Published On: November 3rd, 2020Categories: Automation, Collections, Machine Learning, Predictive Modeling

Make your attempts count.  How Machine Learning models can help with the new CFPB rule

The CFPB recently released a new rule that limits contact attempts to only 7 per week per consumer. 

Maintaining and improving Right Party Contact (RPC) rates is now more important than ever!

This means that you not only have to call the right number, but you have to do it at the right time. The rule will have many agencies scrambling to figure out a way to make the most of their calling strategies to keep their collections rates from dropping.

How will agencies ensure they are staying compliant and still making effective calls?   The answer is Artificial Intelligence. 

For over 5 years, NLP Logix has been building machine learning models to help agencies prioritize and segment their portfolios.  One of the outputs of the ML models is to recommend the best time of day to make a call.

collection attempts

There are several factors to consider – where is the consumer located, what kinds of phone numbers do you have for the consumer, and how many different numbers are tied to this account?  By training a ML model on an agencies’ historical data, the model can recommend the best time of day to call a particular account. By augmenting this data with external data features, such as census and NAICS data, models can paint a pretty good generalized picture of consumer behavior and provide a better understanding of anticipated schedules. This comprehensive approach ensures you are making the most of your limited calls.

One of the great benefits of using machine learning models is that the “best time of day” recommendations is only one of the possible outputs available.  That same ML model can also provide:

  • Propensity to pay score
  • Recommended outreach channel (phone, letter, text, email)
  • Expected Value
  • Payment option recommendations and suggested payment plan amounts

All of these ML models run behind the scenes and all accounts are ingested into the model and scored daily to ensure that up to date information is incorporated.  Accounts are scored, prioritized, and incorporated into your calling strategy.

With many companies currently planning for their 2021 budget, now is a great time to learn how a machine learning model can benefit your operations and help increase collection rates while still being compliant with the CFPB.

For more information on how machine learnings models are trained and put into production, contact one of our team members today.