Latest / Startup Stories with Fexingo: Conversations About Founders, Funding, and Building Companies from Zero / How Nubank Built a 100 Million User Bank Without Branches
Transcript
- Lucas: So picture this: it's 2013 in São Paulo, and a Colombian venture capitalist walks into a bank branch to open a simple checking account. He waits an hour, fills out mountains of paperwork, and when he finally gets a number, it's a 40 percent annual interest rate on credit. Luna: Forty percent? That's not a typo? Lucas: Not a typo. That experience was the founding moment for Nubank, of course. David Vélez, the VC in question, realized Brazil's banking sector was a perfect target for disruption — five giants controlled over 80 percent of the market, margins were obscene, and customer service was abysmal. Luna: And he decided to take them on with… a purple card and no branches. Lucas: Exactly. The purple Mastercard with zero annual fees. That was the entire initial product. No checking account, no savings, no loans — just a credit card you applied for through an app, managed entirely on your phone, with real-time spending notifications. Luna: Which sounds obvious now, but in 2013, Brazilian banks were still charging you for SMS alerts. Lucas: Right. And Nubank's innovation wasn't really about technology. It was about a radically different cost structure. Traditional banks in Brazil have a cost to income ratio around 60 to 70 percent — meaning they spend that much of their revenue on branches, staff, and legacy systems. Luna: And Nubank? Lucas: Consistently below 40 percent. Sometimes as low as 35 percent. That allowed them to offer zero fees and still make money on interchange from merchants — roughly 2 to 3 percent per transaction — plus interest on revolving balances. Luna: But isn't zero-fee credit card a loss leader? How do you make that work unless you have insane volume? Lucas: You need scale, and Nubank got it. They crossed a million customers in three years, then ten million in five. By early 2026, they're at over 100 million customers across Brazil, Mexico, and Colombia. That's the largest digital bank in the world by customer count. Luna: And they achieved that without ever opening a single physical branch. Lucas: Correct. Every acquisition was digital, basically word of mouth driven by a referral program — you refer a friend, you both get something like a higher credit limit or a small reward. Their net promoter score hovered around 90, unheard of in banking. Luna: But there's a catch, right? All those customers, they started with just a credit card. Converting them into full banking relationships — checking, savings, loans — that's where the real economics live. Lucas: Exactly. And Nubank's cross-sell engine is actually quite good. Once they have you on the card, they offer you 'NuConta' — a no-fee digital checking account that yields a bit of interest. Then they layer on personal loans, insurance, and even a crypto trading feature launched in 2023. Luna: And I think that's where the Brazil-specific regulatory environment helped. The central bank launched Pix, a real-time payment system, in 2020. Nubank embraced it fully — you could send money instantly, for free, through their app. That locked users in. Lucas: Pix was a windfall. Nubank processed over 3 billion Pix transactions in 2025 alone. Every time a customer uses Pix through Nubank, they're interacting with the app, seeing offers, building trust. The data from those transactions also helps Nubank assess creditworthiness better than traditional bureaus. Luna: So they're using behavioral data rather than just credit scores. Lucas: Right. And that allowed them to lend to millions of Brazilians who were previously 'unbanked' or underbanked — people the incumbents wouldn't touch. Nubank's default rates actually ended up lower than the industry average because their models were better at predicting repayment based on transaction history. Luna: Okay, but expansion into Mexico and Colombia hasn't been as smooth, has it? Lucas: It's been more challenging. Mexico's banking oligopoly is even more concentrated than Brazil's — two banks control about 60 percent of the market. And regulatory fragmentation across Latin America means you can't just copy-paste the Brazilian playbook. You need local licenses, local compliance, local partnerships. Luna: Still, they've gotten to about 10 million customers in Mexico. That's not nothing. Lucas: And they're profitable in Mexico as of last year. The overall company turned its first net profit in 2022 and has been in the black since. For a fintech that spent years burning through venture capital — over $1.5 billion raised from Sequoia, Tencent, and