Let’s say Sam, a Bank of Someplace customer, has his heart set on Tesla, model 3. He is currently shopping for auto loans to bring his dream car home. He goes on his bank’s website to look up interest rates, but does not start a loan application. Obviously, he is still shopping around. Next day, his 5-year-old daughter’s future is top-of-the-mind priority for him. He again goes to his bank’s website to read up about 529 College Savings Plans. He even calls the bank to make an appointment with a wealth management person to discuss this further. However, the schedules don’t match and he doesn’t end up making the appointment.
A few days go by and Sam has still not taken any action on either the car loan or the 529 Plan. What should the bank do when he next comes back to the website or visits the local branch? A reminder would be good. But a reminder for what — for the auto loan with an amazingly low interest rate, or a great incentive to sign up for the college savings plan? Which is the best offer for the bank to put forward that will nudge Sam to complete one of the two transactions? And when should the bank send this offer to Sam to yield the best possible results?
To make this determination, the bank has to consider multiple factors to determine Sam’s intent and needs. This can be done by reviewing what Sam is doing now and what he has done in the past. Some considerations could be:
Banks are faced with such situations multiple times a day, from numerous customers, on a variety of touch points (branch, website, mobile app, call centers). They are inundated with data about what the customers are doing now and what they have done in the past. These data are stored and analysed to come up with insights and patterns for banks to take action on … send an email or a push notification, maybe a day later.
Considering the large transactional volumes and user activity, it is next to impossible for banks to make a determination about which information best suits the individual needs of every customer with traditional analytical tools. Such personalized, individual level decisioning and offer orchestration at scale has to be automated and lends itself well for the use of machine learning and artificial intelligence. Banks, too, realize the importance of machine learning and AI. According to the research done by Digital Banking Report, 35% of financial organizations have deployed at least one machine learning solution. Of the organizations that have not yet embraced this technology yet, 23% said they believe they would have an AI solution in place in the next 6 months to a year, with another 13% believing they would have a machine learning solution in place within 18 months. While 17% indicated a machine learning solution was on their roadmap in the next 18 months.
However, according to a recent article in The Financial Brand, Banking Providers Must Leverage AI and Machine Learning (But Aren’t), so far many of the machine learning implementations have been in the areas of credit scoring, fraud, and security. While machine learning works really well for security and fraud use cases, adding personalization to this repertoire of use cases is what is needed for banks to enhance customer experience and boost revenue growth.
Jim Marous, co-publisher of The Financial Brand and owner/publisher of the Digital Banking Report, emphasized in the article that, “Machine learning has potential to make banks exponentially smarter. “Smarter” in this case means delivering better customer insights and intelligence, and thus a better customer experience — something most in the banking industry now believe is the key to differentiation, growth and increased profits.”
Great customer experience is also critical in encouraging usage of banking services and deepening loyalty in today’s Millennial population. This segment has the lowest level of engagement with banks compared to other generations and are more likely to switch banks if they are not happy with it. This generation is always online (desktop or mobile), with only 66% of them visiting the brick and mortar branch of their bank. They are also not inclined to wait on the phone to get the answers they need from their bank and they demand instant information. Considering that the Millennials will make up more than 40% of the workforce by 2022, it is essential for banks to offer the digital solutions they want to manage their money and provide contextual, in-the-moment engagement to win their business and loyalty.
This brings us back to the importance of adopting machine learning and AI-backed solutions to offer 1:1 interactions at scale. This need cannot be fulfilled by existing analytical tools that focus on past customer behavior to predict the future. This requires a new generation of technology that will allow banks to manage user experience through a range of possible actions and interventions based on the user’s context. This context needs to be built from user activity tracked on digital channels, as well as from customer attributes, historical transactions, location information, etc. Additionally, machine learned customer preferences such as user’s preferred day / time and channel, along with models based on event sequences and customer attributes can contribute to the intelligence that drives actions and interventions best suited for individual user.
We have seen our customers achieve 5x increase in customer engagement through such contextual and personalized responsiveness, compared to vanilla messages sent to all. Going back to Sam’s use case mentioned earlier, suppose he visits a Tesla dealership. His bank is alerted about it because of the breach of geofence that the bank has set up in all Tesla dealerships. Sam’s repeated browsing for auto loans plus this event at the dealership gives a clear indication that Sam is very interested in buying the car. At this point, a notification from the bank about a great rate for buying electric cars could definitely clinch the deal and nudge Sam to fill out the auto loan form.