It's a fine line between pleasure and pain

The role of friction in preventing fraud?

Digital payments have transformed the way we transact with businesses and financial institutions. With just a few clicks, we can pay for goods and services from the comfort of our homes or on the go. However, with convenience comes risk. Fraudsters are always on the lookout for opportunities to exploit weaknesses in payment systems to steal from unsuspecting victims. The scale of this challenge is monumental, in-fact PwC's Global Economic Crime and Fraud Survey 2022 respondents reported total losses of US$42B. This challenge creates a dilemma for businesses and financial institutions, who must balance the need for security with the need for a frictionless payment experience for their customers.

Businesses and financial institutions have tried to combat fraud by imposing additional security measures such as two-factor authentication (2FA), CVV codes, and CAPTCHAs. While these measures can help prevent fraud, they can also create friction in the payment process. For example, requiring customers to enter a one-time password or answer security questions may add an extra layer of security, but it can also cause frustration and lead to abandoned transactions.

On the other hand, providing a frictionless payment experience can make transactions seamless and easy for customers, but it can also leave payment systems vulnerable to fraud. Fraudsters can use stolen credit card information to make purchases without the need for additional security measures, leading to chargebacks and financial losses for businesses and financial institutions.

So, what is the solution? How can businesses and financial institutions find the balance between friction and fraud prevention in payments?

One solution is to dynamically apply friction where perceived fraud risk is high. This technique allows businesses and financial institutions to leverage advanced technologies including data streaming, artificial intelligence (AI), machine learning (ML) to dynamically apply friction where perceived fraud risk is elevated. These technologies can help businesses and financial institutions detect and prevent fraud in real-time, limiting added friction to the payment process.

Data streaming, as offered by Confluent, enables real-time processing of not only transaction data, but also contextual data from a variety of sources (location, device, past transaction behaviour, website interaction data). Importantly, these diverse datasets have previously been either in the domain of cyber OR fraud teams. Bringing these datasets together enables a broader view of risk to be assessed, the moment login is attempted through to an attempted transaction. Such capability enables organisations to immediately detect and block transactions exhibiting risky traits, from a diverse set of real-time signals. The real-time nature of streaming also means that watch-list accounts can be immediately actioned for subsequent transaction attempts. Ultimately, data streaming enables organisations to make more informed decisions and take proactive measures to prevent and mitigate fraud. Confluent’s ebook “Putting fraud in context” explores these capabilities in detail.?

AI and ML can be used to analyse transaction data in real-time to identify patterns and detect fraudulent behaviour. By analysing a wide range of data points such as device location, user behaviour, and payment history, businesses and financial institutions can quickly identify and flag suspicious transactions before they are completed. This approach not only helps prevent fraud but also provides a seamless payment experience for customers.

Leveraging such techniques enables businesses to dial up or down friction to align with perceived risk. This approach uses a variety of factors such as the device used, location, and behaviour, determined in real time, to assess the user’s behaviour to evaluate the risk of a transaction. By using risk-based authentication, businesses and financial institutions can provide a frictionless payment experience where risk is low and insert an appropriate level of friction where risk profile is higher. All of this can be accomplished in real-time targeted at the most likely sources. The way businesses respond to fraud, both with impacted individuals and improving risk posture to reduce subsequent instances, has a significant impact on NPS (See McKinsey study “A new approach to fighting fraud while enhancing customer experience” Exhibit 1

In conclusion, the balance between friction and fraud prevention in payments is a complex and ongoing challenge for businesses and financial institutions. However, by leveraging advanced technologies and implementing risk-based authentication, businesses and financial institutions can strike the right balance between providing a seamless payment experience for customers and preventing fraud. Ultimately, finding the right balance will not only protect businesses and financial institutions from financial losses but also improve the overall payment experience for customers.

To find out more about Confluent’s fraud capabilities, tune in for our global webinar “How to Build a Context-Aware, Real-Time Fraud Detection Solution in Confluent”?