Identifying Good Borrowers Among the Underbanked: The Role of Non-Traditional Smartphone Data
Leading banks such as JP Morgan Chase and Royal Bank of Canada are turning to alternative data, such as job and educational track records, to fill the gaps in potential borrowers’ credit history. Experian, the global consumer credit reporting agency, will soon begin taking cellphone and utility payments into consideration when calculating the credit score of “thin-files”. This is helping them identify credit worthy borrowers among younger customers and recent immigrants with thin credit histories, and even traditional customers whose credit scores were hit during the recession. Alternative data sources are also becoming pivotal in emerging economies, where credit bureau data on consumers is often inaccurate or incomplete. The result: accelerated financial inclusion and growth.
But as banks increasingly leverage alternative data sources to predict the applicant’s creditworthiness they face challenges in verifying the data. Alternative data sources - such as e-commerce delivery address data or payment histories for utility bills, rent and insurance - can be altered to reflect false creditworthiness. In addition, the risk of accessing personal information without customer consent and increased security threats from such data act as further deterrents for banks and financial institutions (FIs).
Fortunately, the growing internet and mobile penetration presents an unprecedented opportunity for FIs to access and capture customers’ digital footprints. By facilitating day-to-day interactions such as calling, texting and emailing, smartphones have transformed into electronic diaries of individual phone users. According to a paper published by the Federal Deposit Insurance Corporation's Center for Financial Research, even a minimal set of digital footprint data, such as device operating system and email host, could yield predictive models for defaulters.
Here are three ways lending institutions can use non-traditional mobile data to offer innovative lending solutions.
- Evaluate customers’ default risk: Analyzing metadata across categories such as potential buyer’s call history, messages, contacts, calendars, applications and downloads, can provide financial institutions insights into a person’s propensity to default. AI and machine learning algorithms, for instance, can tap into more than 100,000 characteristics from a user’s smartphone to connect the dots in ways that traditional methods cannot. The result: ability to build digital score cards for faster, more informed and profitable lending decisions.
- Reduce the time to yes: Leveraging ML-driven models allows banks to sift through customer metadata faster, bringing the time to assess customer credit-worthiness and offer loans down to an astonishing 30 seconds. Without hampering the customer’s privacy, lenders can analyze the way smartphone devices are used and correlate it with that of delinquent users. Using this method in addition to the existing credit scoring models can help expand the customer base by creating a pipeline of pre-approved customers.
- Segment customers to cross-sell and upsell: Consumers are increasingly using mobile-based payment platforms and mobile shopping applications in their daily lives. Metadata from such platforms offer a clear view into a customer’s spending patterns by geography, enhancing lender’s ability to offer personalized high value products and build niche segments for cross-selling and upselling.
With smartphones expected to account for 77% of mobile connections in 2025, leveraging mobile metadata can result in positive outcomes for lenders such as an expanded customer base and improved profitability – all while reducing risks. Using non-traditional data for developing credit scores can also help lenders reduce their delinquency rates by 20% to 60%. The World Bank, the Asian Development Bank and the Global Initiative of Financial Inclusion have issued guidance notes that encourage the use of digital footprints, including mobile device data, to foster financial inclusion. As the non-traditional data space is poised to reach an inflection point, partnering with an experienced fintech company to harness alternative data can help lending institutions enhance customer as well as business outcomes.