Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Fintechs Use Real-Time Restaurant Data for Dynamic Merchant Cash Advances
Transcript
- Lucas: So Luna, you know that little pizzeria on Atlantic Avenue, the one with the sourdough crust that people line up for? Luna: The one that always has a wait? Yeah, I've been there. Best margherita I've had in Brooklyn. Lucas: That place got a working capital advance last quarter. But here's the twist: the lender didn't ask for a single tax return or a personal guarantee. They looked at the restaurant's real-time point of sale data, and within twenty-four hours the money was in the owner's account. Luna: Twenty-four hours, that's incredible. So they just checked the daily sales through Square and decided on the spot? Lucas: Pretty much. And that's the world we're living in now — where your daily dinner rush is becoming a credit score. Episode one sixty-five, and we're diving into how fintechs use real-time restaurant data for dynamic merchant cash advances. Luna: Okay, but let's back up for a second. Merchant cash advance — that's the product where a lender gives you a lump sum and then takes a slice of your future card sales, right? Lucas: Exactly. It's not a loan in the traditional sense. Instead of a fixed monthly payment, the advance is repaid through a percentage of daily card transactions — usually somewhere between ten and twenty percent. So in a slow week, the owner pays less. In a great week, they pay more. The payment naturally flexes with revenue. Luna: That sounds perfect for a restaurant, where weekends are packed and Tuesdays are dead. But historically, these advances had a pretty bad reputation, didn't they? Lucas: They did. For years, merchant cash advances were the product of last resort — high effective interest rates, aggressive collection tactics, and some operators who were pretty predatory. The industry was basically unregulated, and small businesses often didn't realize they were signing away a huge chunk of future revenue until it was too late. Luna: So what changed? Why are we talking about this like it's a positive story now? Lucas: Because the data layer got much richer. Ten years ago, a lender could see a merchant's total card volume, but only with a lag of a month or more. Today, with point of sale systems like Square, Toast, and Clover, they can see daily, even hourly, sales data. They can track seasonality, peak hours, repeat customer rates — all in real time. And that changes the risk math dramatically. Luna: So the lender has a much better picture of whether the business can actually afford that advance. Instead of just looking at a bank statement from last year, they see how the restaurant is doing this week. Lucas: Exactly. And that's what we mean by 'dynamic' — the advance amount, the repayment percentage, even the duration can be adjusted based on the latest data. For example, a restaurant that's seeing a surge in reservations for the summer could qualify for a larger advance now, with a repayment schedule that stretches into the fall when sales typically dip. Luna: But wait — is this really new? I mean, lenders have been using cash flow data for a while, right? Lucas: They've used bank account data, sure. But bank account data is lumpy — it includes transfers, payroll, utilities, all mixed together. point of sale data is cleaner and more granular. It tells you exactly how much the restaurant is selling, every single day. And it's collected automatically, with no manual statements to upload. Luna: So it's not just about the data being real-time; it's about the data being more relevant to the business's core revenue stream. Lucas: Right. And that's why the big point of sale players got into lending. Square Capital started offering merchant cash advances back in 2014, using the sales data from their own payment processing. Toast, which is huge in the restaurant world, launched Toast Capital a few years later. They both have this natural advantage because they see the daily transactions flowing through their systems. Luna: So it's like they're lending against data they already own. That must make underwriting a lot faster and cheaper. Lucas: Definitely. Square Capital has issued billions in advances, and the average approval time is something like two minutes. Two minutes, Luna. The merchant applies, the algorithm looks at their sales history, and if it passes, the money hits their account almost immediately. Luna: Two minutes — that's insane. But what about a restaurant that just opened? They don't have a year of sales data. How do lenders handle that? Lucas: That's a good question. Some lenders are using predictive models that factor in the restaurant's location, menu type, and even local foot traffic data. For a brand-new place, they might offer a smaller advance based on projected sales, then increase it as real data comes in. But honestly, the sweet spot is a restaurant that's been operating for at least six months — enough data to see a pattern. Luna: Got it. So they're not really lending to unproven startups, but rather to businesses that have a track record, even if it's a short one. Lucas: Exactly. And that's a big step up from traditional banks, which often require two years of tax returns and a personal credit score above seven hundred. For many small restaurants, especially immigrant-owned ones, that's a huge barrier. They might have excellent cash flow, but no credit history in the U.S. or a thin file. Real-time data lets lenders see past that. Luna: So this is also a story about financial inclusion. But I've also heard that these advances can be really expensive. What's the typical cost? Lucas: It varies, but let's be honest: the annualized percentage rate can easily be in the double digits, sometimes even above fifty percent. That's because it's a short-term, unsecured product with higher risk. The key is that the repayment is tied to a percentage of sales, so in a bad month the restaurant isn't hit with a fixed payment it can't afford. That flexibility has value, but it comes at a cost. Luna: So if a restaurant has a slow month, they pay less, but they also end up paying interest for longer. And if they have a great month, they pay more and pay it off faster. It's like the payment exactly tracks their ability to pay. Lucas: That's the theory. And the data makes it more accurate. Some lenders now adjust the percentage — the 'holdback rate' — on a weekly basis based on the merchant's recent sales. So if sales drop sharply, the holdback might go from fifteen percent down to ten percent, automatically. That's