Artificial intelligence as a playing field for credit unions

On November 30, a panel discussion was conducted with a focus for credit unions addressing “Assessing risk and optimizing growth for each member”, hosted by Neuton.AI. This event brought together thought leaders from the industry who shared views on how credit unions can uncover new growth opportunities and mitigate risks by leveraging AI.

Needless to say, the pandemic has caused a seismic shift in how we interact with customers such as how members are now expecting to consume services, their digital expectations which have in turn forced credit unions to rethink the way they interact and respond to member needs. This has subsequently led more and more credit unions to adopt a more data-driven mindset while leveraging innovative technologies such as artificial intelligence or machine learning. Beginning this journey, institutions are faced with a number of challenges such as where you begin, why data is important, what the possibilities are, and how I complete this journey when I may not have the resource or financial capital that is historically required to implement such services. The panel takes these topics head-on providing practical and ready-to-apply solutions to these challenges.

The panel discussion brought together Anne Legg, Founder of Thrive and Author of Big Data/Big Climb, Naveen Jain, Founder of CULytics, Jay Lauer, Senior Innovation Strategist at PSCU, Todd Lindemann, Senior VP of Payments at Redwood Credit Union, Michael Lawson, Creator at CUbroadcast, and Blair Newman, CTO at Neuton.AI, all well-respected professionals in their field of expertise.

Post-pandemic challenges for credit unions

Michael and Blair kicked off the event with a brief fireside chat where they began to discuss and note the challenges that credit unions have faced in light of our new pandemic reality. One notable realization is that members have begun to consume services differently than prior to the pandemic, which has forced credit unions to adjust quickly. The new realization is that it has become more challenging to engage your existing customer base let alone continue to drive growth with new members or additional services. The current state of affairs indicates the need for alternative and innovative ways to service your members as the need for enhancing the customers’ digital experience has come crashing to the forefront now.

“We see that credit unions are at different stages of their journey. Some customers are looking to explore how they can leverage a forward-leaning technology such as machine learning, some are looking to debunk some of the myths about the ability to implement ML solutions, some are ready to begin their journey and some are well on their way. One of the things that we’ve really focused on is eliminating all of those barriers.” — recapped Blair.

Naveen also mentioned that, based on his experience, the majority of data and digital initiatives simply fail to deliver on the promised value due to the lack of understanding of what problems they should solve. “Our mission is to help credit union leaders be successful with their investments in data and digital, with the focus on events, a vibrant community, management consulting, and complimentary advisory services.” — added the founder of CULytics.

Barriers on the way to innovation

Speaking of innovation, Jay Lauer highlighted that credit unions don’t have to invent anything revolutionary but simply adapt to new ways of thinking and doing business. On the other hand, they need to figure out how to go faster and bring innovation to production quickly and effectively. “Nearly 85% of credit unions feel that AI is critical to long-term success. At the same time, however – over 70% feel they are behind when it comes to adopting such solutions”, said Jay. He also mentioned three reasons that hold credit unions back at the moment:

• Lack of focus on the data strategy and effective data management. As data is the lynchpin for both AI strategy and execution, credit unions need to eliminate data silos, shine a light on dark data, and adopt data stewardship principles. Such practices help to build trust in the credit union’s data and in how that data may be used.
• Lack of organizational readiness. Many credit unions feel they have a limited understanding of AI potential and have trouble with use case identification. Educating credit unions on these aspects can be a big step towards their success.
• Poor execution, primarily centered on a shortage of talent or availability of AI technology. Credit unions have to attract talent and up-skill the talent they already have, but they should also be looking at how technology may help them in this area. They could explore low-code or no-code solutions or look at other platforms which might provide some additional capacity for the team they already have.

One more important finding for credit unions, according to Anne Legg, is understanding where member friction exists within a credit union’s ecosystem and building a roadmap to reach the full-enterprise capability. “We see an uptick in critical thinking which enables a credit union to move from a reactionary position to a proactive position, take the right insights and do things with them” said Anne.

Practical use of AI/Machine Learning in Redwood Credit Union

Redwood Credit Union is a model of not only embarking on their data-driven AI-powered journey but also bringing innovation to life, Todd Lindemann shared their experience in making the member experience more satisfying while also realizing organization value.

Firstly, it’s crucial to understand where you are with your data. Once the Redwood team figured out their situation, they started to use data on a daily basis by regularly asking members for feedback, implementing critical thinking, and infusing data into all business processes.

One of the directions that Redwood has taken is moving to a more cashless environment, exploring individual member needs with the help of an automated machine learning platform. They implemented an AI-based predictive member risk model as a baseline service to understand which members were considered high risk and which members were considered low risk. This model was built without the consideration and/or use of the FICO score and was developed strictly on individual Member behavior. Leveraging this baseline predictive model, Redwood CU was able to apply this risk rating model to a members’s ability to withdraw funds during an ATM transaction. Lower risk members would be enabled with higher ATM withdrawal limits above the default limit as well as the ability to boost their access to cash dynamically upwards to $10,000. Implementing such a forward-thinking solution reduced the amount of times members were effectively reaching their static limit which oftentimes forced them to either go inside the branch or contact the Member Service Center. This was the first step in improving the customer’s digital experience, providing more efficient access to cash resulting in improved customer satisfaction.

The member risk rating model is an innovative approach and the first step for Redwood to continue to bring additional value to their members by implementing additional solutions such as implementing dynamic overdraft limits, credit scoring, and member lifetime value just to name a few. More importantly, Redwood is empowered to implement such solutions without a data scientist and realize results in days rather than months.

Conclusion

Anne Legg concluded that the key value of credit unions is their mission-based approach as they make great efforts to build member-centric relationships. As the post-pandemic realities continue to impose their own will, credit unions should amplify their strategies of personalization of products and services, which can be reached faster only by using AI innovations and handling data appropriately.

It is vital to establish data management strategies, remain active in like-minded communities, and not be paralyzed by the use of transformational solutions such as no-code AutoML platforms like Neuton, which provides enormous opportunities to uncover granular insights to drive personalized services, enhance the member’s digital experience while driving value for all credit unions irrespective of their size or point in time in their AI journey.