Pod Insights a16z Podcast a16z Podcast: Fintech for the People About this episode New fintech companies are democratizing access to financial services in different ways, whether it's making food stamps more efficient, no longer waiting two weeks for a paycheck, or enabling anyone with a smartphone in developing countries to create...
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Hanne Tidnam? The content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund. For more details, please see A16Z.com/disclosures.
0:19 Hanne Tidnam? Hi and welcome to the A16Z Podcast. This episode of the podcast is all about new fintech companies that are democratizing access to financial services in different ways, whether it's no longer waiting two weeks for a paycheck, making food stamps more efficient, or enabling anyone with a smartphone to create a small business. Moderated by Angela Strange, partner at Andreessen Horowitz, the conversation includes Jimmy Chen, CEO of Propel, Matt Flannery, co-founder and CEO of Branch, and Ram Palanyapen, CEO of Ernan, and is all about what those kinds of new fintech companies have in common besides a more inclusive approach to finance. How is innovation coming in unexpected ways from unexpected places, and what does this mean for the future of financial services as well as established players?
1:05 Speaker 01 One of the most staggering statistics is that almost 50% of Americans don't even have $400 in savings. Now, the underbanked space, lesser space, has been traditionally dominated by nonprofits and ignored by VCs, but in huge part it's because building a business here is really difficult for all the reasons that you might expect. Customer acquisition is really challenging. Once you have your customers, it's very hard to risk score them. We're joined by three companies today that are building some of the new financial services companies of the future. So Ram, Matt, Jimmy, thank you for being here. One of our premises is that there hasn't been that many great for-profit models here. And so Ram, we're going to start with the genesis of the idea and how you came up with the new business models that you're approaching in this space.
1:50 Speaker 02 I was running another company, also in the payment space in Cincinnati, and I noticed that some of my employees, particularly the ones in the call center, were getting lots of overdraft fees and short-term loans. It didn't make any sense to me because I was paying everybody well. And I didn't think that they had to go to those clearly bad financial products. And so when I spoke with one of my employees, she said the issue wasn't how much we were paying her. The issue was she needed money then. And we would pay her only the next Friday, which is how our payroll system ran. And so I wrote her a check from my account, which is, I think, what millions of small business owners do. And then on payday, she paid me back. And I did that because I owed her money for the work that she had already done. If she could, I would have had to actually pay her that day. But because she wasn't quitting, she'd have to wait till the next Friday to get paid. And so this went on for several years.
2:33 Speaker 02 We sold most of the company and moved out from Cincinnati to the Bay Area. And I had all of my employees on Messenger. And so they weren't enough. I'd still do this for them. And I didn't mind doing it. I knew how to move money. But I also needed to know if they were working or not because I wasn't part of the company. So I just had her give me the username and password to the time and attendance system. And then if she wanted money, she would message me. I would log in as her and see if she had earned as much as she wanted. And if she had that money, then I would push it out to her and then pull it back on the next payday. And so this went on for a little bit, and every time she wanted money, she'd message me. But if I went out to see someone, I'd come back, I'd have this really long list of messages. And I'd have to read through her entire day's story to figure out how much she wanted. And so I was like, this shit's not gonna work.
3:17 Speaker 02 So I built a really simple web form and said, if you want money, just go fill out that form. And then once I had that webpage out, there were people who I didn't know who tried to use it. So I said, okay, this is more interesting. Let me see if I can do it for people who I don't know. So I had them give me access to their employer's time and attendance system. And I could do the same thing for them. And I realized what I was doing was essentially what the payroll system does. Is it looks at how much you've worked, calculates what your take home pay is, and then pushes that into your bank account. But instead of doing that in a batch for everybody, I was doing it for each employee when they wanted it. And when I did this, they were paying all their bills on time. No more overdraft fees, no more payday loans. Like, their life was much, much simpler, just giving them access to the pay that they had already earned. And so that's kind of how this started.
4:01 Speaker 02 Today, we do it in a far more automated way for employees from over 25,000 companies.
4:06 Speaker 01 You're using a little bit more than a spreadsheet, probably.
4:08 Speaker 02 Yeah.
4:09 Speaker 01 Excellent. Now, Matt founded a nonprofit that you've probably heard of called Kiva, which used to do microloans in Kenya. So maybe you could talk about the opportunity that you saw moving into the for-profit space and how that's different.
