April 1, 2014. Three founders in a São Paulo diner, testing whether the credit card they'd spent a year building actually works.
David Vélez swipes. Declined. Cristina Junqueira swipes. Declined. Third try, Edward Wible's card goes through for a snack.
This was Nubank's first transaction. A decade later, the company had signed up more than 100 million customers across Brazil, Mexico, and Colombia; was listed on the New York Stock Exchange above $40 billion; and had become the first digital bank outside Asia to hit that milestone. By mid-2026, its customer count has climbed past 125 million, with a market cap around $66 billion.
Financial services in Latin America was one of the most profitable, least competitive industries on the continent, run by a handful of banks that never had to compete for a customer in their lives. David Vélez had spent years at Sequoia Capital watching technology disrupt every other industry except this one. Cristina Junqueira had just walked away from a five-year career at an incumbent bank, worn out from trying to fix it from inside. Together with engineer Edward Wible, they set out to build what nobody else in the region would touch.
Let’s look at the story with a bigger picture and see what you can learn from it.
Why they started with the riskiest product on the shelf
Neobank playbooks in Europe and the US open with a debit card or savings account, asking a skeptical customer to trust a brand new company with their money on day one. Nubank flipped the order by launching a no-fee credit card first, which accomplished three things at once.
- It put the risk on Nubank's side of the table. A credit card means the company is lending money and absorbing default risk, an easier first yes for a customer with every reason to distrust a bank.
- It created a rapid data loop, since credit spending and repayment behavior show up in weeks, feeding underwriting models faster than a savings product could and letting Nubank approve people traditional banks had never served.
- It meant launching without a banking license, since credit cards sat outside the rules a deposit-taking institution would trigger.
This combination let a twelve-person team, mostly engineers and one designer, get a real product into real hands for under a million dollars, zero of it spent on marketing.
The part where hatred became a growth channel
When we say "zero-dollar marketing," we don’t just mean it for fun. Latin American banking customers had spent decades absorbing high fees and poor service with nowhere else to go, so the moment a genuinely fee-free alternative showed up, they told their friends unprompted. Nubank hit one million customers within two years of launch, three years ahead of its forecast, almost entirely through word of mouth.
The deeper lesson here is that when a category has trained customers to expect mistreatment, removing that mistreatment works as a marketing budget on its own. The product doesn't need to be ten percent better. It needs to remove a pain point people have been quietly furious about for years.
Patience as a moat
The other decision that gets underplayed is what Nubank refused to do. The company stuck to one credit card, in one country, for roughly four to five years before adding a savings account in 2017.
Debit cards, personal loans, investments, SME accounts, and insurance followed only once the previous product had proven itself at scale. Nubank's founders refer to this as earning the right to build the next thing. International expansion followed the same pattern: the company waited until 2019 to enter Mexico and Colombia and still operates only in those two markets today.
Their discipline paid off during a genuine existential threat. A few years in, Brazilian policymakers floated a rule cutting the standard thirty-day merchant settlement window down to two days, a change that would have forced the entire card industry to raise billions in fresh funding overnight just to keep operating. Nubank spent weeks walking regulators and the central bank through the mechanics of settlement timing until the policy got shelved.
Cristina Junqueira signed her Series A paperwork the day after birthing her first daughter in a hospital bed because the company couldn't stay quiet about the round any longer. She now has three daughters, each born during a different milestone: the Series A, the Mexico launch, and the IPO. It says a lot about how much of this growth was built by people playing a long game.
Founder takeaways
If you are a founder building a company of your own, there is a lot to learn from Nubank’s story:
- Put the risk on your side of the table first. If your product asks a skeptical customer to trust you with something valuable on day one, the sale gets harder than it needs to be. Look for the version where you absorb the risk.
- Look for categories running on stored up resentment. The cheapest growth channel available is a customer base that already hates every option in front of them. Removing one pain point can outperform a marketing budget.
- Earn your next product before you build it. Skip launching five things at once. Prove one product deeply, then use that proof to justify the next.
- Treat your data loop as part of the product. Nubank's credit card doubled as rapid feedback for underwriting. Ask what your first product teaches you about your customer that the second one will need.
- Show up in the room before the crisis hits. Nubank's regulatory relationships were strong enough to get a real hearing when an existential rule change appeared out of nowhere. Build that credibility early.
Know a founder who needs to read this? Forward it their way.
