He runs a Uruguayan company that generates more than a billion dollars in annual revenue, although many people use its services without knowing its name. Pedro Arnt is fine with that. Meanwhile, he keeps repeating a phrase to his engineers: “Excuse me, a basic question from someone who isn’t technical.”
A student in Lagos wants to pay for a subscription. A woman in Quito wants to subscribe to Netflix. The owner of a small business in Montevideo wants to use Spotify. They have the money, but the platform does not always accept the way they can pay.
A few years ago, that could force them to borrow an international card or activate a VPN so the system would think they were in the United States. The payment method they used for everyday purchases did not work there.
dLocal handles that problem. If the payment works, the buyer has no reason to know that the company was involved. That is why Arnt says it is almost a good thing that people do not know who they are.
When Google, Netflix, Spotify or Amazon enter a country, they need to accept the payment methods used there. In Uruguay, that might be Oca or a local bank transfer; in Brazil, Pix or Mercado Pago. dLocal connects those platforms with the systems of sixty markets. With a single integration, its clients can operate in different countries without having to address each case on their own.
Arnt says the big digital companies encountered that difficulty when they began expanding. In their home markets, collecting payments with a credit card was largely a solved problem. They had not expected payments to complicate their growth so much.
He has been dLocal’s CEO since March 2024. The company ended 2025 with revenue of $1.093 billion, up 47% from the previous year, and processed around $40.8 billion in payments. Uruguay accounts for less than 1% of its revenue.
Before that, he had spent twenty-five years at Mercado Libre. For the last twelve, he was its chief financial officer.
Tropicalization
When asked to summarize what he learned at Mercado Libre, he begins by pointing out that they also did many things wrong. There are quite a few he would recommend not repeating.
Among the successes, he singles out one he calls “product-market tropicalization.” They closely followed what was happening elsewhere in the world because many of the problems in e-commerce were similar. But to solve them, they had to understand how people shopped and paid in each country, and how they could receive their orders.
For years, they heard that Mercado Libre would do well until Amazon arrived. Later, the warning shifted to Chinese companies. Arnt attributes much of what they achieved to their knowledge of the markets where they had started. Competitors could arrive with resources and experience, but understanding those countries took time.
The other lesson was to think long term while also handling everyday decisions. They tried not to take a shortcut to solve an urgent problem if it would cause trouble later.
That is how they ended up entering logistics and financial services. Arnt believes neither decision would have made sense if they had looked only at the results of the next twelve months.
At dLocal, he took over a company that others had founded and brought to Nasdaq.
It had been established in Uruguay in 2016 and went public five years later. As it grew, its founders discovered that running a global company was not necessarily a job they wanted to do. Arnt arrived in 2023 and shared the leadership with Sebastián Kanovich, CEO from the beginning. In March 2024, he took charge.
He recommends that founders considering leaving that position develop a successor within the company. Bringing someone in from outside can be difficult. He compares it to performing a lobotomy on the organization: removing one brain, putting in another and expecting everything else to keep working.
Sharing the role with Kanovich allowed them to get to know each other through work. Each could see how the other made decisions, discuss what did not sit right and make adjustments before completing the transition.
It also helps reveal that things will not work out, he admits. In that case, it is better to find out early.
For Arnt, there is an important difference between leaving the leadership because someone feels they are not doing it well and leaving because they no longer want the job. When the founder wants to take that step, they have more reasons to help it succeed. If the proposal comes from the board and the founder is not convinced, everything becomes more complicated. At dLocal, he says, there was a willingness to make the change.
The hardest part was explaining that some ways of working also needed to change. The company had grown and was delivering good results. There was no crisis he could point to as justification.
The Basic Question
Arnt acknowledges that his lack of technical training has always been a disadvantage in his work. He feels it more now because fewer and fewer areas of the company do not depend on technology.
He tries to learn by using the tools, reading and listening to podcasts. Sometimes he encounters explanations that are beyond him and looks for something he can understand before moving on.
He also turns to the people around him. “Excuse me, a basic question from someone who isn’t technical: can you explain this to me?” he asks them.
He had to get used to asking without worrying so much about what it might say about him. Being CEO does not mean he understands everything. The engineers he works with know things he needs to learn, and talking with them is one of the parts he enjoys most.
