The Founder’s Guide to Building in Volatile Markets

Jeff Bezos has a famous saying: "Focus on what makes your beer taste better."

The idea is simple. Don't waste time building infrastructure. If you're a brewery, your job is to make great beer—not manufacture bottles or build distribution networks. Buy those. Outsource those. Focus on the thing only you can do.

In Silicon Valley, this works beautifully. The rails are stable, standardized, and reliable.

But that entire playbook assumes infrastructure you can trust. 

In Latin America, none of those assumptions hold.

Regulation changes constantly. Payment systems evolve. Government infrastructure gets digitized—or doesn't. What works in one state doesn't work in another. So the definition of core competency flips.

In established markets, your core competency is your unique product insight—the thing that differentiates you in a crowded market.

In volatile markets, your core competency is infrastructure resilience—the ability to adapt faster than regulation changes, faster than competitors can react, faster than the market shifts beneath you.

Owning your infrastructure isn't a distraction from making your beer taste better. It is what makes your beer taste better.

Let me show you how this plays out through four founders who figured that out the hard way.

THE INFRASTRUCTURE DECISION: BUILD VS. BUY

In 2014, Daniel Vogel at Bitso faces the decision every founder faces: build from scratch or buy off-the-shelf and move fast?

Bitso chose to build.

A competitor entered the market two months after Bitso launched. They took the opposite approach—licensed an off-the-shelf trading platform. Their product was better. More features. Cleaner interface. Built by a vendor who'd been doing this for years.

Bitso's product was rough. Built in-house by a team with zero trading experience.

In an alternative reality, Bitso could have been dead on arrival.

Not in Mexico.

Because while the competitor had a better product on day one, Bitso had something more valuable: control over its destiny.

The competitor had to submit feature requests to their vendor. Every change meant joining a queue. Behind other customers. In other markets. With other priorities.

Within a year, Bitso had introduced improvements that left their competitor scrambling. The competitor would see what Bitso had done and ask their vendor for the same functionality. But they were stuck waiting.

Bitso was just faster.

They killed the competition not by having better features at launch—but by having the infrastructure that let them out-iterate everyone else.

That same principle applies even when infrastructure does exist. It's just impossibly fragmented.

That's where Gringo comes in.

Brazil has more than 80 million drivers. Every single one deals with the same nightmare: documentation, registration, fines, taxes. There's a whole profession called despachantes—private agents who handle DMV paperwork for drivers. It's a massive market. Ten billion dollars. But split across thousands of mom-and-pop offices and a dozen disconnected systems.

Most new players entered through buy-and-sell. Car marketplaces. Classifieds. Financing platforms. They focused on one slice of the automotive chain. High CAC. Low margins. Endless competition.

Caique saw the gap.

Everyone was fighting for their own slice—but no one was looking at the driver as a whole. No one was connecting the full journey of car ownership.

So while other companies were going vertical—building car inventories, becoming banks, hiring armies of agents—Caique went the opposite direction.

He built a software layer that sits on top of the entire ecosystem.

Before Gringo, drivers had to navigate ten different government websites. Pay fines through the post office using paper slips. Calculate taxes on broken portals. Track renewal deadlines manually across multiple systems.

Gringo didn't rebuild any of that. They just made all that fragmented infrastructure usable and convenient.

Once you upload your vehicle documents to Gringo, it becomes your system of record. It reminds you when payments are due. Pre-fills insurance applications. Pre-qualifies you for credit. Even prices your car for resale.

You don't navigate portals. You don't track deadlines. You interact with Gringo—and Gringo orchestrates everything else in the background.

That's the moat.

THE DISTRIBUTION DECISION: DESIGN PRINCIPLE, NOT DEPARTMENT

Stay with Gringo for a second. Because there's another decision here that's just as critical.

Distribution.

For Gringo, distribution wasn't an afterthought. It was baked into the product from day one.

In Brazil, consumers don't want another app. They live in WhatsApp. They talk to friends there. Pay bills. Book appointments. Run businesses. It's not just a messaging app. It's the daily operating system.

So Gringo built for WhatsApp first. The "app" came later.

That single decision flipped the entire distribution challenge. Instead of pulling users into a new environment, Gringo met them where they already were.

And that's the first question every founder should ask about distribution: Where is your customer already living?

Not where would be easiest for you to build. Not where you want them to be. Where are they right now? Every single day?

For Brazilian drivers, it's not email. It's not app stores. It's WhatsApp.

Second question: What existing behavior can you insert yourself into?

Gringo didn't try to teach drivers a new habit—like checking a car app daily. They inserted themselves into an existing one: paying fines, renewing registrations, managing taxes.

Third question: Once you're embedded, can you become the default entry point for that journey?

Because once you're the system of record—the place where the critical data lives—customers have to come to you first. Insurance, credit, resale—every car-related action starts with Gringo.

Now here's what Caique told me:

"Most people you talk to about growth do everything—SEO, paid, branding, out-of-home, podcasts. But 80% of your acquisition is going to come from one lever in the first three to five years. Find it and double down on it."

