AI Creeps Into Credit Union Card Rewards
Mixing AI with card rewards still in many ways requires human brainpower in order to pay off.
It may sound ultra-high-tech, but AI is already taking hold in some parts of the industry. According to a survey of credit unions and banks by financial technology company Infosys Finacle and global trade group Efma, for example, last year 46% of the respondents did not use conversational AI services such as chatbots or voice-based interface to talk to members – but planned to offer those things in the future. Financial institutions have also started using AI for risk and fraud management, the organizations recently noted.
One of AI’s less common uses, however – one that’s rife with opportunity, according to some experts – involves sifting through credit union card transaction data in order to uncover what members like to buy and where they like to shop, and then offering members customized rewards that encourage certain card behaviors.
The work may involve machines, but the advantages are anything but artificial: Stronger member loyalty, top-of-wallet card status and higher interchange revenue, experts touted. But mixing AI with card rewards still in many ways requires human brainpower in order to pay off. Two industry pros shared their tips for making the most of this futuristic technology.
Data Dynamics
One of AI’s biggest roles in the card rewards world involves corralling seemingly infinite heaps of digital information from card transactions. AI applications can sift through that data to detect everything from spending patterns to favorite merchants and other shopping habits. Armed with that information, credit unions can send, usually via text message, tailored reward offers to certain members – 1,000 extra points for using the credit union’s card at a specific retailer in a specific location five times in a month, for instance.
“The member would see a more targeted, personalized type of marketing engagement message,” FIS Chief Product Officer of Banking and Payments Bob Legters said. “More ‘Hey, Bob, you haven’t bought gas in six months. Hey, make sure and grab gas from X; use your credit union card to get gas.’ Those type of things.”
The idea is to reward members for keeping their credit union cards top-of-wallet, as well as reward behavior that drives interchange, noted Cyndie Martini, who is CEO and president of the CUSO and card processor Member Access Processing in Seattle, Wash.
Engaging with members via mobile might also boost loyalty. A member using a mobile app generally owns one more product on average than a member who doesn’t use a mobile app, according to Martini. Plus, mobile app users tend to perform 19% more point-of-sale transactions, 25% more ATM transactions and 13% more ACH transactions per month than branch-only users. Those members are less likely to leave the financial institution, she said.
Recurring transactions are often prime targets for AGI-driven card rewards, Legters noted.
“We incent a lot of gas, grocery, dining, et cetera, because those are the things that happen more frequently. You only go to the home improvement store, really, maybe once every other month. You only go to the pet store once a month, but you get gas once a week. You get groceries once a week. And so a lot of the incentives are driven around what benefits the credit union and what strengthens their relationship with the member,” he said.
However, it’s unlikely that many credit unions are AI experts right now.
“The number of people who are really actively engaging in advanced AI-type stuff is probably less than 5% of the credit unions,” Legters estimated. Many more may already use or subscribe to AI-embedded services, he noted. “But there’s a difference between ‘Hey, I have a solution that does that optimization for me, and I let that solution work for me,’ and ‘I have a solution, and I’m going to get in and really work with it to really get down to the nitty gritty,’” he said.
Robot Shopping
The biggest step, Legters said, often involves acquiring card purchase data from card processors.
“Your processor should be the system of record for all the transaction history for your member,” he explained. “They have all the history of transactions from your member, which is a good baseline to understand.”
Sometimes that data can be had for less than $1,000 a month, Martini added.
However, credit unions don’t typically go out and “buy an AI program” per se – AI is often already integrated into many of the systems credit unions already use. But credit unions may have to do their homework a little differently if they want to launch AI-integrated rewards programs.
“A lot of that is all bundled together in the solutions that you’re getting offered. And if it’s not, the more separate components you buy to create the program, the more expense you’re going to have, without a doubt,” Legters warned.
“There are card processors who offer rewards programs,” he added. “Then there are also rewards processors who do nothing else but rewards, and there are card processors who do nothing else but card.”
Super Humans
Attracting and retaining talented people who are interested and able to wrangle with AI is important, according to Infosys Finacle’s latest survey.
“In what appears to be a conflict between importance and organizational focus, banking organizations worldwide are highly aware of the positive impact of AI, advanced analytics and machine learning in almost all areas of the bank, yet still do not focus on the need to develop these skills or focus investment in these areas.”
And even when credit unions have AI-driven systems in place to gather and analyze all that data, they still have to decide what to do with it all – a task that also requires human intelligence.
“We recommend with our credit unions that they have a digital strategy, which is for their card, but they also have a data strategy,” Martini said. “There are so many tools out there. Financial institutions need to look at what your goals are first, what you are trying to accomplish, and then create a partnership that supports that.”