How community banks solve problems with data

Banks have always had a lot of data to wrangle. Core processors handle transactions by the millions. Loan applicants have their data compared to underwriting standards. Banks reach outside of their walls to obtain credit scores or verify a customer’s identity.

Yet the need for data — complex data — is growing. Data can inform bank decisions from the very beginning with a new customer account through maintaining a portfolio of accounts, mitigating risks and reducing costs along the way. And interconnected data is more valuable than siloed data, meaning that banks are increasingly relying on data other than their own or relying on disparate systems to “talk” to each other.

Can the onboarding process be automated? Can outside data lead to a better internal decision? And can banks wield the power of data transfers through APIs to more efficiently manage their portfolios? The answer to all of these questions? Yes. 

Should they all be addressed at one time? For your own sanity, probably not. Instead, banks should prioritize different projects based on where they have the most to gain, either in speed, reducing errors, or making better business decisions. 

We explored three examples of ways that banks can improve their processes through data, from the earliest stages of the customer journey to analyzing volumes of data across multiple platforms.

Removing delays in account opening

Online account opening skyrocketed during the pandemic, with completed applications for online accounts growing by 40 percent in 2020, according to a report by Bottomline Technologies, a business payments platform. With the right platform, community banks could offer a digital account opening experience that rivals the big banks.

Fintechs have made this process very easy, with lots of gears turning in the background to verify a new customer’s identity. Individual account opening can use technology like government-issued ID validation and knowledge-based authentication to verify identity, taking an applicant through the account opening process in a matter of minutes. 

Complying with KYC/KYB, AML, and BSA requirements is complex, however, and processes can be disjointed when it comes to opening business accounts. Kaylie Hodge, partnerships manager at Middesk, has seen firsthand the pain points that banks experience. “The technology for onboarding businesses has lagged behind because it’s really challenging to access high-quality registration data on businesses, which is fragmented across different states and government entities,” Hodge said. 

When banks require businesses to upload documentation, such as Articles of Incorporation, and then review the documents manually, there is a delay in finalizing the account opening or approving the loan application. Small businesses may be tempted to look elsewhere, to banks that promise digital account opening or loan approval in minutes instead of days.

And the stakes are high: Banks have paid billions of dollars in fines for non-compliance with AML and KYC/KYB requirements. For many banks, the approach to satisfying the regulators seems to be “hire more staff” to handle the volume and regulatory requirements, rather than address the underlying issues of automating the customer onboarding process. 

As a platform, Middesk takes a really aggressive data acquisition strategy with pipelines to the different Secretaries of State, the IRS, and other data sources so that banks can fulfill their KYB requirements automatically for business account opening. But the power of platforms like Middesk isn’t necessarily as a direct vendor to banks: It’s embedding its technology with other platforms so that banks can choose a vendor that meets their digital account opening needs, through API-based integration.

Hodge said that by implementing digital account opening and loan approval products that satisfy KYB requirements, banks can automate the majority of applications so they don’t have to spend time manually reviewing them. “There is increasing efficiency through reducing overhead costs, increasing conversion rates by opening more accounts or issuing more loans, and reducing risk through access to accurate, complete, and up-to-date data,” Hodge said. 

Connecting the loan approval dots through data

As a loan committee weighs all of the data involved in a credit decision, it can be hard to assess the borrower’s financial health. Annually updated RMA comparison data has existed for a long time, giving banks a benchmark for businesses within the same industry, with similar assets or income. 

Yet RMA comparison data only offers the option to compare a business’s financial statements. Commercial real estate is a different ballgame, with the borrower’s financials carrying an additional layer of complexity, detailed in the rent rolls. 

CRE stress testing has become an imperative for modeling different scenarios, but banks are still limited to what is known, based on their own internal CRE data, rather than a larger picture of all CRE in their geographical area. Or even CRE data across the country. Or CRE performance as the result of a global pandemic.

