Progressing with Collection Scoring
July 12, 2010
Scoring can be an extremely beneficial practice for identifying accounts with the greatest potential.
Collectors and asset buyers are constantly seeking methods for collecting smarter and improving profitability. During their ACA convention workshop, Progressing with Collection Scoring, Dwayne Banasiak, director of consumer scoring sales for Predictive Metrics, and Mark Detrick, chief financial officer of Capio Partners LLC, provided a detailed look at how statistical collection scoring can help achieve these goals.
Statistical collection scoring is a predictive decision system designed to accurately rank order and segment recoveries based on mathematics and applied statistics that is proven through validation analysis. According to Banasiak and Detrick, scoring helps with optimizing the assignment of collection resources, focusing on accounts with the greatest potential and knowing when to limit or extend collection efforts.
“Today, you can't continue to do business the way you always have,” Banasiak said.
They examined four types of collection scoring and the benefits and drawbacks to each:
Judgmental/rules-based scoring: This model uses a combination of internal and external bureau data. Collection management and staff subjectively set the rules. It analyzes who is most likely to pay.
Bureau scoring: Generic scores are developed by the credit bureau using its data, typically derived from credit reports.
Placement scoring: This model uses statistics to build mathematical models that are used to predict the probability of payment and the amount to be collected. Internal placement data drives model performance and is supplemented with socio-economic and demographic data.
Blended scoring: This type of scoring uses statistics to build mathematical models to predict the probability of payment and expected dollar amount. It blends fresh credit bureau data with account-level data.
Because working accounts that cannot pay is a waste of resources and because most dollars come from a small portion of accounts, scoring can be an extremely beneficial practice for identifying the accounts with the greatest potential.
