Debt Collections Process Optimization using AI

About Client

The client is Canada’s largest consumer finance retailer operating hundreds of stores across the country offering a wide range of financial services for their customers.

The Background

Optimizing collections process will see limited improvement with industry best practices such as multiple sales training and leveraging omnichannel technology to mass spam customers in order to maintain performance edge.

The Challenge

The client had a process of calling every customer before the loan due date. This meant calling the customers who had a record of paying off their debt on time. This naturally annoyed most of the customers, and affected the performance of the representatives as well as company revenue. To avoid the angst of customers, the Customer Service Representatives weren’t calling up at all, and just marking the customers as called. The entire process needed a new solution.

Our Solution

  • Past data of customers were used based on the criteria decided by the client themselves.
  • Data cleansing was performed to eliminate invalid numbers.
  • The data included insights such as the expected payments, and fees with interest owed which were analyzed in-order to classify the customers.
  • Variables were identified which could give better insight into the customer’s credit history and income modes, thereby, helping in strategizing ways to tackle them.
  • A DB schema was developed for the past data which used SQL database and machine learning algorithm to find the default rate of each customer, after which a score was assigned to each loan. Higher the score, higher the risk of default on those specific loans.
  • We built a business solution where we assigned “Risk Score” to each loan that was due, based on the customer’s history and loan data.
  • Daily output of various collection campaigns were reported back to the client after discussion with the in-house Business Intelligence

The Outcome

An accurate forecasting of the Cash Balance reduced the risk of holding surplus cash overnight across stores. This decreased the cost of idle money for the company and in turn added to the top-line growth of the firm.

By calling just

26%
of customers, the client was able to cover

65%
This meant better customer experience for those who paid their loans on time.

Parallely, this ensured better Customer Service Representative bandwidth, for Sales activities, instead of collections that decreased the idle time of agents.