Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Bank Cut Loan Approval to 28 Minutes
Transcript
- Lucas: So there's this regional bank in the Midwest, about forty branches, and they just cut their small-business loan approval time from two weeks down to twenty-eight minutes. Luna: Twenty-eight minutes? That's like, you apply over lunch and you have an answer before you finish your sandwich. Lucas: Exactly. And it's not a tech unicorn, it's a traditional bank that's been around since the 1960s. They didn't build a new app from scratch. They re-engineered the underwriting process itself. Luna: So what was the old process? Two weeks seems like an eternity for a small business waiting on capital. Lucas: It was paper-heavy. The applicant would fill out a stack of forms, bring in tax returns and bank statements, and a loan officer would manually verify everything. Then it went to a credit analyst, then a committee. If anything was missing, it went back to the applicant. That back and forth alone ate up days. Luna: And I'm guessing the new process digitizes all that. Lucas: Right. They moved to a fully digital intake. The application is online, and it pulls data directly from the business's accounting software and bank feeds. That eliminates the manual data entry and verification step entirely. The system also runs a credit check automatically. Luna: So the human is out of the loop for the initial screen. Lucas: For the initial screen, yes. The new model scores the application based on cash flow, credit history, and a few other factors. If it passes a threshold, it moves to a loan officer for a final review. If it doesn't, it's declined automatically — though the applicant can appeal. Luna: That's a big shift. But what about fraud? Quick approvals usually scare me because they feel like they might skip checks. Lucas: That's a fair concern, and the bank thought about it. They built in a fraud-detection layer that flags anomalies — mismatched documents, suspicious IP addresses, those kinds of things. If the system flags anything, the application goes to a manual review team. So the fast lane is only for clean, straightforward applications. Luna: So the average time dropped, but the human still has a role. Lucas: Exactly. And that's the key operational insight. They didn't just automate everything. They kept human review for the edge cases where judgment matters. Luna: And I love that you mentioned the appeal process. That's a nice touch — it gives the customer a path if the algorithm says no. Lucas: Yeah, and it also gives the bank a second look at borderline applications. A loan officer can override the system if they see something the model missed. That's how you get the best of both worlds. Luna: So what's the actual impact? Has it moved the needle for the bank? Lucas: The bank says loan applications are up about 30 percent since they rolled this out. And the approval rate actually ticked up slightly — they're not just approving more, they're approving better. Luna: Interesting. So the speed didn't come at the cost of credit quality. That's usually the fear. Lucas: Right. Their default rate hasn't budged, which is remarkable. They attribute that to the richer data they're pulling from accounting software — they see real-time cash flow instead of a snapshot from a tax return six months old. Luna: So the data is actually better, not just faster. Lucas: Precisely. And that's the lesson for any operation: you can often improve speed and quality simultaneously if you replace manual data entry with direct data feeds. Luna: It's like the difference between reading a company's financials off a PDF and getting a live API feed. Lucas: Exactly. And the ripple effects go beyond the loan department. The bank's small-business customers can now make decisions faster because they know quickly whether they have the capital. Luna: And that's got to improve their customer relationships, too. Business owners hate waiting. Lucas: It does. They survey customers after every loan, and satisfaction scores are up sharply. The number one complaint used to be the wait. Now it's not even in the top three. Luna: So what's the catch? There's got to be a downside. Lucas: The catch is the upfront investment. They spent about two years and a significant amount of money building the integration with accounting software and training the risk model. It's not a plug and play solution. Luna: And I'm guessing the human team had to change too. The loan officers' roles probably shifted. Lucas: They did. Instead of data entry, loan officers now focus on relationship management and complex cases. The bank retrained everyone. And they did have some attrition — not everyone wanted to change how they worked. Luna: That's real. Operational change is never just a tech project. Lucas: And that's why this case is so instructive. It's not about the twenty-eight minutes. It's about the deliberate design of a process that uses automation where it's strong and humans where they're strong. Luna: And that's a lesson that applies way beyond banking. Lucas: Absolutely. Every business has a process where a customer is waiting on a decision. The question is, can you give them an answer faster without making worse decisions? Luna: So what's the one thing you'd take away from this episode? Lucas: Start with the data. If you can get the data automatically, you can cut the wait. The rest is just workflow design. Luna: And if you're thinking about your own loan process, maybe start by mapping where the data sits and who's moving it by hand. Lucas: That's the first step. Thanks for listening, everyone.