been working with eCommerce companies of varying stages provide a secure, customer centric, high-performance experience. More recently I have moved into the use of AI to provide fast, reliable, and accurate detection of fraud and security events.
about Todd Ludwig
Todd Ludwig is an accomplished Cyber Security service professional with 20+ years of experience in the industry. Starting in the early days with a focus on education and training on information security best practices and advanced training for government agencies and Fortune 500 organizations. For the last ten years, I have been working with eCommerce companies of varying stages provide a secure, customer centric, high-performance experience. More recently I have moved into the use of AI to provide fast, reliable, and accurate detection of fraud and security events.