Signifyd’s Developed An AI-Powered Response Solution To Combat Retail Fraud

In an era that has given rise to e-commerce and digital transactions, fraud has become an ever-pressing challenge for retailers. Cybersecurity Ventures estimates cyber-crime costs organizations $1.9 million every minute, and are predicted to grow by 15% to a staggering $10.5 trillion annually by 2025. Businesses are grappling with the need to optimize their revenue streams while safeguarding against the downside risks of fraudulent activities.

Some alarming statements from the report:

“If it were measured as a country, then cybercrime …would be the world’s third-largest economy after the U.S. and China.”

“This represents the greatest transfer of economic wealth in history, risks the incentives for innovation and investment, is exponentially larger than the damage inflicted from natural disasters in a year, and will be more profitable than the global trade of all major illegal drugs combined.”

Amid this complex landscape, Signifyd, a company driven by big data and machine learning, has emerged as a champion of “fearless” commerce, providing retailers with a powerful shield against fraud and chargebacks. I spoke with Raj Ramanand, Co-Founder and CEO of Signifyd to discuss how they are tackling this pervasive threat.

Unraveling the Impact of Fraud on Retailers

What is the impact of fraud on retailers today? Traditionally, when an online fraudulent transaction occurs, retailers tend to bear responsibility. In today’s system, Ramanand explains the mechanisms in place for each individual transaction:

“To put it simply, if I were to steal your credit card and make a purchase from an online retailer, four parties would be involved: myself (the thief), you (the cardholder), the retailer, and the bank. In the current setup, when someone makes an online purchase, the retailer bears full responsibility for any fraudulent transaction. This means that the banks and consumers are shielded from any liability. That’s how the transaction flow operates in the present system.”

As a result, retailers have started putting protections in place, effectively adding friction across the customer journey to manage the liability for fraud. In return, they end up turning away or declining hundreds of millions of dollars annually in legitimate sales.

Traditional Fraud Policies Further Exacerbate the Fraud Problem

Ramanand explains that retailers today are trying to remediate on average 1% in total fraud from all transactions, which can mean declining up to 15% of good transactions. If this is applied to the physical store, a retailer would be turning away one customer for every ten new customers. The same rules-based system, that can be fraught with human biases, is applied to digital transactions, where volumes are much higher. This is what further exacerbates the problem leading to the average decline rate of 10-15% on those genuine transactions.

Ramanand illustrates this:

“The reason for these transaction declines is due to the reliance on traditional rules-based systems by most retailers. If an IP address is detected as originating from Vietnam, and the item is being shipped to Mexico City at 2:00 AM, a rule instructs the system to automatically decline or reject this transaction. There may be valid reasons for such patterns, such as someone legitimately traveling or sending a gift to another person. These reasons don’t indicate any malicious intent or wrongdoing on the part of the customer.”

Signifyd’s Paradigm-Shifting Solution

Signifyd offers a paradigm shift in how retailers combat fraud. Unlike traditional methods that focus on a single transaction variable, Signifyd leverages AI and machine learning to examine thousands of variables in real time, enabling a comprehensive approach to fraud detection.

Leveraging data from thousands of retailers to analyze patterns of behavior, Signifyd can determine whether a particular pattern of behavior signals good or bad intentions. As the transaction unfolds, Signifyd intervenes on behalf of the retailer, confirming the legitimacy of the transaction. If a mistake occurs, Signifyd takes complete financial responsibility for any resulting liability.

Ramanand explains, “At Signifyd, our role is to assess transactions on behalf of the retailer and determine their legitimacy. When we confirm that the transaction is valid, the retailer receives full coverage from us, ensuring their protection against potential risks. This approach empowers retailers to achieve massive conversion rates on their top line while simultaneously safeguarding them from potential downsides.”

The Power of AI in the Face of Evolving Fraud Techniques

Fraudulent communications and more sophisticated phishing attacks are more prevalent today. Using Generative AI, cybercriminals increasingly employ more refined tactics that urge recipients to log in and complete a transaction. This is compelling retailers to adapt swiftly to protect their customers.

Signifyd’s AI-driven solution thrives on a robust feedback loop, constantly learning and evolving to identify new fraud patterns.

Signifyd’s systems are designed to effectively manage this challenge by analyzing thousands of variables in real-time during a transaction. The machine learning model is trained to recognize patterns in behavior, such as where the customer typically logs in, makes purchases, and ships items. Each transaction is assigned a risk-score in real time to identify potential fraud.

In addition, device fingerprinting techniques help to identify unique characteristics of the device used for the transaction and helps detect and prevent fraudsters who may be using stolen or compromised devices.

