Using data science to mitigate car reposession impact

This is a guest blog post by Stuart Wells, chief technology officer, FICO

Debt is an age-old, yet very current problem: a recent UK survey revealed that over 20% of people under the age of 30 were forced to move back in with their parents to help clear their debt. This is one of many examples of how difficult it can be to get your debt back on track, but it is not the only one. Home or car repossessions are also important.

Now, car repossession is a last-resort option for lenders and car dealers, but a far more common issue than you might think. That’s what Dr. Jim Bander, from Toyota Financial Services, set out to change.

Helping struggling customers

Following the 2008 financial crisis, more people than ever were finding it difficult to keep up with loan repayments. The Toyota Financial Services team saw that, for the first time, over 100,000 customers per day were delinquent and needed help. Dr. Bander and his team realised that carrying on with the same collections approach they had before the global recession was not sustainable, especially if they wanted to help financially stressed customers.

The team at Toyota Financial Services wanted to adopt a more scientific approach to debt collection in order to achieve more flexibility, which is what automotive customers needed more than ever before. The goal was to keep as many drivers as possible in their cars. The challenge in this was developing a strategy that put the struggling customer first, without damaging the profitability of the firm.

Putting the customer first

The spike in delinquency cases following the 2008 financial crisis had exceeded the reach of Toyota Financial Service’s internally developed decision framework, which meant that new, more powerful tools had to be integrated into it. Bander used prescriptive analytics from FICO to optimise debt collection in this new era. The team believed that applying smart analytics would help them develop a more solid and reactive decision framework that supported their customer-focused collections strategy.

The Toyota team can now apply optimised strategies to collections customers, and gather data for future iterations to improve their decision making models. The new and improved framework, powered by self-learning analytics, allows collections agents to get in touch with customers sooner, better educate them about their options, and offer them a realistic repayment plan to help them get their accounts back on track.

The most successful new approach to date is what the Toyota Financial Services team calls ‘skip a payment’. Each customer is assigned a risk code that contains information about their debt and payment history. Collections agents can use this code to make an informed decision about whether that particular customer can afford to skip a payment and stay on track to clear their debt or not. The new software also allows collections agents to get in touch with customers earlier, which helps financially stressed customers to get more notice about the status of their accounts. This saves thousands of people from becoming delinquent in their payments.

Data for good: keeping drivers on the road

Prescriptive analytics and a smart customer-centric collections strategy have helped many Toyota Financial Services customers with their debt problems: “The number I’m proudest of is that 1,600 car repossessions have been avoided using this system,” says Bander. Toyota Financial Services has also helped more than 50,000 customers avoid a state of delinquency that would have negatively impacted their credit rating.

This is a great example of how data science can help organisations efficiently and effectively balance objectives.  It also shows that number-crunching doesn’t have to mean a “computer says no” approach — analytics can help businesses and their customers.

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

-ADS BY GOOGLE

SearchCIO

SearchSecurity

SearchNetworking

SearchDataCenter

SearchDataManagement

Close