Inside Look: Kabbage automates small business lending

Automated lending fintech Kabbage is no stranger to disruption.

Founded in 2009 in the wake of the Great Recession, the fintech’s lending solutions were one of the first to leverage automation and dynamic data analysis to improve efficiency in small business lending. Investors took notice, raising the company to unicorn status in 2017 with a $500 million debt financing round, before putting the company’s valuation at more than $2 billion in 2019.

During the COVID-19 pandemic, Kabbage became the second-largest lender of Paycheck Protection Program (PPP) loans by application volume and, in August 2020, the company was acquired by American Express in a deal reportedly worth $850 million.

All roads for the Atlanta-based fintech, however, lead back to automation.

“My partner Rob Frohwein was exposed to an eBay API in 2007,” Kabbage co-founder Kathryn Petralia told Bank Automation News. “It was one of the first public APIs that allowed third parties to get access to seller and transaction-level data. He wondered if we could use this data to automate a lending decision to a small business.”

Kabbage is building on what it learned in 2007, keying in on application programming interfaces (APIs) to provide flexibility for its lending offerings, which move away from the static nature of traditional solutions by adapting to changing client forecasts. Security is another concern, one Kabbage meets with automated internal support and an API-enabled framework.

“We were only able to grow because of the proliferation of APIs,” Petralia said. “That’s the germ of how Kabbage started.”

Data-driven flexibility

Kabbage led JP Morgan Chase and Wells Fargo in approved PPP applications during the pandemic. Central to its prevalence was the fintech’s ability to digitally service micro-businesses low on the lending ladder that did not have the credit depth to receive traditional assistance.

“What’s important about our data is that we don’t just collect it one time; we stay connected to it,” Petralia told BAN. “We give customers a line of credit, which is very hard for traditional applications to do for a small business because they only get data one time.”

As opposed to one-time “analog” forms of credit reporting, like bank statements or tax data, API-centric customer data gives Kabbage and similar lenders like OnDeck and Fundbox a 360-degree view of a business’s creditworthiness.

“A customer comes to our site, then they authorize us to collect their data. They do that by logging in at third-party sites, like eBay, which shares the data with us. We have a token that allows us to get the data.”

Kabbage uses API tokens, which allow a user to authenticate with cloud apps and then retrieve data from the instance through representational state transfer (REST) APIs to collect critical financial information, including checking account data, payments and processing details, and shipping account history. Cash flow can then be determined and a line of credit can be extended, which is mobile and adaptive to changes in business. All information for both Kabbage and the user is accessible through the Kabbage app, a vital touchpoint for payment and repayment.

APIs are important in automated lending, David O’Connell, strategic advisor at Aite-Novarica, told BAN.

“There should be always-on API’ing into a variety of data sets so that you know not only how a borrower is doing now, but how you expect them to do in the future,” he said. “Then you can be the trusted advisor with a capital T and capital A.”

Automated security for small business lending

Credit determination is not the only area of automation for Kabbage. Security issues like identity verification are a chief concern, and one that has always been met with “robust and digital” measures, Petralia noted.

“One of the reasons we were such a large lender during PPP was because we had automated the identity verification, anti-money laundering (AML) process,” she said.

APIs connect information from a range of data sources for Kabbage’s verification procedures. Identity is “triangulated” using bank accounts, credit reports and accounting platforms, boosting confidence in the security of the onboarding and lending process.

“If you can log into a bank account that matches the credit report that also matches your accounting platform, then you have confidence that you’re verifying identity,” Petralia added.

While third-party sources are used in the verification process, aggregation and application of data is internal, encrypted and automated. This fosters speed and mitigates risk, a critical factor in winning the automated lending battle and ensuring that both client and provider can operate securely.

“Most Kabbage customers can get through the application process and have access to capital in less than 10 minutes,” Petralia told BAN. “No humans are part of that process.”

Bank Automation Summit, taking place March 1-2 in Charlotte, N.C., is the first and only event to focus solely on automation in banking. The event will feature the brightest minds from across financial services on intelligent automation strategies and deployment. Learn more and register for Bank Automation Summit 2022.