By Ray Birch
METAIRIE, La.—Providing financial counseling to the underserved is nothing new for credit unions, but using artificial intelligence to do so more effectively certainly is, says Judy De Lucca.
With help from a grant from Treasury’s Rapid Response Program (RRP), New Orleans Firemen's FCU has just begun using AI to identify accounts with slow payments, NSF/overdrawn activity, and payday loan payments so it can intervene prior to a potential bankruptcy risk.
The Community Development Financial Institution (CDFI) serves a large underserved market in an impoverished area of Louisiana.
The $226-million CU has also begin using AI to identify members with sky-high car payments from title loan and finance companies to get them into a credit union loan at a much lower rate.
“What we do now is look at our transactions each night and sweep through them, searching for anybody with NSF overdraft, late payment, members with payday loan payments…,” explained De Lucca, NOFFCU CEO. “We then get those members’ names to our Financial First Responders team so they can reach out and schedule counseling sessions.”
A Team of Five
NOFFCU’s Financial First Responders are a team of five who are trained to be good listeners and counselors, helping members whose financial lives are headed the wrong way, De Lucca explained. During counseling sessions Financial First Responders review members’ banks statements, credit bureau reports, FICO scores and other information. NOFFCU also has 60 certified credit union financial counselors on staff in both the front and back office.
“We help them, for example, understand what a FICO score means, how it goes up and down. We review their cash flow and do a budget analysis and help put them into credit union products that can help their financial lives,” explained De Lucca, whose organization was named this year by NAFCU as Credit Union of the Year in the $250 million in assets and under category in part for its work in serving low income communities.. “What we do a lot of is get members out of the payday loan trap. These people don’t have generational wealth. They don’t have a great deal of financial knowledge or even access to financial education outside the credit union, and we do our best to help them.”
But to do that, to identify those members who are getting into financial difficulty can be time-intensive, noted De Lucca.
Not Fast or Sexy
“What we do is kind of old school credit union, and it's not the fastest or sexiest thing,” said De Lucca. “We hope to do this more efficiently with the use of AI to target these members, as opposed to using staff. That will not only make us more efficient, it will extend our ability to help more of these members who are getting in trouble. Through the use of advanced technology we could get to members sooner, identify them faster, and really help them in their lives. That's what we're supposed to be doing as a credit union. We're supposed to be in their lives helping them, not just like the bank around the corner that does not care about them.”
Among the results from the Financial First Responders initiative has been strong date. When the credit union gets to members who are in need, bankruptcy filings go down, De Lucca said.
“We saw a 20% decrease in 2020 over 2019 in the number of our members filing for bankruptcy,” said De Lucca. “Members in the program, too, generally see their FICO scores rise 25 to 30 points.”
De Lucca acknowledged paying for the technology to find financially troubled members faster and more efficiently will be an additional expense.
“But so is all the time our staff puts into looking through transactions,” she said. “That is a lot of salary expense.”
The credit union worked with Tacoma, Wash.-baed CU Strategic Planning to obtain the $1.826 million RRP grant. As CUToday.info reported here, some 244 credit unions in total received $401.8 million in awards from the Treasury Department’s new Rapid Response Program (RRP) as part of $1.25-billion in COVID 19 relief funds that are going to 863 community development financial institutions (CDFIs) overall.
The CU this summer was also struck by Hurricane Ida, suffering significant damage.
A Detailed Analysis
De Lucca said the credit union has run a detailed analysis of the impact of using AI to spot candidates for counseling sessions, and estimates in the first year of employing the technology to sweep through accounts AI will allow NOFFCU to reach three times the number of members compared with the current approach. Financial First Responders last year worked with more than 500 members.
“This is really going to take away the burden from our staff having to look through account histories and transactions that are coming in, and I think our first-year estimate is conservative,” De Lucca said. “We’re talking with a couple different companies now about what is the best way to extract the data.”
In addition to reaching more members faster and earlier and ideally, prior to their financial troubles mounting, using AI will free up Financial First Responders to have more in-depth conversations with those who are struggling, De Lucca added.
“This just will give us a huge return on investment for our members,” said De Lucca.
A Smart Ride
Auto loans is another area in which the credit union plans to employ AI by identifying members who are making payments on ultra-high-rate loans from title loan and finance companies.
“We have a Ride-Smart loan that gets members free of the high-rate car loan racket,” De Lucca said. “Most of the time we can reduce their rate by more than 50%. Plus, the loan has a built-in savings feature to help members if their car breaks down or they need new tires. It’s all about getting them a reliable car so they have transportation to get to and from work and to the grocery store.”