In the last year we saw the digital financial ecosystem explode. One of the hottest segments was Buy Now Pay Later (BNPL) with millions of new consumers using this credit option (yes, it is still credit orientated) for the first time. With that growth comes greater responsibility.
Incidents of fraud and consumer complaints rose alongside this massive growth. Fintechs and financial services firms have to find the right balance of friction while onboarding customers in a seamless fashion.
With an esteemed panel consisting of Philip Belamount, CEO and founder at Zilch, Mindy Davidowski, SVP Global Customer Acquisition Products and Platforms at American Express, Killian Brackley, CTO at Sezzle, and Bence Jendruszak, COO at SEON, the topics of establishing the balancing act of onboarding, how data should be used to target the right customer for your product, how the right technology reduces risks and manual reviews, the rewards of good behaviour and building loyal customers and learning lessons from the credit card industry were all on the agenda to be discussed.
The Last 18 Months
Todd Anderson, Chief Product Officer at LendIt Fintech kicked off the conversation by asking the panel how they had found the past 18 months. Brackley was the first to answer, stating, “I think one of the things we learned in the last 18 months, is how we tune in with security communities to really understand security threats consistently, and I see the fraud community being very similar. They use very similar tactics… so if you’re looking at more sophisticated fraud rings, really you have to look to everyone in the space as being a part of a community.”
A theme that became abundantly clear when this question was posed to the panel was the need for automation. Davidowski said, “We’ve been trying to balance three things: one is continuing to strive for even more automation at the point of decisioning, and application without sacrificing any credit or fraud loss risk increase. As the servicing team experienced a spike in demand we needed to free them up to focus on other pockets of our customer life cycle.” Manual intervention from the team was what wanted to be avoided.
Belamount said the main focus in the last 18 months for Zilch had been balancing the product they were providing. “The challenges really to balance what product you’re providing to a customer, how responsible that product is for the customer, is the customer who they say they are? This is what we tend to find in our business: are you who you say you are and is this token you provide us yours?”
As he continued he acknowledged the growth they had seen in the last year and a half but stressed “This balancing act between onboarding people at speed and making sure they are who they say they are and giving them the appropriate amount of money to defer, is definitely something we’ve spent all our time worrying about.”
Partnerships and Risk Consideration
Belamount concluded his thoughts on the last 18 months by establishing that during the pandemic, “microservices have certainly helped us navigate this time, and then we can plug providers in and out appropriately. We can make use of a combination of these providers to make the best decisions for the customer.”
When asked how partnerships are used within Sezzle, Brackley said, “We lean on partners to help supplement data and run models… You’re not going to see fraud, either first-party or third-party for 90+ days… With our product, we’re not going to know for a while so we need to be able to run those [models] at least from a customer perspective, all the time. run the tests to see how a decision would have been made, and then look back at it in 90 days and see where would this have helped?” This is all done in an attempt to reduce the amount of false-positive cases of fraud.
Jendruszak also added to the discussion on working with partnerships when the need for two-factor authentication, and higher forms of verification were being demanded. “Consumers want everything immediately nowadays.” The decision making of partners is crucial, so the speed at which they can help the company is extremely important.
As the conversation turned to risk consideration, Belamount spoke about how “in the pursuit of understanding if this really is you, and if this is in fact your car, there are different things that you can do in order to prove it.” Essentially stating that the old fashioned way of identifying someone was evolving.
Onboarding With the Right Amount of Risk
Davidowski introduced the idea that customers actually expect some sort of friction when applying for these sorts of things and actually understand the need for it. “With some of our seamless acquisition journeys, we’ve observed in user testing, that when a journey starts to get too easy or streamlined, our applicants start to get a bit nervous and feel uneasy.” She went on to establish that users need to understand why you want information and make it clear that there needs to be some friction in order for the customer to feel secure giving up their details.
In response to this Brackley said, “what we do is promote different options for customers to continue [their journeys]. So we have different processes, and we promote different ones that we believe will get them higher scores on those attributes, but ultimately I think there are a number of different ways that you can get your risk score bumped up so that we’re comfortable approving your account and transaction.”
Synthetic Identity Fraud
Using social media checks and other forms of verification, Jendruszak explained how synthetic identity fraud was combated. “They set up an email address or phone number that would look to mimic an everyday consumer, then we’re able to easily spot that through our social media checks, through our use of HLR based on phone numbers which actually pings the number to see if its virtual or not. There are a variety of tools which enable us to help merchants and operators decide whether that consumer is trying to fake an identity or not.”
Belamount echoed this thought saying there is “a combination of checks that help solve the problem.” Like, social media checks and phone pinging. Belamount explained how information was so accessible in the UK through junk mail: people’s full name, address, date of birth and often other details are handed to criminals as they can search through junk to find someone to impersonate. He went on to say that there must be a combination of things to prove one thing matches the other to prevent synthetic fraud.