4:22 Speaker 03 So my story goes back right after I graduated from college in the early 2000s. I was a Java developer at TiVo, and I was really restless with that and had a lot of ideas. And so I ended up traveling all over Africa and getting really interested in entrepreneurship there and volunteering at various NGOs, making videos of borrowers who were using a $500 loan to start a restaurant or get supplies for a pharmacy or things like this. And being an internet person, I eventually made a website for this, sort of like Rom's story, started lending to these people in Uganda, Kenya, and Tanzania over my own website and getting the money to Africa via wiring money back and forth with an NGO.
5:03 Speaker 03 And really got into this. And in about 2005, all of a sudden, it got really popular and we were on TV and we were on Oprah and we got a lot of press. And so we eventually expanded that system called Kiva to about 80 countries. We lent over $1 billion to low-income people in rural places that way. But along the way, I got really interested in the business of microfinance and what is good about it and what is not. And I realized that a lot of the people I was working with had mobile phones and digital money, and you could lend money directly to them without going through an NGO in Kenya or India or wherever. So I kind of got obsessed with this idea of lending to someone's phone and
5:44 Speaker 03 eventually left and started this new company called Branch, was essentially doing artificial intelligence-based lending to people on their phones. Right now, we work in Nigeria, Kenya, and Tanzania, and we're about to launch in India.
5:58 Speaker 01 Okay, Jimmy, some of the stats that I used in food stamps, as it's better known, $70 billion a year in one out of seven families. So you think of large markets, you would think that there'd be lots of entrepreneurs flooding into this space. That hasn't been the case.
6:12 Speaker 04 So I grew up in a loving and supportive household that also had trouble putting food on the table. I ended up getting a full ride scholarship to Stanford, where I studied CS and cognitive science, and then ultimately landed at Facebook as a product manager. In 2014, I left Facebook, looking to apply the skills of Silicon Valley tech to solve some of the social challenges that I cared really deeply about. And so moved out to Brooklyn to join a nonprofit to do a fellowship program. Through that program, started to learn about the food stamp program. And so as Angela said, it's a massive program that touches about one in seven families throughout the United States, but the user experience of it, coming from Facebook and thinking about things from the point of view of what we expect from these modern software companies, seems just totally out of whack. We learned about the program by applying for food stamps ourselves, and so my team and I went to the food stamp office in Brooklyn to apply.
6:54 Speaker 04 And one of the things that was really striking about that is that you go to that line, you see hundreds of people standing in line to talk to a human caseworker and fill out a paper form, and the majority of people are passing the time and have a smartphone in their hands. So the problem is not a hardware one, it's kind of a software one here. The other thing that we realized is that once you get through that line of applying for food stamps, you get the benefits on an EBT card. People who have their benefits on an EBT card have to call a 1-800 number to go check their balance, and then are expected to manage how they spend that balance over the next 30-day period. There are lots of challenges where lots of families run out of benefits around week two or week three of each month, and then spend the next two to three weeks of the month using food pantries, trying to borrow money from friends and family. And so we think that there's an opportunity here to apply a lot of the techniques used
7:38 Speaker 04 in modern consumer software and in modern financial services to this population. So our core app is called Fresh EBT, works kind of like a mobile banking app for somebody who's on food stamps, that helps them to manage their EBT benefits. But we really use that as a hook towards broader financial health by introducing our users to the types of products that help them to save money, to build assets, to find savings on groceries, and so on.
8:00 Speaker 01 I teed up a couple of the core challenges, risk scoring, and I want to dive in there. 90 million Americans are mispriced, a lot of customers have no credit score at all. One of the big challenges of entering this space, Ram, how do you decide who to lend money to, especially if you've never seen them before? If you don't have a credit score, what other data can you pull in that's going to give you a good indication of who to lend to?
8:22 Speaker 02 So ours is very simple. We don't say no. If you worked, you get your pay. We don't do any of the decisions. We don't ask for social security number. Don't ask for your phone number. Don't ask for your home address. It's if you worked, you get paid. We make sure that they're actually working and we calculate how much they've earned. We'll also look at their bank accounts so we can calculate what is the rate at which they earn money. So if they worked for eight hours, we know how much that is in dollar terms. And so we pull all of that data to construct how much somebody has earned. And once you've earned it, we let you take it out. And so all that you have to do as an employer is give us access to the pieces of data that we need to show that you've worked. And so as long as we can see that you've worked, then we'll let you take out money.
9:02 Speaker 01 Matt, if you're not going to link to their bank account, you need to come up with a variety of different signals. And one of the core challenges if you talk to a lot of the lending companies in the US is they'll run machine learning algorithms over all of your data. And then, let's say, for instance, I decide that I don't want to lend to this front row right here. But the front row all has purple hair, and purple hair is a protected class. Now, all of a sudden, I'm running afoul of a regulation in the US called disparate impacts. And although I might not actually be discriminating, it's going to come out that way. And so there's a lot of techniques that have been harder to use in the US. Matt, what would not be able to do if you were in the US? And do you think that's helpful to consumers?
