By John San Filippo
Finopotamus was onsite for the 2024 CULytics Summit, which was held March 25-28 on the Microsoft campus in Redmond, Wash. Among the many distinguished speakers was Breana Wolfert, vice president, data and innovation, at the $4.7 billion, Philadelphia-based American Heritage Credit Union. Wolfert’s 20-minute session was titled “Turning Data into Growth: A Case Study of How American Heritage Credit Union Utilizes Datava’s Data Activation Platform in Nearly Every Area of Its Organization to Drive Member Satisfaction, ROI, and Employee Effectiveness.” After her presentation, Finopotamus met with Wolfert, along with Denver-based Datava’s CEO Gordon Flammer, to discuss American Heritage’s data successes in greater detail.
Charting New Territory
Although Wolfert was only appointed to her current title in August of 2023, she told Finopotamus that she’s been laying the foundation for this role since 2017, when she started at American Heritage in a project management role.
“For the last seven years, I've been trying to facilitate data change, but I don't have the technical expertise to do that myself and we didn't have the organizational structure for me to lean on,” said Wolfert. That’s why she suggested to the credit union’s executive team that they engage with Datava, a company Wolfert knew from a previous credit union. Datava has supplemented the credit union’s internal team ever since.
“I originally ran a project management department under branch operations,” she continued. “We became the facilitators between business units and data,” performing discovery within individual business units.
She further explained that her approach to any data project is to “Start at the end. What do you want it to look like at the end? What do you need to know? And then we'll back into how do we get there. We were playing translator between what we know about the credit union space and the business units, and then relaying that over to Gordon (Flammer) and his team at Datava to execute on the business unit needs. We did that one business unit at a time, sometimes a couple business units at the same time.”
Wolfert’s data initiatives proved so transformational that credit union CEO Bruce Foulke decided to move her role out from under branch operations. “(Foulke) was like, we just need a dedicated data initiative,” said Wolfert, “so he moved me into that role and we brought in an SVP of data innovation from the outside, not from the industry at all.” Wolfert is this SVP’s direct report.
Data Challenges
Before American Heritage implemented this new approach to data, it was difficult for departments to get the data they needed. “One of the challenges we had was that everything funneled through IT,” explained Wolfert. “If you wanted data, if you wanted information, if you wanted a report, if you wanted anything to help your retail staff, you had to go through IT. And IT’s resources were more than over allocated.” While her role still falls under the credit union’s CIO, having data and innovation as a separate group has eliminated this issue.
According to Wolfert, another challenge is making sense of the siloed member data that’s stored in 20 or more different systems. “We have all these systems with all these different UIs (user interfaces) that our frontline staff has to navigate. And our members are spending a lot of time staring at our staff members’ cheeks. [Employees] are trying in session to figure out the member's story so that they can figure out what to talk to them about next.”
Wolfert noted that the “member story was everywhere. Everywhere. There's a little bit of information in the LOS (loan origination system), a little bit of information in the MOS (mortgage origination system), in the core and the online banking admin system. It's everywhere. How does that help our retail staff figure out where to go next?”
Multiple siloed systems inevitably give rise to the quest for a “single source of the truth.” However, Wolfert claims that such an absolute is not always required to resolve the current need. “Single source of the truth is such an interesting term for me,” she said. “Now we’re looking at something called sufficient truth. My boss always says, ‘Don't let perfect be the enemy of good enough.’ Sometimes good enough really is good enough, for example if you're trying to identify trend lines using data that seems like it's pretty close to the mark.”
Wolfert added that she’s been able to address all these issues and more with Datava’s data activation platform.
Supplementing Staff
“It's very expensive to bring in the kinds of talent that we want for our data initiative,” said Wolfert. “Like a modeler for AI (artificial intelligence) and machine learning. Like a data architect. A data engineer. Like all the super-skilled talents that we know we need. We tried to bring some of those specialties in-house a couple years ago.” She explained that the credit union hired a database administrator, but couldn’t afford redundancy in that function. So when that employee left, the credit union was left without a database administrator. The solution for avoiding these gaps: fractional employees sourced through Datava.
As an example, Wolfert explained that “fractionally, maybe we need a quarter FTE (full-time equivalent) as a data modeler, so 25% of that Datava FTE's time is spent working on stuff for us. We won't have a gap anymore because it's [Datava’s] job to fill that gap. And one of the coolest things that's most attractive is they're fed by the Colorado School of Mines. They have a constant flow of the type of technical talent we need at the credit union.”
“Redundancy risk is a huge problem for credit unions,” added Flammer. If you hire internal people, you’d better have redundancy in house. Otherwise, if that person leaves, you have a problem. Then you look at the total cost to go hire some firm and it ends up completely destroying the ROI calculations for other systems. That was one of the things that we really were looking to solve.” He noted that in the same way credit unions are looking to their data to solve problems, Datava is looking to solve those same problems on behalf of its clients.
Flammer continued, “It's like saying I need to get from one side of the country to the other, so I'm going to buy a plane. It's going to take a lot to get that plane from one side of the country to the other. Having that plane is a lot like software. Just having a piece of software that theoretically can do something is very different than having it consistently, reliably and without crashing do that task.”
“One of the speakers earlier today said that you can create the next new idea, or you can execute innovatively,” noted Wolfert, “but the answer really is, you need to do both.”