For Gringo, that lever was organic search and word-of-mouth. Every feature they built had a distribution strategy baked in.

To me, this is one of the most underappreciated ideas in tech. Distribution as a design principle. Not a department.

In consumer, it's almost intuitive. CAC is high, attention is scarce. So the smartest consumer founders design for distribution from day one. Think Revolut's referral loops. Duolingo's streaks. Gringo building inside WhatsApp.

But in enterprise this mindset is just as critical.

Because the cost of acquiring and onboarding a single customer can exceed tens—even hundreds—of thousands of dollars. Sales cycles are long. The number of potential buyers is small.

You can't afford to bolt on distribution later. You have to engineer it into the product itself.

The most successful B2B companies—Snowflake, Notion, Figma—don't just build software. They design go-to-market architectures.

The same logic applies in Latin America. Maybe even more so. Capital is scarcer. Brand trust takes longer to build. Many enterprise buyers aren't actively looking for solutions.

So the companies that win are those whose products carry their own distribution—by embedding themselves into existing workflows, platforms, or partner ecosystems.

Whether it's consumer or enterprise, São Paulo or San Francisco—distribution isn't a department. It's a design principle.

THE VERTICAL DECISION: INEFFICIENCY AS MOAT

There's a whole cohort of Brazilian companies that have built multi-hundred-million-dollar businesses by solving problems global players can't even see. Problems that don't exist in developed markets. Problems so specific, so local, so operationally complex that they're invisible to Silicon Valley.

Take Contabilizei. They automate accounting and tax filing for Brazilian SMBs. They've raised over $185 million and now serve north of 50,000 clients.

Their entire business is built on one insight: Brazilian tax compliance is so absurdly complex that entrepreneurs spend 2,600 hours a year just managing taxes. That's seventy full work weeks. Every year. Just to stay legal.

Or Omie. A cloud-based accounting ERP for SMBs and mid-market. Today, they serve 130,000 customers, generate over $160 million in annual revenue, and recently raised $150 million in secondaries—Brazil's largest deal of 2025.

And Asaas. Payments, receivables, payables, working capital, ERP—all stacked vertically. They've raised $150 million from some of the best investors in the world and now have 37 revenue streams.

These companies aren't competing with Stripe, QuickBooks, or Salesforce. They're competing with Excel spreadsheets, handwritten invoices, and the operational complexity that crushes most small businesses before they scale.

SMBs in Brazil don't die because they lack ambition. They drown in paperwork before they can even compete.

Piero Contezini understood this deeply.

He'd spent years running a software house in Joinville, building custom systems for clients drowning in exactly this kind of complexity. So when he and his brother sold the software house and started Asaas, they didn't try to solve everything at once.

They started with one pain point: Help small businesses get paid.

Not payments plus inventory plus accounting plus payroll. Just one thing: issue invoices and actually collect the money.

They made it absurdly simple. You could issue an invoice before creating an account. They built an automated communication engine that would follow up with customers through email, SMS, even voice calls—until payment happened.

One problem. Solved completely.

Piero didn't stop there though. He didn't branch out into horizontal diversification. He went vertical instead, solving sequential, interconnected problems that made the product stickier with every layer.

Here's how the stack worked:

Layer 1: Help SMBs get paid (receivables). Getting paid is existential in Brazil. Asaas made invoicing simple and collection automatic.

Layer 2: Help them pay others (payables). Once Asaas knows when money is coming in, they can help merchants manage when money goes out—suppliers, employees, taxes—without manual reconciliation.

Layer 3: Help them manage cash flow (anticipation of receivables). Brazilian SMBs pay 20 to 30 percent interest on loans. But if you have confirmed receivables coming in, Asaas can advance that cash at more attractive rates.

Layer 4: Help them run operations (ERP and financial services). Once Asaas manages money in and money out, they expand into ERP software, credit cards, financial management tools—helping merchants run their entire business from a single platform.

Each layer made the previous one stickier.

You can't manage payables well if you don't know your receivables. You can't anticipate receivables if you're not managing payments. You can't optimize taxes if you don't have integrated operations data.

This is vertical product stacking in a market where operational complexity is the real competition.

Vertical SaaS—and vertical monetization models more broadly—represent one of the biggest untapped opportunities in Latin America. The inefficiency that makes these markets hard to operate in is the same inefficiency that makes them defensible once you figure them out.

Brazil alone has more than 20 million small and midsize businesses. Each one facing the same tax, payment, and compliance pain points.

That's not a niche. That's a massive market hiding in plain sight.

When you solve one operational bottleneck well, it opens the door to monetizing a dozen adjacent workflows. That's why companies like Asaas can build 37 revenue streams—not through aggressive diversification, but by capturing value across an entire vertical where complexity is the barrier to entry and the moat at the same time.

THE M&A DECISION: INORGANIC GROWTH AS PRODUCT STRATEGY

In Silicon Valley, early-stage M&A is met with skepticism. The belief is that great companies build, not buy. Investors worry about integration risk, cultural friction, loss of focus. In mature markets—where APIs are standardized, infrastructure is reliable, capital is abundant—that mindset makes sense.

In Latin America, the playbook is different.