9:40 Speaker 03 In lending in the US, there are great laws that protect people from being discriminated against in that process based on the race, gender, protected properties like that. But alternative data sets, like the data you can get on someone's mobile phone, can be extremely helpful in making more and more rational decisions about just how much money to lend to somebody at what price.
10:00 Speaker 01 What's something that you're doing specifically that you think would be harder to port over to the US?
10:05 Speaker 03 I think the main thing in the US is when you make a loan, you have to have a defensible line of reasoning for that that's easily explainable to a regulator. Whereas when you're using deep learning or artificial intelligence, different methods of machine learning, it's really hard to articulate exactly why someone got exactly this loan. And robots would be much better at making that decision than people. In the end, it's a very mathematical decision that's multi-dimensional. And so you can't really explain to the state of California why exactly this person got this loan or why this person exactly got rejected because it's just a deep learning network that makes that decision.
10:37 Speaker 01 Jimmy, I want to flip over to consumer acquisition. And we have this expression that we use a lot, which is the battle between the incumbent and the startup is whether the incumbent can get innovation before the startup gets distribution. And this applies to any industry, but especially financial services. Consumer acquisition is very, very difficult. Talk about how you've managed to scale and grow a fairly difficult-to-acquire customer.
10:58 Speaker 04 Yeah, I think our lens on that problem is a little bit, it's kind of odd, because the incumbent, in this case, is actually a partner of ours, in the sense that it is a government program that we're providing a skin on top of. So we're not trying to displace the incumbent necessarily. We're trying to augment the service that the incumbent offers, as we think there's a lot more to gain from the consumer's point of view. We bring a lens of consumer software that's kind of intuitive in Silicon Valley, but far from intuitive when you talk to state governments, when you talk to federal governments, and the folks who run programs like the food stamp program. And so what we try to do is kind of blend both. We try to blend the best of the food stamp program with kind of the philosophies around how you can make something that people really want, and that solves a real consumer need, and then bring that to the market at scale. We try all different types of acquisition strategies, but the most important of them is word of mouth.
11:41 Speaker 04 So we have a million monthly active users across the country now who use Fresh EBT at least once a month. These are all families that are on food stamps. About 60% of them found the app through some friend or family member telling them that this is something that they should try. We think that's the core, kind of, core signal that we're solving a problem that real people have, that they're willing to promote us to their friends and family.
12:00 Speaker 01 And how do you get over, you know, like many other early fintech companies, consumers are generally skeptical of the banking services in general. How do you think about getting over that hurdle, or what have you encountered?
12:11 Speaker 04 Solve a real pain point for people. We started with a very niche pain point of, it's hard to check my balance on my EBT card. I've got a call this phone hotline, it takes a couple of minutes, and it's like, it's not the worst thing, but it's kind of annoying. If we can take that to a three-second thing that you can check on your smartphone, the same way that you would check your card balance for your Chase debit card, we think that that kind of is enough to get somebody to engage in the product. And then our opportunity is to use that entryway to create real value for people who are lower income.
12:38 Speaker 01 Rum, have you found consumer trust to be a hurdle in the growth of Vernon?
12:41 Speaker 02 Yeah, I mean, so our product, when people hear about it the first time, it's you can get your pay whenever you want, and we don't charge any fees or interest, so we let the users just pay whatever they would like to pay. So it's get your pay whenever you want, and pay whatever you'd like for the service. They're super skeptical of it. They're like, this sounds too good to be true, should be a scam. But reality is like what they're used to, overdraft fees and payday loans, like, those are too bad to be part of our future. And so we're not going to compromise the product because it's much better than what people expected. I think as they see others using it, there's much more acceptance of it. And then they completely hate the other products that were always taking advantage of them.
13:15 Speaker 01 And so how do you get over that hump, though? Is it word of mouth?
13:18 Speaker 02 So from the way we get customer acquisition, word of mouth is very, very big. There's also a really strong community aspect to the product because we let users pay whatever they think is fair. Very often, users will cover the cost for another user. And so if I go and I offer a tip more on the app, it'll show me that I'm covering the transaction cost for another user. And as soon as the other user comes in, it's going to say, "I'm already paid for you. I'm already covered the cost of your transaction." And then that person can send me back a thank you message. And so every day, there's thousands of thank you messages going up and down between our users. On the acquisition side, too, we actually put it back in the community. So if like, if you got one of your friends to use the product, we're going to say Angela got one of her friends to use the product. And people from the community can actually thank Angela with money. And so like, even our acquisition
14:02 Speaker 02 is funded by the community.
