Staying ahead of online marketplace fraud

Interview

Staying ahead of online marketplace fraud

Jean-Paul Feidt
Department Head of Credit and Fraud
OTTO Payments

How prevalent is fraud within online marketplaces today? What are the most common areas fraudsters exploit?

Fraud in online marketplaces remains a major concern, often operating in the shadows. A lax approach to security can result in severe financial losses, sometimes reaching millions.

Fraudsters exploit any vulnerability they can find:

In the B2C sector, common methods include abusing new customer discounts, committing payment fraud, or manipulating return policies. Every payment method has inherent vulnerabilities, and the notion of a ‘completely secure’ payment system is a myth.
In the B2B sector, fraudsters are just as active. On marketplaces, fraudulent partners exploit process errors, sell counterfeit goods, or commit triangle fraud. Weak Know Your Customer (KYC) processes are often misused, and in some cases, marketplaces inadvertently make it too easy for fraudsters to operate.

These methods are sometimes combined, leading to attacks that span both B2B and B2C, increasing the complexity of fraud cases.

Much of the damage could be avoided, but negligence, overconfidence, and naivety are common. Often, minimal viable products (MVPs) are rushed to market, leaving critical security measures overlooked.

Companies frequently realize the severity of the problem only after substantial damage has occurred. Even in well-established setups, the smallest error can cost thousands of euros.

What strategies can marketplace operators employ to mitigate these risks effectively? To what extent can organizations be proactive rather than reactive?

The key to mitigating these risks lies in adopting a proactive and holistic approach.

Preventive measures are typically more effective and cost-efficient than reactive actions after a fraud incident. However, organisations need to ensure a balance between customer experience and fraud prevention measures. Calls for strict measures may be understandable but are often impractical.

The most critical elements are operational excellence and cross-department collaboration. Fraud is often treated as an isolated issue, with different teams addressing it independently, without proper coordination. One team may be working on one aspect while another focuses elsewhere, and internal politics can cause delays.

Moreover, there is often a lack of holistic understanding of e-commerce processes. For example, while IT is working on building a system, Fraud Prevention might be handling matters separately, leading to fragmentation.

A holistic approach to fraud prevention is essential but rarely fully achieved. Many companies do not possess a comprehensive understanding of all potential fraud risks. One solution is to map out a “Fraud Landscape” that identifies all relevant processes and potential attack scenarios. This is similar to an architectural blueprint but specifically tailored to fraud risks. Such an approach requires unified KPI management and cross-department collaboration to achieve optimal results.

What are your thoughts on AI in this context – what are the main opportunities and risks in your view?

AI plays a pivotal role in fraud detection and prevention in online marketplaces, offering both significant opportunities and notable risks.

Opportunities:

Efficient Real-Time Monitoring: AI analyzes vast amounts of transactional data in real-time, detecting anomalies and patterns that indicate fraud much faster than human operators.
Optimized Pattern Recognition: AI excels at uncovering hidden patterns, such as fraudulent account creation or abnormal transactions, predicting massive attacks before they happen.
Faster Learning Systems: AI constantly adapts to new fraud tactics, evolving to stay ahead of increasingly sophisticated threats.

Risks:

Data Quality: AI relies heavily on accurate data; poor-quality input can lead to incorrect fraud classifications.
Error Cascades: Faulty decisions by AI can trigger a chain of unwanted reactions, such as misidentifying legitimate patterns, leading to higher rates of false positives.
Privacy Issues: AI must adhere to data protection laws like GDPR while ensuring robust fraud prevention.
Inadequate Collaboration Between Human Expertise and AI: Human oversight remains critical. AI, combined with human expertise, creates the best outcomes. However, this requires a fundamental shift in how fraud detection and mitigation are approached.

In summary, AI is a powerful tool for fraud prevention.

However, its successful implementation depends on striking a balance between automation and human oversight to manage risks and maximize effectiveness. When done right, AI can deliver significant efficiency gains.

Enjoyed this interview? For more insights, be sure to attend MoneyLIVE Payments Europe taking place on the 20-21 November 2024, Amsterdam, where Jean-Paul and 500+ payments leaders will unite to define the future of payments.

Jean-Paul Feidt
Department Head of Credit and Fraud
OTTO Payments GmbH

Jean-Paul Feidt is a seasoned finance professional known for his expertise in credit risk management and fraud prevention. As the Head of Credit and Fraud at Otto Payments, he plays a pivotal role in overseeing the company’s credit assessment processes and implementing robust fraud prevention strategies. His dedication to innovation and efficiency has significantly contributed to Otto Payments’ reputation as a leader in the fintech sector.

Interview