The tools have also changed how he finds out what is happening in the company. Previously, he would ask a vice president, who would consult a manager. The answer might come back three hours later. Now, seven times out of ten, he finds what he needs in an internal tool.
When he thinks about that, he wonders how a CEO managed before email. He imagines someone receiving information late, fairly isolated from what was happening, yet still having to make decisions. A man in a glass castle.
Africa
Arnt says African markets have taught him the most. There is an enormous opportunity, but also difficulties with technological infrastructure and frequently changing regulations.
Operating there requires finding solutions that can later be useful in other countries of the Global South. The interviewer reminds him that there is no single Africa. He accepts the point: differences between countries matter a great deal.
The acquisition they made on the continent was linked to what they expect over the coming years. Arnt looks at the size of the countries, their young populations and the growth of the middle class. Digitalization still has some way to go, and some of those changes will take time. Looking ten years ahead, he sees enormous potential. They bought to maintain the position they had already established.
There is something he learned while opening new markets that he recommends paying particular attention to: the first hire in a country.
Finding that person is difficult. The company has just arrived, nobody knows much about it as an employer, and it cannot always attract the people it wants. But the first person in charge will choose the rest of the team. If that choice goes wrong, the problem can spread.
He explains it through a familiar classification: A players hire people better than themselves; B players hire people a little worse; C players hire people much worse. A weak initial hire can end, a year later, in a board meeting to explain why the expansion did not work.
At dLocal, they also had to reconsider what kind of experience they needed. They sought professionals with established careers and knowledge of the regulator, but the results were not what they expected. They did better with people who had started businesses. Those people understood the speed at which they needed to work and were accustomed to building something, with the problems and uncertainty that entails.
They found several of those people through Endeavor. In one market, they had three leaders who did not work out before finding the right person.
Working with teams in more than forty countries also changed his opinion of cultural differences. At Mercado Libre, he heard that Colombians were different from Mexicans and thought it was an exaggeration. Now he sees how much culture matters when leading people.
It shows when pointing out a mistake or saying that something is not working. A way of speaking that might be commonplace in the United States or Germany can be poorly received elsewhere. Explaining that it is a work matter and should not be taken personally does not always help.
Even so, he believes a team needs to be able to talk about its problems.
He wants people to feel confident enough to report what is going wrong, ask for help and listen to an uncomfortable observation. Also, to acknowledge what someone else has done well. He does not expect everyone to speak the same way, but he does want them to be able to say things to each other.
He dislikes “corporate athletes,” those who show only their good results. And he wants anyone to be able to challenge an idea, regardless of their position. Often, he says, the best proposal came from someone with less experience.
Making that happen has been one of the hardest tasks.
What Is Not Being Discussed
Arnt believes artificial intelligence could be the most important revolution humanity has faced. What strikes him is that almost the entire technology conversation, and much of investors’ interest, has become concentrated there.
It seems as though nothing else were happening. People also talk about winners and losers as if they could already know what will happen in each sector. He believes there will be many more nuances.
What concerns him most is employment. He is not convinced that as many jobs will be created as disappear, even though that happened in previous industrial revolutions.
He has a feeling that many more could be lost this time. He makes clear that he does not know. He believes technology will generate wealth, but wonders how it will be distributed and what will happen to those who lose their jobs.
He sees little discussion of those questions, either in companies or in the public sector. He suspects many of those who should be thinking about what to do have not yet noticed the speed of change. Those who use the tools see it with every new version.
He notices it himself when looking at dLocal and calculating how much its cost structure could change. He clarifies what he means: he is also talking about the number of employees.
He is also concerned about the money the United States and China are putting into computing and electricity infrastructure. He does not know how countries unable to invest on that scale will keep up.
Over the past twenty years, he recalls, several emerging countries managed to narrow the gap. Now he fears they will fall far behind again. “I don’t know how we avoid falling way behind,” he says.
He used to advise his children on which career to choose. Now he prefers not to. Changes are happening so quickly that he no longer knows whether what he would recommend today will be useful when they start working.
He does want them to learn to manage when things become difficult and to start again if necessary. He believes they will have to do that more often than he did. His generation, he says, lived through a fairly favorable period: many years of peace and technological advances that improved life without bringing major social problems.
He does not expect his children to have the same stability. He still believes that, with time, a better world is possible. But he does not imagine a smooth path to get there.
That is why he settles on three recommendations: learn to work with others, be able to recover when things go wrong and stay curious.
