Old Second Bank Uses Machine Learning to Predict Customer Behavior & Improve Marketing Effectiveness

Predicting High-Propensity Churners with Predictive Analytics

Written By // Richard Lamorena

Financial
500
Chicago, IL

The bank knew that, while they were successfully adding more checking customers every year, they also stood to lose a certain percentage of those customers due to insufficient methods for predicting high-propensity churners. In an industry where customers are known to frequently switch or drop accounts, leveraging data to understand customer behavior and improve marketing effectiveness had become a top priority.

Old Second Bank was looking to leverage predictive analytics to better understand customer behavior and improve marketing effectiveness. By leveraging the power of Machine Learning, they could analyze trends and identify correlations much faster and more accurately than a human ever could, giving them a significant leg-up up in a rapidly changing business landscape heavily dominated by mega corporate banks.

Old Second Bank

Old Second Bank is a Chicago-based bank that’s been serving the local community for over 140 years. The bank remains dedicated to maintaining a local presence with strong commitment to community, which has been the guiding principle throughout their expansion across the Chicago metropolitan area over the past century.

While Old Second Bank remains dedicated to their founding mission of preserving community banking and serving local people and businesses, they also recognize that mastery of Big Data and digital transformation is a key driver that will enable them to stay true to that mission in a rapidly changing business landscape heavily dominated by mega corporate banks.

Gaining Deeper Insights into Customer Churn

Old Second Bank was looking to grow the business and maintain their long-standing reputation for customer satisfaction. They specifically wanted to identify the variables that were impacting a customer’s decision to close checking accounts, and to do this they needed deeper insights into their business lines.

Prior to partnering with SWC, the bank was using customer surveys to identify at-risk accounts. This method was tedious and the turn-around was slow. By the time they were able collect and analyze the customers’ responses and implement targeted marketing and sales strategies around those insights, it was often too late to save the account.

The bank knew that, while they were successfully adding more checking customers every year, they also stood to lose a certain percentage of those customers due to insufficient methods for predicting high-propensity churners. In an industry where customers are known to frequently switch or drop accounts, leveraging data to understand customer behavior and improve marketing effectiveness had become a top priority.

Backed by an unwavering sense of financial responsibility, they knew something had to be done to stay ahead of customers.

Unlocking the Value of Data

As data becomes more feasible and accessible to use for decision-making, the bank’s leadership was looking ahead for answers, which they knew could be found in the massive amount of data they had already collected over the years. However, like most mid-size organizations, understanding what data was needed and then transforming that data into predictive analytics models was outside the scope of the organization and something their internal teams had never done before.

After exploring multiple options for outsourcing certain aspects of the project, SWC’s holistic approach, compelling business case, and ROI analysis was enough to convince leadership that partnering with SWC was not only a safe bet, but would prove to be a valuable partnership throughout their journey to becoming a data-driven organization.

Creating a Predictive Analytics Model

SWC worked in collaboration with numerous department leads to identify specific business goals, review available data to support the predictive model, and align those data points and business objectives with the right technology options. The sales, marketing, and operations teams each had their own set of questions they needed answered, but to get those answers, they needed access to the right data at the right time.

After understanding the business goals, SWC gathered the relevant historical data files and aligned them to various business requirements. With the right data gathered and organized, SWC’s data analytics team used Azure Machine learning and predictive analytics algorithms to create a predictive model that accurately predicts bank clients that were most likely to close an account.

Once the organization was able to successfully utilize their data, SWC set out to create a business dashboard that would compile and organize that data into actionable insights. After confirming the accuracy and business impact of the model, SWC worked in collaboration with Old Second Bank to operationalize the model and integrate predictive reports into lines of business.

Big Data Leads to Big Results

With the successful launch and integration of predictive analytics throughout the organization, Old Second Bank is now able to more efficiently act on insights gleaned from their data. The report summarizes predictive insights, including:

  • List of accounts at risk of closing
  • The likelihood of closing based on percentages
  • The top three reasons why a customer may be at risk of closing their account

The true power of the machine learning model is that it is able to run analysis on trends and correlations much faster and more accurately than a human ever could, giving the bank a significant leg-up on their competition in an increasingly fast-paced marketplace.

With the predictive analytics report, Old Second Bank is empowered to improve relationships with their customers and solve real-life business issues with enhanced reporting and proactive communications. The bank is more effectively targeting the right areas for business development and driving more sales towards their bottom line.

Plans for the Future

Happy with the foundation they’ve built, Old Second Bank is excited about the future. With the momentum of recent success, plans to expand their partnership with SWC by exploring other areas of advanced analytics are underway, including visualization in Power BI and integration/automation with data analytics.

Questions? Get in Touch

Just Released! 2018 Tech Insights ReportDownload
+