How Credit Unions are empowering their lending and marketing teams with Gen AI

Gen AI can prove to be a powerful tool for institutions that struggle to keep up with larger institutions but don’t possess the resources, and time to match their throughput. 

Today’s micro case studies dive into how two CUs are improving their lending decision making and marketing efforts through incorporating Gen AI tools in their workflows. 

1) How Commonwealth Credit Union is improving strategic decision-making in lending through Gen AI 

When it comes to analytics, 61% of lenders find the large swathes of customer and lending data available in the market overwhelming and 73% report that their limited ability to leverage data impacts their competitiveness, according to research. 

This points to a significant gap between data sources and FIs’ ability to extract useful insights. Lenders that can understand their positioning compared to their peers in factors like delinquency, chargeoffs, and interest rates are better able to provide competitive products. 

Commonwealth Credit Union is particularly aware of how this impacts the firm’s ability to compete. “Our competition isn’t waiting weeks for data. They’re making decisions today on data that they got today,” said Jaynel Christensen, EVP, at Commonwealth Credit Union.

The backstory

Recently, the $2.5 billion, Kentucky-based CU decided to fill in this gap by integrating a tool by Zest AI called LuLu Pulse, which uses Gen AI to consolidate multiple data sources like NCUA Call Reports, HMDA, and economic data. This ultimately allows lenders to gain insight into how their products and services compare to their peers by querying the platform.

The recent integration of LuLu Pulse builds on the Commonwealth’s long standing partnership with Zest AI, through which the CU has also utilized underwriting resources and fraud protection tech. 

The biggest value prop for the CU is the turn around time and efficiency. 

“A manual underwriter for our organization is expected to underwrite in five to seven minutes, looking at 15-20 data points while probably answering a chat message, taking a phone call or talking to the person that’s walking behind them, Zest is making a decision in 2.4 second, looking at somewhere between 180 – 230 data points. That’s significant in the ability to determine risk, which has helped us be able to lend deeper and expand those that we can say yes to,” she said. 

The masterplan

The launch of Gen AI-powered tools have required firms to go back to board rooms and build policies anew. Christensen recalls that, in late 2023, the Gen AI adoption curve required the CU to construct policies around what the technology could and could not be used for. Since then, the firm has revised its policies in some respects and prioritized identifying the best Gen AI tools to be utilized within the firm. 

At the 2024 Governmental Affairs Conference (GAC), Christensen came across LuLu Pulse for the first time, and had been workshopping how the CU could leverage the tech for internal strategic and analytics improvements ever since. “Everybody is in a race to stay on top of AI and Generative AI technology. We’ve been very similar in that approach,” she said. 

While Lulu Pulse is not used for credit underwriting, it does help CUs with improving their strategic decisioning in board rooms as well as understanding the twists and turns of the market. 

Currently, the firm’s Finance and Data Analytics teams are using the tool along with its leadership.  

“We have been working with the lending team and challenging them that the next step we want to take is, how do we start diving deeper and making those tweaks to adjust the risk tolerance and be able to make adjustments in the underwriting decision scoring model quicker,” she said. 

2) How DUFCU’s marketing department is improving segmentation and targeting with AI

While Gen AI plays a significant role in content generation across the internet, it is unclear what part it plays in content creation for highly-regulated industries like financial services. But this may be changing. 

The backstory

For small CUs in particular, Gen AI’s potential to improve workflows and efficiency is much greater. 

“Smaller credit unions are resource constrained. There’s a desire to provide more personalized member engagement in the credit union space, but that runs up against having lean teams and limited resources,” said Vertice AI’s CEO Mitch Rutledge. 

One CU that is experimenting with how the new tech can enable it to expand reach and build a stronger marketing funnel is Duke University Federal Credit Union (DUFCU), which recently integrated Vertice AI’s copywriting tool called COMPOSE. 

For DUFCU’s Director of Marketing Jennifer Sider, purpose-built tools focused on the financial services space offer her a significant advantage over free Gen AI tools available to the public. It’s also better than the manual alternative of managing the whole copywriting process alone. 

“To be able to have more control of the content and create our brand identity, consistent messages, personalization, and the compliance piece,  it had all these different elements to it that would serve as, really like another part of the marketing team and getting us started with content,” said Sider. 

The masterplan

COMPOSE’s integration is already changing how Sider is allocating resources within the organization. “In the past, I would most likely outsource the content, and I still do a bit of that, but now I can do more in house. The content is already aligned with our brand, and that’s helping in the workflow, speeding things up, in crafting our messages,” she said. 

Before CUs start using COMPOSE, the tool first learns what the firm’s brand voice and identity is, either from a brand guide or the firm’s website. COMPOSE also stays updated with NCUA requirements, helping compliance and marketing teams ensure that their messaging aligns with changing regulatory stipulations. 

For DUFCU’s Sider, the integration’s success is going to be demonstrated by the turnaround time and volume produced by the marketing team, as well as the engagement with its messages. “It could be as simple as likes and clicks, it could also be inquiries, calls, and product adoption – so various metrics that we can use along the way to see if the messages are resonating with our members,” she said. 

Post-integration, Sider is focusing more on building messages that focus on member growth. She feels her efforts now may be more effective due to the tool’s ability to provide better segmentation and targeting. 

“The marketing team can prioritize delivering high-quality content that drives new member growth. COMPOSE is equipping us to elevate our standards of excellence, while streamlining our efforts, ensuring our acquisition campaigns are highly personalized, on-brand and efficient,” she said.