DISCOVER FINANCIAL SERVICES 

Achieving Safer Portfolio Growth


Discover Logo-02
“We found the best way to drive benefits faster was to complement our internal efforts with a partnership with Zest” 
Roger Hochschild |CEO and President, Discover Financial Services 
INDUSTRY
Consumer Credit
CHALLENGE
Improve profitability in personal loans
RESULTS
Significant reduction in charge-off rates

Some lenders get the value of machine learning. Some don’t. Discover Financial Services (NYSE: DFS) is one that definitely gets it. Its new CEO and President Roger Hochschild has this to say: “Banks that fail to invest in machine learning will end up fundamentally uncompetitive in a couple of years.” 

Discover, a leading U.S. direct bank and credit card issuer, turned to ZestFinance in 2018 to help build one of the largest AI-based credit scoring solutions in the financial services industry. Zest data scientists and analysts worked closely with Discover to tap its trove of consumer spending data to build a new model for its $7.5 billion personal loans business that assesses hundreds of applicant data points, up to 10 times more than Discover’s credit model had used before, to get a more accurate view of borrower risk.

In a successful trial, Discover and Zest found that the inclusion of more data and more sophisticated math reduced default rates significantly without added portfolio risk.

The new underwriting process will also help Discover approve applicants it might have previously denied. The companies expect to put their first ML model into production before mid-2019.

“A lot of banks struggle just to get their AI out of the lab,” said ZestFinance founder and CEO Douglas Merrill. “Discover is a leader in the use of AI in financial services because it has the vision to make machine learning core to their business. Using machine learning responsibly means deploying accurate, explainable and fair models.”

“We found the best way to drive benefits faster was to complement our internal efforts with a partnership with Zest,” says Hochschild.


Want to read this later?