others — that's a big milestone. Luna: It's worth noting that Nubank went public on the NYSE in December 2021, at a valuation of over $40 billion. The stock had a rough 2022 — like most growth stocks — but it's recovered nicely. Lucas: Yeah, it's been a volatile ride. But the underlying business metrics are solid: over 50 percent of their customers use Nubank as their primary bank now. And they've started layering on higher-margin products like installment loans and a 'credit card with rewards' that has an annual fee. Luna: So the strategy is: hook them with free, then upsell them on premium. Lucas: Exactly. It's the classic product-led growth funnel, but applied to banking. And it's working. Their average revenue per active customer has been climbing steadily, from about $6 per month in 2020 to over $10 now. Still low by US standards, but in Brazil that's real money. Luna: You know, this kind of deep dive into how a company actually built its economics — it's exactly why people listen to this show. I know I personally find it really useful for thinking about my own work, even if I'm not launching a bank. Lucas: That's nice of you to say. And honestly, if today's episode moved your thinking forward even a little, if it was worth a coffee to you, there's a link — buy me a coffee dot com slash fexingo. It's the smallest possible way to keep these conversations coming, completely ad-free. Luna: Yeah, listener support is what lets us spend the time digging into these stories. No pressure, just if it's useful. Lucas: Right. So back to Nubank — the big question now is what comes next. They've conquered credit cards and basic banking. The next frontier is insurance and investments. Luna: Insurance in Brazil is famously opaque. Lots of fine print, slow claims. If Nubank can simplify that, it's a huge market. Lucas: Exactly. They launched Nubank Vida — life insurance — in 2023, with a fully digital claims process and no medical exam for policies under a certain amount. Early numbers suggest decent uptake. And their investment platform, Nu Invest, lets you buy Brazilian government bonds, ETFs, even crypto. Luna: So they're trying to become the full financial super-app. Lucas: That's the ambition. And their biggest competitive advantage is the data moat. They know where you spend, when you get paid, how much you save, who you send money to. That's incredibly valuable for underwriting and for targeting offers. Luna: But incumbents aren't standing still. Itau, Banco do Brasil — they've all improved their digital offerings. And there are other fintechs like C6 Bank and PicPay nipping at their heels. Lucas: True, but Nubank has two advantages: brand trust and cost structure. Their NPS is still miles ahead of any traditional bank. And they don't have the legacy branch network dragging down their margins. That means they can compete on price and still be profitable. Luna: So the takeaway from Nubank's story isn't just 'digital banking works.' It's that a relentless focus on the customer experience, combined with a fundamentally cheaper operating model, can topple an entrenched oligopoly. Lucas: Exactly. And it took a decade. David Vélez and his co-founders, Cristina Junqueira and Edward Wible, were patient. They didn't try to boil the ocean — they started with a single product, perfected it, then expanded. That's the template. Luna: And it's a template that's now being replicated across Latin America. I'm thinking of Ualá in Argentina, Albo in Mexico, other challengers. Lucas: Right. But Nubank is the proof of concept. If you can do it in Brazil — with its complex tax system, high inflation history, and regulatory hurdles — you can probably do it anywhere. It's a genuinely inspiring startup story. Luna: Alright, I'm curious: if you were starting a fintech today, would you still do a credit card or would you pick a different wedge? Lucas: That's a great question. I think the wedge depends on the market. In a place with high interchange fees, credit cards make sense. In low-interchange markets like Europe, maybe you start with payments or savings. But the principle holds — start narrow, nail it, then expand. Luna: And that's a lesson that goes way beyond banking. Lucas: It does. For now, Nubank's 100 million customers are a testament to the power of simplicity, transparency, and operational discipline. What do you think, Luna — is there a comparable story in another industry that we should cover? Luna: I've been thinking about how Mercado Libre essentially did the same thing for e-commerce and payments in Latin America. Maybe that's one for a future episode. Lucas: Good call. We'll put that on the list. For today, thanks for listening.