Finding a smarter way to manage large data sets

Predictive modeling helps debt collector CBE Group round up accounts receivable

From its early roots in 1933, CBE Group has evolved into a national leader in accounts-receivable management. Today the company is defining the future of debt collection by embracing a powerful combination: an innovative analytical infrastructure and an companywide commitment to ethical and professional interactions with consumers.

The result: efficient operations, superior results from a data-rich environment and stronger brand reputations for its clients.

The Analytics Solutions Group (ASG) is the heart of the company's initiative to improve business insights using SAS^® technologies for a variety of tasks, including data manipulation and statistical model building.

Each day, CBE receives high-volume data updates to huge client databases that store information about consumers and debt statuses. That data can be unorganized, unformatted, redundant and sporadic, requiring reformatting and transformation into structured data sets for analysis – a previously cumbersome and lengthy process.

Seeing all areas of the business

To more efficiently manage those databases, ASG relies on SAS to aggregate data from various sources. The resulting time savings enable analysts to spend more time exploring, mining and analyzing the data to uncover trends, patterns, and insights that improve business operations.

According to CBE Analyst Shane Roberts, the value from SAS is extensive. "SAS is an invaluable tool for data manipulation and data mining," he said. "For the large databases we're analyzing, SAS is essential for us to drill down on each client, measure performance and create granular reports.

"SAS provides reporting and dashboard capabilities that we've never had before … whether it's inventory management, payment statistics, or dialer performance. Now we can easily and quickly see what's happening in all corners of our business," said Roberts.

Optimizing profitability and efficiency

Specifically, SAS solutions are helping CBE improve debt-recovery efficiency and collection-agent performance. To maximize profitability and efficiency, CBE needs to focus its team on the accounts that present a projected optimal combination of recovery amount and ease of recovery. Determining the ranked priority of accounts to pursue is a key task for ASG.

The team turned to SAS Analytics Pro and SAS Enterprise Miner to create predictive models for each client group that identify the best candidates for debt collection. These well-developed models help CBE understand how certain consumer characteristics and variables can influence their ability to make payments.

Assumptions and distributions in many CBE models will change over time, making it necessary to constantly monitor and adjust models to maintain or improve their efficacy. Analysts in ASG use SAS to monitor the applicability and performance of their models by creating distribution charts for all variables to detect population changes with each new wave of inventory (accounts) received.

This SAS routine, which CBE calls its Model Distribution Program, combines data manipulation and data analysis to transform raw new inventory data into the desired format, append it to the base data set, and then compare the variable distributions for different inventory periods through histograms and Pareto charts. Instantly, the ASG analysts can see if modifications are needed in their models.

Exploring the data

Using SAS data visualization software, ASG analysts can ask the exploratory questions and find patterns, interactions and stories hidden within the massive data volumes.

For example, James Zheng, an ASG analyst, built a collection-agent performance measurement system that enables CBE to compare each agent's actual versus expected performance.

"SAS helps us capitalize on the wealth of real-time and historical data to create a data-driven business that sets the standard for debt collection," said Chad Benson, Chief Operating Officer for CBE.

"SAS processes large amounts of data and offers flexible data manipulation – both of which help solve critical business and decision-making needs across our company," continues Benson.

"As our analytics team has grown to more than 20 analysts, our ability to efficiently analyze data and perform more complex analyses makes SAS the perfect fit. SAS will further benefit our clients as we move into real-time decision making and begin using predictive analytics capabilities, like neural networks, to drive our collection efforts into the future."