10 Steps To Better Business Intelligence

Like many people, Juan Jimenez stumbled into the credit union business by chance.

In 2016, he was working in senior analytics role as a reporting consultant for Verizon’s Southwest region, focusing on call center metrics, when he learned his job had been eliminated. That’s when he learned from a friend about a job opening at FirstLight Federal Credit Union for a portfolio analyst.

“I took him up on it,” Jimenez says. “I applied, and I got the position with no experience in credit unions or finance — and immediately fell in love with credit unions.”

Over the past eight years, Jimenez has worked with CFOs, EVPs, CMOs, and lending execs at three large credit unions — FirstLight ($1.6B, El Paso, TX), GECU ($4.4B, El Paso, TX), and SF Fire Credit Union ($1.8B, San Francisco, CA) — who helped him learn the business so he could deliver the right insights.

At each credit union, Jimenez has helped move the needle from analytics reporting to full-fledged business intelligence programs. He joined SF Fire Credit Union 18 months ago as assistant vice president of business intelligence, with the mission of transforming the program — moving disparate data warehouses to the cloud, cleaning and making sense of vast stores of data, and communicating insights with the organization.

“I’ve always been a huge believer that data drives every business,” he says. “It’s such a great feeling being a part of the department that handles the data that helps drive these decisions.”

According to Jimenez, it takes the right people, the right tools, and the right processes to build a BI department and develop a data-driven culture that runs throughout the organization. Here, he provides credit union leaders 10 steps to do that.

Step 1: Define Objectives And Goals

A credit union must clearly outline the objectives and goals of the division to align it with the organization’s overall strategic objectives. This includes identifying the key stakeholders and their requirements for data and analytics.

Jimenez recommends looking at the organization itself. If the major goals are around lending, it might make sense to embed the BI team in the lending and product marketing group. If the goals focus on income, risk, and costs, put the program under the chief financial officer. If data centralization is important, put the BI program under the IT department, he advises.

Jimenez has worked under each of those models and sees advantages in each one.

“Here at SF Fire, we do everything,” he says. “This is the only credit union I’ve been with so far that BI falls under the umbrella of the IT department. It’s great because we have access to great technology and the budget of the IT department.”

Step 2: Assess The Credit Union’s Current State

Evaluate the credit union’s existing data infrastructure, systems, and processes with an eye toward identifying its strengths, weaknesses, opportunities, and threats (SWOT analysis) related to data management and analytics capabilities.

Then, assess the credit union’s current state. What’s the budget? Who does BI staff report it?

“Do you have an existing data infrastructure?” Jimenez asks. “Are you using just what came out of the box with your core system? Do you have any existing processes? Do you have anybody doing any reports? Identify those strengths.”

Many organizations, for example, already have personnel doing analytics for various business units. Those resources might have greater impact as part of a centralized BI team sharing the same tools and processes.

Step 3: Develop A Roadmap

Create a comprehensive roadmap outlining the steps and milestones required to build the BI division.

Define the scope, timeline, budget, and resources needed for each phase of the project. The size of the credit union and the budget it’s willing to commit to BI will affect key decisions about the roadmap. For example, a fully functioning BI team needs analysts, database administrators, architects, engineers and data scientists. The job market for these roles is highly competitive. Big Tech firms such as Google, Meta, and Microsoft pay up to $250,000 for technical resources.

“For credit unions, not-for-profit organizations, it’s difficult for us to compete with the bigger institutions,” Jimenez says. “It’s even more difficult for the smaller credit unions to hire talent that can drive that transition. The board has to approve the positions, and the CEO has to create the positions and the pay scales. Sometimes it’s hard to convince folks that you need an engineer, and that engineer is not the same as the architect, and the architect is not the same as an analyst.”

Even the BI tools a credit union chooses can affect its roadmap, he adds. Tableau is one of the most popular tools used for data analysis and visualizations, but finding skilled analysts to use the software can be challenging. Many smaller organizations might be better served with Microsoft’s Power BI tool, he says, because the software comes free with Office 365 licenses.

Step 4: Establish A Governance Framework

Define data governance policies, standards, and procedures to ensure data quality, security, and compliance. Establish roles and responsibilities for data management, including data stewards and data custodians. Don’t be discouraged if the BI team faces competing demands and shifting priorities from various parts of the business.

“Once you have a governance framework, it’s harder for anybody to be upset over how you’re running the department,” Jimenez says. “You have clearly defined responsibilities for everybody.”

Step 5: Build The Data Infrastructure

 Invest in the necessary technology infrastructure to support BI initiatives, including data warehouses, data lakes, ETL (extract, transform, load) tools, and BI platforms. Ensure scalability, reliability, and performance of the infrastructure.

