The financial industry has experienced an exciting digital transformation in recent years. Most notably following the pandemic, where financial organisations had to work harder than ever to compete with FinTechs to offer time-saving digital onboarding and seamless mobile experiences.
Whilst this digital transformation presents many benefits, including improved accessibility (with regard to both availability and inclusive designs and build), optimised operations and increased opportunities for personalisation, it also presents significant fraud risks. Indeed, fraudsters have continued to find ways to outfox fraud detection measures and exploit this proliferation of digitalisation in banking.
Digitalisation has also transformed consumer habits. In a 2020 Signicat report, the company found that 68% of consumers expect 100% digital onboarding and 32% are refusing to start an application if they are required to bring identification to a physical branch. As the digital age brought about the desire for ‘instant gratification’, customers expect to apply for, and interact with, a bank anytime, anywhere, and very quickly.
As customers continue to demand simple, frictionless experiences and digital banking becomes the norm, how can lenders provide excellent customer journeys whilst mitigating fraud?
We asked this question to three Jaywing experts.
Malcolm Clifford - Customer Experience Director
“Providing better customer experience whilst also preventing fraud is complex. But not for the reasons you may think. The truth is that user experience (UX) and risk teams are often working in silos with their own unique objectives. On one hand, you have the UX team wanting to attract more customers and simplify the user journey and, on the other, you have risk teams working to detect and prevent fraud losses. The reason why the question is complex is because this conversation has been dominated by the notion of “UX versus fraud" - pitting the two against each other and looking at each in isolation. The conversation needs to pivot to a more collaborative approach from the start – “UX and fraud: how these teams can work together”.
It’s a mistake to judge a fraud strategy solely on the amount of fraud that is identified. The negative impact on the customer experience is a cost as much as the failure to find fraud. This will be measured by poorer conversion at account opening, customer dissatisfaction, resolving complaints, and over time increased attrition. Organisations should review their fraud strategy with UX and customer experience (CX) teams to assess the impact it has on customers, analyse where it is creating the most friction, and question: is the friction necessary? Can small amounts of fraud risk be traded for large improvements to the customer experience and vice versa? There is a compromise to be found and a ranking approach (i.e. prioritising the importance of different fraud and customer experience outcomes) can be a useful way of identifying the best trade-offs. This collaborative approach will help organisations find that sweet spot where they mitigate fraud risks and provide customers with a seamless, positive experience.
What is clear is that the overall customer experience is at the very heart of it, as delivery of an excellent service is critical. Where additional process is necessary to mitigate fraud risk, it should be tailored to be as non-invasive or delaying as possible and delivered alongside messages which explain what is happening and why. Good messaging borne from customer empathy can even turn some level of friction into a positive outcome if the overriding feeling felt by the customer is one of being protected from fraud rather than being inconvenienced unnecessarily.”
Nick Sime – Head of Modelling
“Fraud is extremely expensive, not only in terms of monetary losses but also for reputational damage. Many banks prioritise fraud prevention as they believe that adding friction to the customer journey, and perhaps losing some customers along the way, might be cheaper in the long run. However, we have noticed that organisations are becoming increasingly focused on minimising friction within their application processes and automating lending decisions. This certainly adds to the challenge of making approval models and processes as slick and powerful as possible.
Despite this, I believe fraud prevention can actually benefit the customer experience. Just as technology has advanced customer expectations, it has also enhanced fraud detection and prevention capabilities.
Models built from artificial intelligence (AI) or machine learning (ML) software are allowing banks to predict fraud risk more accurately than ever before and, thus, allows customers to quickly and simply complete processes or transactions. These models are far more easily updated and refreshed than traditional scorecards and linear approaches, so also allow banks to keep up with emerging fraudster techniques and update their system as soon as a threat emerges or even on a reactive basis, if needed.
The speed and accuracy of these AI techniques are essential to early fraud detection. Detecting and preventing fraud early on will allow you to invest more time in giving your genuine customers the best experience possible. Indeed, using this advanced technology to proactively fight fraud is key to improving customer loyalty and ensuring both your team and customers feel secure.
That being said, even the most effective fraud detection models are not 100% accurate. Risk teams should prepare strategies for responding to ‘false positives’ or ‘good’ customers who may have had their user journey negatively impacted. In the first instance, the risk team should understand why that false positive came about. This is often a problem with AI or ML models, where teams encounter the ‘black box’ issue in which mechanisms in the model that lead from input (eg, customer activity) to output (eg, customer flagged as suspicious) are opaque. We solved this problem with Archetype, our AI fraud software, that builds fully controllable and explainable models to identify high risk behaviour and also provide reason codes to summarise the activity that led to the referral. In cases where a false positive has been generated, the team must work with marketing and UX experts to respond to these customers and provide them with appropriate, empathetic communications which they can do on an informed basis.”
David Rubery - Head of UX
“In an increasingly competitive market, finance brands are grappling with the commercial need to provide a best-in-class user experience and the moral and legal requirement to protect customers from fraud. I often hear that these objectives conflict, but I don’t believe that has to be the case.
The word "friction” has become synonymous with UX; “it’s the job of the UX team to remove friction from the journey". I often find that people mistakenly associate reducing friction with reducing complexity (i.e. removing things). Friction in UX is where two things interact that are not in alignment; in this case the human being and the software / AI provided by the financial organisation.
So, in 2022, do we need to redefine “friction” in the context of online fraud mitigation?
The answer is of course, yes. Removing friction isn’t about reducing complexity (being less intelligent), it’s about using data and research to understand customers motivations, needs, fears and real-world pressures and then designing experiences sensitive to those insights. Reducing friction with fraud prevention isn’t about “how do we make this simpler”, it’s a matter of “how do we explain to the customer that we have shared aims”.
For instance, innovations such as facial recognition and three factor authentication (3FA) shouldn’t automatically be treated by UX teams as “friction”, they should seek to explore (research/test) how to present this in a context users can intuitively process and interact with. The good news is, customers also don’t want to be defrauded, so there is no actual conflict of interest; the conflict comes from an implementation that is software-biased, as opposed to human-centred.
There is no ‘silver bullet’ answer to ‘how can lenders provide excellent customer journeys whilst mitigating fraud?’’. It is a case of good process; deploying behavioural research and meta experimentation analysis to understand where software and users are misaligned; experimenting with levers such as semiology, trust signals, authentic imagery, social proof and system feedback. A simple example can be seen in a project undertaken with Insurance2Go, where our behavioural science team increased conversion rates by 34% by using AI image analysis to predict the most persuasive image motivate users to trust and uptake insurance with the brand online.
The big risk to financial organisations online, is that customers are not aligned with their attempts to protect them. My final advice would be to treat your fraud strategy as a risk project, brand project, UX and CX project, AND content project. Your aims and customer’s aims are not misaligned, only miscommunicated, so to protect yourself and your customers in the best possible way, you need to be user-centred (outside in) in your approach to designing solutions”