How Elements Financial Took on Data Transformation

How Elements Financial Took on Data Transformation

How Elements Financial Took on Data Transformation

Elements Financial, a $1.4 billion credit union based in Indianapolis, is tackling the challenge of data transformation head on. In recent years, Elements has evolved from a single SEG credit union for employees of Eli Lilly and Company to one that serves 90,000+ members throughout Indiana and even nationwide with an increasing focus on digital excellence. Underpinning that effort is a transformational approach to data that will enable Elements to make smarter strategic decisions, shape future product and service offerings and provide a member experience that is both fast and personalized.

“Five or six years ago, we really started to think about bringing all of our disparate data into one place,” says Chris Sibila, EVP & chief information officer. “We didn’t begin by building a data warehouse, but we built data repositories, so that we could collect our data from transactions, our core system, mobile and online banking. By taking these steps, we knew where that data was and could use it to run analytics.”

At the same time, the digital experiences members were having made a serious impact on expectations. “Our competition out there is not banks,” says Sibila. “It’s Uber, Netflix and Amazon. They’re continuing to refine their data processes and strategies to look even smarter for their customers. That makes doing the same an imperative for us.”

Converting data into usable insights for the digital era is a simple proposition on the surface, but a massive undertaking in reality. As luck would have it, Elements found the ideal candidate for the job. Adam Arffa worked in IT at Eli Lilly for 26 years. He also volunteered for 12 years on the Elements board of directors. “I had always thought Elements would be a good place to work when I retired,” Arffa says. “When this opportunity opened up, I saw the perfect chance to leave Lilly and begin a new adventure.” He became Elements’ VP of business intelligence in March, 2017.


Arffa has been deliberate in his approach: “This is not a five-week exercise,” he says. “We have an 18- to 36-month roadmap to get to the point where predictive and prescriptive analytics are in mainstream use at Elements,” and data is playing a significant role in operations and decision-making. Here are the steps that will take Elements to that point and beyond:

Aggregation: “When you start on this journey,” Arffa begins, “you’re going to have disaggregated data. You have individual systems and each system has data you want. But you don’t have the ability to assimilate that data across the systems to see how things are going. The first step is to aggregate data from various sources. That’s a hard step, but it’s where you have to start.”

Simplify and integrate: Aggregating data is a big project, but it isn’t the final goal. “Once data was aggregated, we entered a phase where things got more interesting,” Arffa says. “We had data, but you practically needed a PhD to understand how to use it. We were stuck in a spot where we needed the few people we had in BI to work on reports. That’s just not scalable. If we in BI wanted to focus time on more advanced uses of the data, we needed the ability for business users to create on their own reports.” For Elements, this meant simplifying data so that non-IT users could readily understand and use it. “We are simplifying and integrating the data. This allows end users to spend 20 percent of their time selecting data, and 80 percent of their time getting answers to business questions from the data.” says Arffa. “We are enabling the business to create self-serve reports which gives them freedom to explore and use the data on their own, and helps to free up our BI staff to work on more advanced analytics projects.”

Being predictive: One of those forward-looking projects is predictive modeling, to give data a new role in shaping organizational intelligence. “When you’re using operational reports, you’re looking backwards, not forward: You’re looking at what happened to evaluate results,” Arffa explains. “Using predictive analytics, we can use models to predict what would happen if we added another product or targeted a specific group of members.” Predictive modeling elevates analytics to the strategic level.

On to prescriptive: “Once you have predictive capabilities, you can get into prescriptive work,” says Arffa. “We can use the predictive models which run against our data, and make strategic suggestions and automate tactical actions to be taken.” Using this kind of intelligence could provide game-changing insights to strategic planning, or create entirely new levels of service. For example, Arffa notes, “If someone were walking into an auto dealership, we might automatically send them an alert to let them know they’re pre-approved for a loan.”

Data governance: Maintaining the currency and integrity of the data is the final piece to the puzzle. “If people find out there are inaccuracies in the data, they’re going to be reluctant to trust it,” Arffa says. “Data governance initially focuses on what business opportunities to focus on, what data is needed to support those business opportunities, and what quality of that data is required.  Data governance is not a one and done exercise, but rather a continuous quality activity. What we’re looking for is applicable, consistent and appropriately qualified data, so that people trust and use the data.”


The intensive, highly specialized work Arffa and his team have been doing points to an even more intensive and universal goal: “We’re trying create a data culture,” Arffa says, “where people aren’t operating from their gut level, but are using data to understand our members and to make better decisions.”

“It’s critical for us to get to this kind of model if we hope to compete in the digital era,” Sibila agrees. “When you think about how companies in this industry like Visa and CO-OP are investing in their own data capabilities, it’s clear just how critical this is.”

At the same time, Sibila recognizes how difficult it can be for smaller credit unions to follow suit. Still, he says, “If you don’t think you have the size, resources and talent to undergo data transformation, reach out and ask your partners what they are doing – especially those partners who [capture] your data. They may be able to help you understand what’s possible or be willing to work with you as you try to get up to speed.”

Fellow credit union leaders are another resource: “Get help from peers in the CU industry,” Sibila urges. “I was just at a roundtable with CIOs where we focused on problems we are trying to solve. Don’t get hung up on the technology. Start with some basic problems you’d like to take on, and don’t feel like you’ve got to do it all yourself. That’s onerous.”

Arffa’s best and final advice: Start now. “You can sit around and talk about this for years and still not get it right,” he says. “You’ve got to do this in an iterative fashion. Pick a scope. Get some early success: You’ll also get feedback from the organization that this work is valuable.” Data transformation isn’t a quick fix, but its value is long-lived. “We expect that new uses for the data and types of analytics will start to unfold with artificial intelligence – many that are unimaginable through today’s lens,” Arffa says. “This is only the beginning.”

Learn more about how to enact your own data transformation in the upcoming Data Strategy issue of THINK Review, the magazine for credit union intrapreneurs. Get more information and subscribe today.