IC System, one of the largest receivables management companies in the United States, adopted a machine-learning powered approach to increasing recovery rates by creating a comprehensive contact strategy for each consumer. The strategy was centered around an Artificially Intelligent Collections (AIC) product developed by NLP Logix, a Jacksonville, Florida-based AI solutions company, and leveraged the decades of consumer behavior embedded in the IC System databases.
“First and foremost we wanted to ensure that we protect the consumer and use the data in a compliant and discreet manner” said David Gunderson, Director IC System, “while ensuring an approach that maximized recoveries for our clients.”
The approach used by IC System and NLP Logix is based upon training a machine learning algorithm developed by NLP Logix specifically for the accounts receivable market. The algorithm identifies patterns in consumers historical payment behaviors and applies a probability-to-pay score to similar or like consumers with outstanding accounts. This was further refined down to specific industries of IC System’s customers, such as healthcare and telecommunications, to capture the different ways consumers approach their doctor bills vs. phone bills.
“IC System had an excellent scoring system prior to us implementing the AIC solution” said Katie Bakewell, Lead Statistician, NLP Logix. “It provided a great performance benchmark for our teams to reach a goal of exceeding those recovery numbers.”
An additional significant advantage to using this unique approach is that credit bureau scores were not used to score the consumer’s probability to pay. Not only is there a significant cost to IC System to obtain the credit bureau scores but there is a potential negative impact on the individual consumer. In addition, The Federal Trade Commission found that 1 in 5 consumers had a confirmed material error in their credit score.
“A twenty percent error rate is much too high for our approach” said Bakewell. “While there is some correlation between a consumer’s credit bureau score, a combination of other factors, like debt amount, age of debt, etc. is much more predictive of whether a consumer will pay a bill or not.”
About IC System
IC System is one of the largest receivables management companies in the United States. Celebrating its 80th year, IC System is a family-owned, privately held accounts receivable management firm in its third generation of family ownership. IC System provides customized, tailor-made debt recovery solutions for healthcare, dental, financial services, small business, government, utilities, and telecommunications industries on a nationwide scale. With their Core Values (People, Pride, Integrity, Performance, Innovation) driving every business decision, the nationally licensed IC System maintains a 16-year average client tenure, has more than twelve million debts for $6 billion placed annually in their inventory, and is endorsed by over 450 professional healthcare and trade associations/societies. IC System is home to over 550 employees and has been named a Top 150 Workplace by the Minneapolis Star Tribune.
Follow IC System at http://www.icsystem.com, on Twitter at @icsystem or Linkedin.
About NLP Logix
NLP Logix is an artificial intelligence/machine learning product and automation solutions provider, which has grown from a vision in 2011 to one of the fastest growing teams of deep learning practitioners. NLP Logix delivers its solutions through LogixStudio, the company’s proprietary automation platform that gives a customer the ability to quickly leverage the growing library of algorithms and deploy them into their workflow. NLP Logix is delivering automation and machine learning solutions to customers across a wide swath of industries, including financial services, energy, healthcare, government, human resources, and others. NLP Logix was recently recognized as one of the top 5000 Fastest Growing Companies in United States by Inc. Magazine. For more information, please visit https://www.nlplogix.com.
Follow NLP Logix at https://www.nlplogix.com, on Twitter at @nlplogix or LinkedIn.