Ramanand stresses the focus is on the transaction behavior rather than identity. This approach enables Signified to accurately predict when a transaction may be suspicious: “We’re not looking at the individual identity of that person, but we’re looking at the behavior of transactions, and therefore we can predict with ease that this particular transaction is not looking good.”

Signifyd relies on a global intelligence network of data sources to supplement their data in share information about fraudulent activities. This collective intelligence helps identify and prevent cybercrime across different organizations.

Maximizing Retail Revenue with AI-Powered Insights

Beyond thwarting fraud Signifyd does help retailers optimize revenue streams by creating an improved customer experience.

As Raj explains, “We’re looking at thousands and thousands of variables and train models to be able to learn what is good and bad based on that outcome or that feedback loop we get, and by doing that, we can predict where something doesn’t look right because the machine has been able to look at so many different signals.”

In developing effective AI models, the significance of the feedback loop within transactional data sets is crucial for effective learning. Ramanand emphasizes the importance of this feedback loop and the protection guarantee they offer retailers if an error were to occur. This commitment encourages retailers to provide prompt feedback on any errors, which leads to rapid model training at scale.

In their approach, Signifyd takes a holistic view of transactions, considering a multitude of variables—like IP address, distance from IP to delivery, time of day, and thousands of other signals–during their model training. The feedback loop allows the company to more effectively distinguish between good and bad behavior patterns, enabling them to better predict when something appears amiss.

This comprehensive strategy contrasts with point solutions in the market, such as device fingerprinting or address lookup, which can provide valuable inputs but lack the holistic view necessary for accurate predictions.

The scale at which Signifyd operates, encompassing over 600 million wallets worldwide, is substantial. Two pivotal trends underpin this scale and growth.

The first trend revolves around the ever-increasing push into the digital realm, as people naturally gravitate towards online purchasing. The Pandemic accelerated this shift, prompting retailers, en masse, to venture online to explore strategies to boost sales frequency. Initiatives like “Buy online; Pick up in Store” gained momentum post Pandemic.

The current state of the economy revealed a second trend: a notable surge in fraudulent behavior among first-party buyers. These buyers, although typically regarded as legitimate customers, have exhibited buyer’s remorse and have expressed increased dissatisfaction with retailers. Ramanand points out, “Concert-goers buy tickets, attend the event, and then allege that the tickets were fraudulently purchased. These chargebacks are becoming more prevalent today. Similarly, customers receiving delivered items have falsely claimed the items were never received. The ease of return policies has contributed to a practice known as “wardrobing,” where customers wear products for a period before returning them within the return policy period.”

These disingenuous behaviors pose a challenge for merchants who struggle to differentiate between genuine and fraudulent actions. While it may seem simple to deny a purchase due to a customer’s history of returns, making such decisions becomes complex when considering these returns could be legitimate and the customer’s lifetime value could be significant.

Ramanand explains the nuances of analyzing behavior at scale while growing the retail business itself,

“The strength lies in the vast network of retailers at our disposal. Within this network, we gain valuable insights into buying behavior and patterns of customers. Naturally, we strive to enhance the customer experience by eliminating friction from their interactions with the retailer. However, it’s equally crucial to distinguish customers with high lifetime value from those who may not contribute significantly to long-term growth.”

For the retailer, the key lies in nurturing and expanding the existing customer base. Signifyd can play a pivotal role in identifying those valuable customers and guide the retailer in effectively allocating resources strategically to support their needs.

Nurturing a Culture of Transformation and Resilience

As businesses embark on the path of digital transformation, the urgency to combat fraud while delivering exceptional customer experiences becomes imperative.

Looking ahead, Ramanand points to the rising issue of account abuse in the retail space. As the economy deals with seemingly never-ending dance between increased interest rates and inflation, the uncertainty has spurred a notable emphasis on cost reduction. Signifyd distinguishes itself by enabling the retail to dramatically improve revenues, while minimizing the cost of ownership. Considering the broader context, the 10% to 15% of orders turned away for fear of fraud translates into a staggering $400 billion annually being declined in the economy.

“What our role is in the world today is to be able to continue to put that $400 billion back into the world.”

The automated solution Signifyd provides allows businesses to infinitely scale up their fraud and abuse prevention operations without the additional headcount. That, in itself, creates boundless opportunities, and reduces the financial burden on business.

2023, for Signifyd, is earmarked as a year of dedicated cost reduction for retailers, ensuring they can continue to scale and remain resilient in the face of potential account abuse. Signifyd is driven by a mission to empower fearless commerce, enabling retailers to not only think differently about how to effectively manage fraud, but how to optimize the business overall. Will this be the company to reshape the retail industry, ushering in much needed resilience and safeguarding the future of commerce?