14:04 Speaker 01 Yeah, so creating this element of social proof to help bootstrap new brands to come up. So I want to talk a little bit about the future and how you guys see this playing out. There's lots of nonprofits trying to help this space and now starting to be more smart entrepreneurs coming in and creating new business models. Do you think, like, is this a fundamental shift we're seeing where it's expensive to be poor is not going to be the case kind of 10 years from now? And if so, like, what needs to happen?
14:30 Speaker 03 Yeah, I'll just say some things that are really obvious to me, probably obvious to every mood, but really important. First of all, we have this whole emerging market, which is growing and a very young group of people. And everyone's getting on smartphones, and it's going to be really hard to buy a feature phone in the coming years. At the same time, most of these people are going to have access to some sort of digital wallet or digital bank account. Yet, there's no credit card infrastructure, no credit bureaus or traditional infrastructure to build out a lending layer on top of that. So credit cards just aren't going to spread. You know, the middle class of India is not going to have plastic credit cards. Nigerians are not going to have Visa cards or Mastercards anytime soon, because there's nowhere to swipe them, and people can't get credit scores. So it's gonna be through apps.
15:15 Speaker 03 And so the credit card of the rest of the world is gonna be an app. It's really obvious and the plastic cards are not gonna get there. So there's just a huge opportunity to layer on other services on top of credit. So we're just using credit to acquire millions of users. And then hopefully you can add to the things like payments, savings products, money transmission, remittances, etc.
15:34 Speaker 01 Yep, so new platforms is a way to accelerate it. Jimmy, what do you think is gonna be? Like, are we at an inflection point that we're gonna start to see more business models in and around your space?
15:43 Speaker 04 I think we're getting there. I think that the problem's not gonna solve itself. And I also think charity alone is not gonna be the solution to solving the financial needs of low-income Americans. I think charity certainly has a role to play, to play that role. But I think for-profit companies can play a really unique role in this space of helping low-income Americans just understand different opportunities for how they can improve their own financial health. I also think it requires us to be a little bit more creative, requires us to think of different models of business and how to generate revenue, such that it doesn't pollute the goals of what we're trying to build.
16:13 Speaker 01 So Ram, you've tackled payroll, which has been incredibly rigid since the beginning of time. Do you see that these sort of sticky, resistant-to-change models are eventually gonna get pressured into new models?
16:24 Speaker 02 So first of all, I wouldn't say that payroll's been rigid since the beginning of time. I think the concept of this rigid paycheck is only a few centuries old. I mean, till a few centuries ago, people were actually paid every day. As a society, we did not accept that people could go home without their pay. There's references to this in the Bible where you're required to pay employees before sunset. Clearly, employees would rather get paid before they worked, and companies would rather pay the employees after they worked. And the fact that all companies today pay employees after they work just shows you where the balance of power is, that it's so skewed towards the companies away from the individuals. But it wasn't always like that. There's different categories of incumbents. And each one of them have got institutionalized and have got things that keep them in place.
17:06 Speaker 02 Like, you could look at banks, and banks have so many advantages. They have access to payment networks. As a startup, you can't get into payment networks. They have access to capital. As a startup, we really can't take deposits, so they got really low-cost capital. They've got privileged access to regulation, where they can get a charter, and they can do all these things, where we have to go get licensed for these things one by one. So there's like a huge hurdle for us. Even if you just look at the whole concept of like pay cycles, that's been institutionalized as well, because lots of your overtime calculations are actually on a weekly basis or a two-weekly basis. That makes it a little bit more complicated now for you to do daily payroll, because... We built these laws after we were used to getting two-week pay cycles. And so the payroll companies that are used to doing all the compliance are used to calculating
17:48 Speaker 02 overtime on a weekly or a two-weekly, whatever it is for that particular state. So lots of these things, I think, are institutionalized. For sure, we'll all try to do what's right for the customer and see if the incumbents either do the same thing or they try to stand in the way.
18:01 Speaker 01 We'll see the reaction. Excellent. I want to thank Ram, Matt, and Jimmy for joining us, and thanks, everyone, for being here.
Previous a16z Podcast: Of Governors and Mayors, and Tech Policy Next a16z Podcast: Revisiting the Gene Page updated September 30, 2026 · transcript and summary are machine-generated