top of page
Writer's pictureJohn San Filippo

2022 Tekkie Award for Artificial Intelligence: Lebanon Federal Credit Union

By John San Filippo

Lebanon Federal Credit Union had a goal to increase the number of cross sales among existing members. The $400 million, 30,000-member credit union, based in Lebanon, Penn., decided to focus on its digital users by providing relevant and personalized offers that would be meaningful for members. To accomplish this goal, the credit union used a new type of campaign utilizing artificial intelligence (AI) within its digital marketing platform DeepTarget DXP called Target by Predictive Model. These targeted campaigns use machine learning (ML) models to identify the top users with the highest propensity to purchase a specific deposit or loan.


The offers are then presented to these selected top users of online and mobile banking as they use these banking services. Predictive campaigns are essentially automated targeting methods using AI, built on years of advertising and demographic insights, enabling community financial institutions to easily deploy techniques and insights previously reserved for only the largest financial institutions.


The credit union had already been a DeepTarget customer for a number of years when Marketing Director Alaina Smith was presented with the opportunity to deploy this AI-based enhancement to DeepTarget’s DXP target marketing platform. According to Smith, deploying the AI enhancement to DXP was a natural progression.


An Easy Decision


“DeepTarget makes it easy [to deploy] new products and new enhancements to the platform,” Smith told Finopotamus. “I always try to implement the new stuff and see how well it works.” She said, adding that she has yet to be disappointed.

Alaina Smith

Smith said that by leveraging AI-based predictive analytics, DXP now takes a lot of the guesswork out of target marketing. “In the past, you would create a role,” she said. “For example, maybe I’d want to target a checking account for members over the age of 18 who don’t have a checking account, but do have an auto loan – something like that.”


Now Smith just needs to identify the product. “The predictive analytics model identifies which members are most likely to want that product,” she explained.


The credit union launched its first set of Predictive Model campaigns in March of 2021 and saw immediate success. Then the credit union added more campaigns, reaching a total of seven predictive model campaigns across seven different product categories by May 2021. These seven campaigns ran through December 2021, resulting in 337 new accounts openings over a nine-month period. These campaigns targeted members for auto loans, checking accounts, credit cards, consumer loans, mortgages, IRAs and home equity loans, Smith noted.


A Win-Win for the CU and Its Members


“I like to run campaigns that aren’t necessarily product driven, but more community driven,” said Smith. “I don't want to overload members with product information.” She said that while DXP lets her identify members most likely to want a product, the fact that it excludes members not likely to want a product is also beneficial.


“If we present products to members who might not be interested in those products, they’re more likely to become annoyed by tour marketing,” she said. This approach makes for an overall better member experience, she added.


Plug-and-play AI-based predictive campaigns present the most relevant offers to individual members with the highest propensity to purchase specific financial products with virtually no effort. The Predictive Model campaigns save the marketing team time by allowing them to simply select the product type and an age range if desired and directly upload a relevant ad image for the campaign. Predictive model campaigns go to work engaging selected members for that campaign and even updating the target audiences as member data changes.


Leveraging a Treasure Trove of Data


“Maybe two years ago, we realized that we had collected all of this information from the results of our customers,” DeepTarget CEO Preetha Pulusani told Finopotamus. “We decided to use AI and machine learning models to figure out if we can identify consumers that have a propensity to purchase a specific product.” She added that the answer was a resounding: “Yes.”

Preetha Pulusani

Pulusani noted that in the absence of AI, credit unions would need to spend a considerable time and effort developing rules-based campaigns. She said the amount of time saved by relying on AI to do the heavy lifting depends on the credit union.


“It depends on how often they're refreshing their campaigns,” said Pulusani. “We’ve seen weekly updates of campaigns to quarterly updates to annual updates.” She said that the data that’s driving DeepTarget’s AI engine is updated daily, so the company recommends weekly campaign refreshes.


“If you're focused on making sure that your campaigns are refreshed frequently, AI saves a lot of time,” she added. “Marketing departments are severely under-resourced. They have so much stuff going on and very little staff. Now they don’t have to think through the rules for each campaign.”


Helping Digital Transformation


“This is absolutely the first step in something much bigger,” said Pulusani. “Everybody talks about two things: digital marketing and data. The volume of data is increasing exponentially, but using that data to run campaigns is another matter.”


She added that to drive digital transformation at the institution level, the technology must be as easy to use as possible for the employees. It also needs to be accessible to smaller institutions. To this end, she said that DeepTarget is now collaborating with a much larger AI company.


“All they do is AI, and they've been doing it for different sectors,” she said. “And for very large enterprise customers, like JP Morgan Chase and United Healthcare. It's going to be a great collaboration because they've been trying to figure out how to take all of this technology and provide it as a service to smaller customers.”


She continued, “We are working with them and trying to simplify our technology so that we can actually go to our customers and say, this month, here are the product campaigns you should run. And here is a population for each campaign and these are the results you will get if you run it.”


Effective and Trackable


“I think AI is the way to go,” concluded Smith. “A lot of marketing is word of mouth, but when you have something like this and you can track it and it's targeting people who already have the interest, you can see the results and how it’s affecting your members. We’re really getting the most out of our marketing dollars now.”

Komentar


bottom of page