SAN JOSE, Calif.-The $1-billion Meriwest Credit Union is using predictive analytics to identify potential borrowers.
Tony Cortez, VP of marketing and interactive banking, told Credit Union Journal Meriwest has sought to increase wallet share by analyzing members' credit usage, identifying those members with car loans, mortgages and credit cards at other institutions, and then determining which of these would most likely consider switching to Meriwest.
Once identified, those members are targeted with "focused and tailored" e-mails with messages such as "Shrink Your Payment" and "Auto FICO." Invitations were issued offering to save money by moving loans and credit cards to Meriwest for better rates and lower payments on the outstanding balances.
"We look at the members' history, looking for trends on what they are doing," he explained. "At the same time, we use a little bit of predictive analysis to determine a likelihood to buy."
Meriwest said the response to the e-mail campaign was "very successful." Since launch in early 2012 through the end of the year, MCU refinanced $5.3 million in auto loans, a $3.6-million increase in direct auto loan production over the same time period the previous year.
The campaign also targeted members with credit cards and mortgages at other institutions, resulting in nearly 70 new credit card-holders and $9.1 million in refinanced mortgages its subsidiary, Meriwest Mortgage.
Drilling Down Into Data
What's driving the response? "It is a little bit of everything," Cortez said. "There also is taking a look at the data and determining what we are seeing. There is a lot of data, and it takes time to find something that is useful for the purpose we want to use it for. We have great people here and we are truly drilling down into the data."
At deep levels the behavioral analysis side works closely with the data analysis side, Cortez continued. For example, if the member has recurring payments going to Bank of America, that is a good indicator Meriwest might be able to talk to them.
Some clues are easier to determine than others, he noted. "If someone has a payment to Toyota Financial Services we can assume they have an auto loan. Our next step is to talk to the member, maybe offer our 'Shrink Your Payment' plan, which is an auto refinance."
Cortez said there eventually was a second campaign using analytics and electronic marketing that targeted members close to paying off a Meriwest loan-with the CU sending them an e-mail offer for their next auto loan.
Meriwest did not run the latter through a formal pre-approval process, Cortez said, but it had ideas of ranges given the member's FICO score.
"We are trying to narrow our bandwidth and doing refined data analysis," he said. "We are not sending an offer to everyone who does not have an auto loan. That makes for a targeted campaign to people who are more likely to act on it, creating a better response rate."
Narrowly Targeted Messages
The data mining and targeted e-mails are on top of traditional advertising, which Cortez said builds awareness . MCU continues to look for ways to make its marketing messages more concentrated and focused. For example, in Meriwest's online banking the banner ads are designed to appeal specifically to the member who is logging in based on the person's history.
"One member will see a completely different set of ads than another member, which is working out well for us."
Currently, the data mining and electronic marketing program is on hiatus to refresh the database. Cortez said this will allow Meriwest to have updated information on its members. It plans to relaunch in the spring.
"We are pleased with both sides of the campaign," he said. "We have a very robust database system, as does every financial institution, but what we get out of it is the tricky part. It is all about managing the data."