By Ray Birch
EAST LANSING, Mich.–With a liquidity crunch facing financial institutions of all kinds, one credit union said it has been able to use artificial intelligence to boost CD deposit growth by 1,000%.
As a result, the $7.4-billion CU said it now plans to use AI to raise performance of several other products.
“We have been using Trellance’s data analytics platform for the last few years,” said MSUFCU Chief Financial Officer Sara Dolan. “One of the components of that solution is predictive models, which integrate into that platform. We reached the point where we were ready to engage with those models earlier this year.”
Dolan said the move meshed well with some new marketing campaigns that were scheduled, including one for certificaters.
As CUToday.info has extensively reported and as just about every credit union is aware, there is an ongoing war for deposits, as financial institutions struggle to find dollars to fund lending.
The Next Level
“Historically, we've done some filters of our membership to be able to personalize offers and provide marketing context that fits the members’ situations,” Dolan explained. “But we needed to take things to the next level.”
Dolan explained the Trellance predictive models look at transaction data, demographic data, and behavioral data over time, and then generate insights based on that information.
“Those AI decisions give us the next best product recommendation for each member,” Dolan said. “We were able to layer that into this recent certificate campaign.”
Those insights led to a 1,000% increase in CD dollars attracted with the latest campaign compared to the previous certificate effort, which ran last year.
Sara Dolan
Testing is Conducted
"To test the effectiveness of Trellance’s Next Best Product AI, we conducted A/B testing on a group of members selected by the Next Best Product predictive model versus a group of members selected using MSUFCU’s traditional selection criteria,” Dolan explained.
Testing included sending the two groups the same promotional materials, a mailer and an email campaign, as well as in-app promotions, promoting their certificate of deposit program.
The group selected by Next Best Product AI had an average account balance of $90,000 and opened 160 CDs with a total revenue of $7.4 million. The group selected by MSUFCU had an average account balance of $7,000 and opened a total of 53 CDs with a total revenue of $666,000, Dolan said.
Instant ROI
“The instant ROI from this has been amazing,” said Dolan. “I think running the A/B testing really allowed us to see the value of the Trellance product. We’re making plans to include this sort of predictive modeling in our future marketing efforts now…We're seeing significantly better results using these AI models.”
Dolan compared the AI-led marketing efforts to its traditional targeted marketing campaigns.
“We used to do campaigns that were personalized, but not to this extent. This is like super-targeting and it certainly the next level of personalization,” she said.
What Dolan believes makes the AI directed efforts very effective is AI’s ability to analyze data over time and essentially “learn” each member, and their needs.
Dolan explained that the AI solution’s ability to “learn” about members isn’t an instant process, and may require a few months to gather enough data and then to really get to know each member. IN short to "build up history.”
Learning Patterns
“The AI learns the habits of your membership, which does take some time,” Dolan noted. “It builds patterns of behavior. The solution connects into our data warehouse—it's seeing transaction data trends and engagement trends over time, and that better allows it to create an engagement score and Next Best Product score.
“I am not saying that in the past we weren't good with our marketing efforts, but we simply wanted to go to the next level,” Dolan said. “Some of this was testing, just to see what we might be able to do with AI. And I can say we are extremely pleased with what we have found.
“This was our first test of the Trellance solution,” continued Dolan. “And we are always looking for ways we can improve how we make offers to our members. We have always done A/B tests, and I have to say this is the most significant difference in results we have seen. This was our first attempt, and obviously we hit a home run here. So, now we're looking at where we can incorporate this AI solution for other product campaigns.”
Final Point of Emphasis
Dolan emphasized the credit union is just at the very beginning of its data analytics journey.
“My goal is to build effort on an ongoing basis to take it eventually across all of our marketing campaigns,” she said.