Step 6: Integrate And Manage Data

Integrate data from disparate sources across the organization to create a single source of truth. Implement data quality processes to ensure the accuracy, completeness, and consistency of the data.

Integrating data from other systems such as credit card processing or separate mortgage or lending applications is important for getting the whole picture for analytics.

Step 7: Develop Analytical Capabilities

Train staff on BI tools and methodologies and build analytical capabilities within the BI division, including data modeling, data visualization, and advanced analytics such as predictive analytics and machine learning.

Jimenez says one of the most powerful reports he used in El Paso helped predict when members dealing with financial problems needed early intervention by collections. By assigning a point value to members for lower-than normal checking account balances and higher-than-normal credit card balances, the system can show when the borrower is late but not yet in delinquency. A representative can call the borrower and offer to skip a payment for a modest fee.

“You have a member that’s happy because the credit union reached out to them to offer help,” Jimenez says.

But that’s not all.

“You potentially made a couple hundred dollars or thousands of dollars in fee income during a time when there are no deposits. You still grow that interest income, your delinquency ratio is low, and you have less of your membership going into delinquency,” he adds.

Step 8: Foster A Data Culture

Promote a data-driven culture within the organization by educating stakeholders on the value of data and analytics. Encourage collaboration and knowledge sharing across departments to leverage insights for decision-making.

Jimenez says fostering a data culture across the organization is his favorite part because the availability of data across multiple departments helps spur collaboration throughout the enterprise.

“When you have a meeting with your heads of departments and they all know what the data looks like for everybody else, it’s easier to make decisions,” the AVP says. “You get less pushback on these decisions.”

Step 9: Focus On Continuous Improvement

Monitor and measure the performance of the BI division against predefined key performance indicators (KPIs). Solicit feedback from stakeholders and adapt the credit union’s BI strategy as needed to address evolving business needs.

After creating the reports and visuals, however, the job isn’t done. The BI team should continuously monitor success against KPIs and ensure the metrics are addressing the right questions. A good practice, Jimenez says, is reviewing which reports are being used and which ones are gathering dust.

“If I see the report hasn’t been used in two weeks, I will email the end user and say, ‘Hey, I noticed you didn’t use the report. Are you focused on something different now? Would you like me to change the report? Is there something wrong with the report that no longer fits your needs or the needs of the department? Can I help to make this report more useful for you again?’” Jimenez says.

Step 10: Communicate The Results

Communicate insights and findings derived from BI initiatives to key stakeholders in a clear and actionable manner. Demonstrate the value of BI investments through tangible business outcomes and return on investment (ROI) analysis.

For example, if members are transacting more business through the mobile app, the data might justify the credit union investing more in mobile banking. If the data shows call center volumes are up and branch visits are down, perhaps the credit union could invest in call center staff and tools or consider reducing branch hours.

“It’s not just about profit or lending or collections data, it’s about having transaction data on your members to know how they’re spending their money,” Jimenez says. “If you’re going to have a rewards credit card, where do you want to put those rewards — in groceries or fuel? What are your members doing? How are they transacting? We’ll have of this information with good business intelligence.”

The Future Of Analytics

The past is littered with poor decisions made without checking the data. Jimenez recalls a Texas credit union that had just opened its first virtual branch in a growing part of town. The branch was so successful the credit union decided to retrofit another branch across town with interactive teller machines.

That branch across town, however, supported older neighborhoods. In fact, a nearby retirement home regularly brought residents to do their banking at the branch. Without warning, the credit union was flooded with calls from members who didn’t know how to use the ITMs. The credit union dispatched customer service representatives, and years later, they’re still staffing that virtual branch.

“If they had a report of the demographics of who transacts here — and the age groups of the members — they could have said, ‘Hey, we should probably not touch that branch,” Jimenez says.

For more than a decade, analysts have dreamed of the “democratization of data” — when anyone in the organization can pull the information they need to get insights. The tools, however, are still too complex for most employees to create their own reports. But that’s changing with the emergence of generative AI tools such as ChatGPT, Jimenez says.

Jimenez already sees promise in ChatGPT creating queries that can be created through a conversation with the chatbot and then plugged into analytics tools. It’s only a matter of time before employees outside of the BI team will be able to access their own data through an AI bot, he predicts.

“It’s really exciting that other people are going to be able to do queries as much as I do,” the AVP says. “Conversational AI is going to be huge. You might even get ideas from an AI model. For example, you might tell it your collection ratio is high this month, and it might say, ‘Have you considered this?’ No, I hadn’t considered that. How can I go about implementing it? ‘Here are 10 ideas on how you can do that today.’”