Writer: Andrew Allen
Editor: Lillian Guevara-Castro
Reviewer: Adam West
Key Takeaways
Slope, a financial technology company focusing on bringing flexibility to B2B payments, has unveiled a new product that aims to help lenders more accurately determine the creditworthiness of small and medium-sized businesses (SMBs).
The system, which Slope calls SlopeScore, is the first cash flow underwriting tool a company has designed especially for SMBs, according to a press release on the product.
A business’s ability to obtain outside funding when it needs it can mean the difference between a banner year and a lackluster one. But securing a loan can be a challenge for many small and medium-sized businesses.
A Federal Reserve survey of small businesses revealed that only 51% of survey respondents who had sought funding in recent years received full approval for the amount they applied for. That means nearly half of the businesses that participated in the survey weren’t able to secure adequate financing at that time.
SMBs can fail to acquire the funding they seek because their cash flows are complex. And lenders may not have the resources they need to accurately gauge a small business’s cash flow.
Many companies can encounter obstacles when trying to secure the funding they need to grow their operations.
We sat down with Lawrence Lin Murata, CEO and Co-Founder of Slope, to learn how SlopeScore can eliminate difficulties in understanding the cash flow of SMBs and allow lenders to gain more confidence in their business lending practices.
Lin Murata’s parents run a wholesale business in Brazil. While he was growing up, he sometimes helped them run their shop. He told us that experience opened his eyes to some of the problems a business can face when attempting to acquire a loan.
He and Slope’s other Co-Founder, Alice Deng, spoke with additional business owners to learn about their success, or lack thereof, in securing financing.
“We saw this as a big pain point, and we had some new ideas on how to underwrite and what we could do to help solve this problem,” Lin Murata told us. “More and more, we began to realize that there is a big market for something like SlopeScore as a standalone product, and that’s why we launched it.”
Transforming Raw Data Into Actionable Findings
SlopeScore relies on a categorization engine powered by a language learning model to take raw transaction data and transform it into structured information that lenders can use to make informed credit decisions.
Lin Murata told us that making lending decisions on business customers can be cumbersome due to the number of income sources a company can have.
Many consumers may only have one primary source of income, but it’s a much different story when it comes to businesses. Companies can receive payments from a large number of customers, and those payments may be in the form of card transactions, checks, or cash.
Tracking a business’s expenses can also pose problems for business lenders who may have to wade through information from multiple suppliers to get a clear picture of a company’s cash outflows.
Businesses can have many expenses and income sources, which can make their cash flow difficult to determine.
“Very often a business will be making sales to a company that isn’t a mainstream business that you would know the name of,” Lin Murata told us. “And businesses can have short-term loans as well as long-term loans and credit cards.”
“That’s why this requires a more sophisticated approach, and it’s why we built a large language model to handle that,” he added.
Lenders who use SlopeScore may find they’re able to approve more small and medium-sized businesses for loans. The product not only allows a lender to grow their loan portfolio, but it opens the door to more opportunities to cross-sell to those companies a lender approves for financing.
Slope has piloted its new tool with a leading bank in the U.S., and the results are promising. Lin Murata said the bank tested the output that SlopeScore provides and compared it with results from other sources, including their own machine learning models.
“We didn’t touch the assessment,” Lin Murata explained. “And they came back to us and said they found that we were able to simultaneously decrease losses while increasing approval rates.”