the 'dynamic' part. Luna: I love that. But what about the risk to the merchant? If a restaurant takes an advance and then a new competitor opens across the street, their sales could tank. What happens then? Lucas: That's the downside of revenue-based financing. If sales dry up, the advance still has to be repaid — the percentage is just lower. But if the restaurant goes out of business, the lender often can't recover the full amount. That's why the real-time data is so important for the lender — they can pull back on the holdback rate or even pause collections if they see a restaurant in distress. That's a huge improvement over the old days. Luna: So they're using the data to be more proactive. Instead of waiting for the merchant to default, they can see trouble coming and adjust. Lucas: Exactly. And that's where the technology gets really interesting. Some lenders are using machine learning models that analyze not just the merchant's own sales, but also broader trends — like how other restaurants in the same neighborhood are performing, or even weather data. A rainy weekend might mean fewer customers, and the model can anticipate that and lower the holdback before it happens. Luna: Wow, so it's predictive, not just reactive. That's a real shift from the static credit models of the past. Lucas: It is. And it's not just the big players. There are fintech startups like Forward Financing and Kapitus that are building their own underwriting engines using point of sale integrations. They might not have their own payment network, but they connect to the merchant's existing system with the owner's permission and pull the data they need. Luna: So how does the restaurant owner feel about giving up that data? Is it a tough sell? Lucas: It can be. Some owners worry about sharing their daily sales numbers — it feels invasive. But the tradeoff is access to fast, flexible capital without needing a perfect credit score. And the data is usually read-only; the lender can't change anything in the system. Plus, with the rise of open banking, people are getting more used to sharing financial data for better services. Luna: That's a fair point. But there are also regulatory questions. These advances aren't technically loans, so they're not subject to the same truth-in-lending rules. Is that changing? Lucas: It is. In 2020, the Conference of State Bank Supervisors, or CSBS, issued guidance that merchant cash advance providers should be subject to state lending laws. And some states, like California, have passed disclosure requirements that force providers to clearly show the annualized percentage rate. So the industry is moving toward more transparency, even if it's not full loan regulation. Luna: I think that's a good thing. But I wonder, what about the merchant's perspective? Have there been any horror stories with these new dynamic products? Lucas: There are still cautionary tales. If a merchant doesn't understand the terms, they can get into trouble. For example, some providers still use a fixed holdback rate, and if sales drop, the percentage might not adjust, so the merchant ends up paying a big chunk of a shrinking pie. That's why it's crucial for lenders to be transparent and for merchants to read the fine print. Luna: So the data can be a double-edged sword. It helps the lender, but only if it's used fairly. And speaking of fairness, what about bias? If an algorithm is using historical data, could it inadvertently discriminate against certain types of restaurants? Lucas: That's a real concern. If the model is trained on data that reflects past biases — like lenders who approved more advances for pizzerias than for food trucks — it could perpetuate those patterns. That's why regulators are paying attention. The Consumer Financial Protection Bureau has been looking at algorithmic underwriting to make sure it's not discriminatory. So the industry needs to be careful. Luna: And that's where the conversation gets really interesting. It's not just about the technology; it's about how we use it responsibly. Lucas: Absolutely. And you know, this episode is a perfect example of why we love covering fintech — it's where cutting-edge tech meets real-world impact. If you're finding these conversations useful for what you're building or running, and you want to keep them ad-free, you can support the show at buy me a coffee dot com slash fexingo. Every little bit helps us keep diving into topics like this. Luna: Yeah, and we appreciate every one of you who's been with us for a hundred and sixty-five episodes. Now, back to the future of restaurant lending — what do you think the next big shift will be? Lucas: I think we're going to see even more integration with delivery platforms. Right now, a restaurant's sales on DoorDash or Uber Eats are separate from their in-house point of sale data. If lenders can combine those streams, they'll get an even clearer picture of total revenue. That could open up capital for restaurants that rely heavily on delivery, which is a growing segment. Luna: So imagine a ghost kitchen that has no dine-in sales at all, but huge delivery volume. With current data, they might look risky, but with combined data, they could be a great borrower. Lucas: Exactly. And I've seen some startups working on exactly that — pulling data from multiple sources into a single underwriting model. They're also using real-time data to offer 'dynamic repayment' where the holdback rate adjusts not just weekly, but even daily, based on the merchant's sales forecast. That would be the ultimate in flexibility. Luna: Daily adjustments sound amazing, but also a bit unpredictable for the merchant. How would they plan their cash flow if their payment changes every day? Lucas: That's the tradeoff. For some owners, a stable payment is easier to budget for, even if it's higher. For others, especially those with volatile sales, a dynamic payment is a lifesaver. The key is giving the merchant a choice. Maybe they can opt into a dynamic plan or a fixed plan. That's where I see the market going — more customization based on what the merchant prefers. Luna: So it's not just about the lender's algorithm; it's about giving the small business owner more agency. That's a nice note to end on, I think. Lucas: It is. Because at the end of the day, the goal of fintech should be to level the playing field for small businesses. And using real-time data to offer fair, flexible capital is a big step in that direction. Next time you order that margherita, just remember — your pizza might be helping a lender decide the restaurant's credit limit.