Tony Lesicka, CEO of Coverlease, has been deeply rooted in commercial real estate for most of his career. He became fascinated with the trend toward data and the ability for banks to use data to drive decision-making. This led him to take graduate school classes in data science, where he learned that “not all data is created equally.” 

“There’s data that’s more important or impactful to what the business gets done and there’s availability of the data,” Lesicka explained. “The lateral is there also, and the ability to understand and analyze the strengths and weaknesses in the leasing activity for a property type for a particular area. What direction is it headed in? Are the terms getting better or are they getting worse?”

Throughout his career, Lesicka saw banks attempt to create their own internal databases to address these questions, using rent rolls and appraisal data collected throughout the institution. However, this is a very manual and tedious process. But the payoff can be huge, especially if banks could look at historical trends and use the data as a guide. Would Covid-19 have a similar impact on CRE as 9/11? Would it be similar to the Great Recession? 

Coverlease compiles CRE data from all of its customer banks across the country, anonymizes it, and makes it available for all of its users. The large dataset makes trends easier to spot. 

“As the data started coming in, you got to see what was actually happening,” Lesicka said. “And, at least from our observation, what happened was that certain property types and certain locations were doing just fine and other property types in other locations were doing terribly.” 

By allowing the comparison of CRE data from other institutions in similar geographical regions or with similar property types, banks can gain insights that inform CRE decision-making.

Lesicka compared using large datasets to driving a car. Appraisals, he noted, are always looking in the rearview mirror — back at what has already happened. In his data-driven world, the car has an augmented-reality dashboard. “You see the winding road and you see the car ahead of you, and it’s telling you the space between you and the other car, or the speed you’re going, how sharp the turn is,” he explained. “It’s giving you all that assistance so you can make a more informed decision on where you’re taking this car.” 

Banks that can effectively compile, access and utilize large datasets can have a clearer path along the road. 

Relying on APIs to connect bank systems

As loan and deposit customers make their journey through bank products and services, they may be touching several different platforms, such as the aforementioned digital account opening or loan approval process. Data can quickly become scattered throughout the bank, making the data difficult to use and even harder to see a “big picture” of the activity across the various platforms.

“There is a growing need to enable different software products to integrate with each other in [banks],” said Mike Bonner, executive director of engineering at nCino. “The promise of APIs is to extend and integrate software systems more easily,” Bonner said.

nCino is a cloud-based platform that addresses many bank needs, from digital account opening to loan origination to analytics. APIs allow products like nCino to provide integrated experiences, including document preparation and automated decisioning.

“Data, when cleaned and normalized, can provide a wealth of insight into a bank’s portfolio,” Bonner said. “Enriching that internal data with data from other sources can provide a view across industries.” 

For example, nCino relies on data for the commercial pricing and profitability modules within its product. APIs enable integrations to occur more quickly and with less investment from a bank.

The investment may cause some pause for community banks, particularly the resource investment. With the growing number of fintechs boasting API capabilities, some are good at writing APIs that can connect systems together, and some are not. A bank needs to have someone who can evaluate the strength of a vendor’s API and understand if it will fit within the bank’s business strategy.

The good news, according to Bonner, is that working knowledge of APIs is becoming more common. Graduates from engineering and computer science degree programs would understand how to use APIs. He believes that pairing a small team of engineers and bank experts on the business use of an API can allow banks to swiftly develop integrated systems as the bank implements new products. 

Hodge noted that some banks use third-party developers or platforms to do some of the integration work on a consulting basis. “The harder part is taking the data … and deciding what to do with it,” she said. “Like how to build the automation and workflows on top of the data.” That’s where an internal bank expert, who may not have technical knowledge but has knowledge of the use case for the data, will be critical.

Like Lesicka, Bonner sees the benefit of analyzing large datasets. “I think we will see more emphasis on data science to analyze the data, providing new perspectives on business processes, loan portfolios, and bank customers,” he said. “Making that information easily available … will be key.”