Inorganic growth has become one of the most scalable and defensible growth strategies. There's a whole generation of companies that built multi-billion-dollar platforms through disciplined, strategic acquisition.

Take iFood, Brazil's food delivery giant.

Over the last decade, iFood has executed a long string of acquisitions to consolidate food delivery in Brazil – including Central do Delivery, Papa Rango, Alakarte, Hellofood Brasil, SpoonRocket’s tech team, and later SiteMercado to expand into online grocery.

And they kept going. In 2024 and 2025, iFood took minority stakes in Shopper, CRMBonus, and Saipos, tightening its grip on the broader food and restaurant ecosystem.

Today, iFood serves 55 million customers across 1,500 cities. Processes 120 million orders each month. If priced today, it would likely be worth well above ten billion dollars.

It didn't win by being first or by outspending everyone on marketing. It won by systematically acquiring competitors and capabilities, then integrating them into a single, unified platform.

And here's the irony: iFood itself was acquired early in its life by Movile—and later outgrew its parent to become one of the most aggressive acquirers in Latin America.

Olist is following a similar path. But for a different reason.

Tiago Dalvi started Olist as a horizontal marketplace aggregator—connecting small Brazilian merchants to large marketplaces like Mercado Libre, Submarino, Via Varejo. Classic aggregation play. Asset-light, software-enabled, growing fast.

He had 15,000 merchants, 700,000 products, month-over-month growth.

But the economics didn't work. Retention was weak. Margins were thin. Merchants would churn every time a marketplace changed its fees or algorithms.

Here's what Tiago realized: When you're just a horizontal layer sitting on top of someone else's platform, you have no leverage. The marketplaces control everything—the fees, the data, and ultimately, your customer relationship.

So in 2021, Olist shifted course.

It acquired Tiny ERP—one of Brazil's leading SMB management platforms—followed by Vnda (e-commerce) and Pax (logistics). Each acquisition wasn't just about scale. It was about convergence—pulling together all the disconnected pieces of a merchant's back office into a single operating system.

As Tiago put it: "Without those acquisitions, Olist would still be just another marketplace aggregator."

But here's the thing: Getting M&A right is extremely hard.

Roughly 70 to 90 percent of deals fail to deliver the value they promise. Few companies manage to merge workflows and UX into one coherent platform. Most stop at merging teams, brands, or spreadsheets.

Olist's post-merger playbook looked less like corporate finance and more like product strategy.

Here's how they did it:

Step 1: Acquire for capability, not for revenue. Don't buy a company just because it has good ARR or a complementary customer base. Buy it because it gives you a capability you need to own the full customer workflow—and you can't build it fast enough. The ERP acquisition (Tiny) became Olist's centerpiece. Not because ERP had the best margins, but because managing inventory, sales, and financials in one place transformed Olist from a marketplace connector into merchant infrastructure.

Step 2: Preserve founder talent. Keep the DNA that made each product trusted. Don't acquire and lose the people who built it. Keep them engaged and give them ownership over integration.

Step 3: Integrate the back office fast. Billing, support, ops—collapse these immediately. No customer should feel like they're dealing with three different companies under one brand.

Step 4: Collapse UX and workflows into one platform. This is the hardest part—and where most companies fail. Olist didn't keep Tiny as "Olist ERP powered by Tiny" or Vnda as "Olist E-commerce by Vnda." They integrated them so merchants experienced one seamless system, not a bundle of loosely connected products.

So remember what Daniel did at Bitso? Building infrastructure from scratch. And what Caique did at Gringo? Orchestrating infrastructure that already existed.

Olist shows how inorganic growth—when done right—can become a product strategy in itself. A form of infrastructure acceleration.

THE TAKEAWAY

For the last decade, we've been told that emerging markets are "harder" to build in. More bureaucracy. More fragmentation. More risk.

And that's true.

But we've been asking the wrong question.

The question isn't: "How do we make these markets easier?"

The question is: "What if the difficulty IS the opportunity?"

Complexity is only a barrier if you think about it as a barrier. But if you're willing to go deep—if you're willing to become the expert in something nobody else wants to understand—complexity becomes your unfair advantage.

Eighty million drivers navigating fragmented DMV systems? That's Gringo's $200 million exit to Corpay.

Merchants drowning in disconnected platforms? That's Olist's $6 billion in GMV.

Merchants who can't get paid without chasing invoices for months? That's Asaas's $150 million raise and 37 revenue streams.

Mexican crypto users with no way to convert Bitcoin to pesos? That's Bitso's $2.2 billion unicorn.

So here's my question for you: What problem are you avoiding because it's too complex? What market are you ignoring because the infrastructure doesn't exist? What opportunity are you missing because everyone else thinks it's too hard?

Because the four founders who sat across from me this quarter looked at those same "impossible" problems.

And they turned complexity into cash.

Thanks for reading,

Olga 

 P.S. If this issue was valuable to you please share it with a founder who needs to hear it. Let’s build LATAM’s next tech leaders—together

🎙 The J Curve  is where LATAM's boldest founders & investors come to talk real strategy, opportunity and leadership.