Your app has a problem

 This Week’s Essay

I asked Guilherme Horn, who runs WhatsApp across Brazil, India, and Indonesia, one question on stage at South Summit Porto Alegre.

What business is WhatsApp actually in today?

He didn’t hesitate.

“We’re not competing with SMS or email anymore. We’re competing with iOS, Android, ERPs, and CRMs.”

Let that sink in.

The app your customers open before they get out of bed is now the layer where business actually happens. Where your dentist confirms your appointment. Where your bank processes your Pix. Where your client sends the brief and your accountant invoices you.

All of it. One thread.

Nobody has put a clean dollar figure on the WhatsApp economy in Brazil yet. That’s not because it’s small. It’s because it’s everywhere — woven into transactions, customer service, commerce, and communication in ways that are hard to separate from the economy itself.

This didn’t happen because someone planned it.

It emerged from a collision: near-total penetration, a business culture built on personal relationships, a population that communicates by voice by default, and Pix making transactions instant and free.

No coordinated strategy. Just the right infrastructure meeting the right behavior at the right moment.

The result is a country that accidentally built the clearest preview of where global commerce is heading.

A world where the storefront is a conversation and the checkout is a message.

The number most AI companies still haven’t absorbed:

Brazil sends 4x the global average in audio messages on WhatsApp.

Four times.

Any AI model that doesn’t handle voice natively is largely irrelevant here. The interface isn’t text-first. It was never text-first. And that shapes everything — from how AI agents are trained to how 150 million people communicate with businesses every single day.

Pix solved payments. Now the battle moves upstream.

More than 30 banks now offer Pix directly inside WhatsApp. The transaction is no longer the friction point.

So what’s the next frontier? Everything before the payment.

Discovery. Trust. The conversation before the conversion. That’s where AI agents are quietly taking over. Horn shared the Magalu example — a customer opens WhatsApp, describes their living room, asks what size TV fits the space. Gets an answer with pricing in seconds.

No app. No search bar. No waiting until Monday.

Horn called it a return to something old.

“It’s very similar to a very old experience — when you used to go into a store and talk to someone and ask for advice. Now you can do that through WhatsApp and AI agents.”

AI isn’t replacing human sales. It’s restoring the personal, conversational buying experience that scaled commerce stripped away.

Brazil got there first. Because the infrastructure was already in place.

Every business is about to become 24/7. Operating hours as a concept is ending.

A small retailer in São Paulo with no IT team and no budget can deploy a conversational agent today that handles inquiries, qualifies leads, and closes sales at 2am on a Sunday.

The moat that larger businesses held by having more people — more coverage, more availability — is evaporating.

Within months, consumers won’t accept a business that goes dark on weekends. It’ll feel as strange as a website that only works on weekdays.

The biggest AI mistake most companies are making:

Treating it as an efficiency play.

Horn was blunt: AI isn’t about efficiency. It’s about redesigning the whole business. The companies falling behind aren’t behind because they lack tools. They’re behind because they’re layering AI on top of broken processes and calling it transformation.

Horn called the cost of this “cultural debt.”

It compounds quietly. And eventually it becomes existential.

The companies winning aren’t running pilots. They’re rebuilding. Top-down. Total. Starting from the CEO.

If you’re building the next hot AI-powered app — read this before you write a line of code.

First question: how often will someone actually open it?

Frequency is everything. Low-frequency products don’t build habits. Products that don’t build habits don’t survive.

Map the user journey honestly. How many times a week does your user need what you’re offering?

Once a month or less?

You don’t have an app problem. You have a distribution problem.

And in Brazil, that distribution problem has already been solved for you.

WhatsApp isn’t just where your customers live. It’s where they’re making decisions, asking questions, and sending money. The smartest GTM move right now isn’t a growth hacking playbook. It’s showing up in the thread where the conversation is already happening.

Let WhatsApp be your distribution before you earn the right to ask someone to download something.

The platforms aren’t coming for your customers.

They’re already there.

This Was Missing

Next Tuesday, we’re launching TJC Debrief.

A monthly show where we break down the most important deals, IPOs, and power shifts in Latin America — and place them in global context.

Hosted by yours truly and Paulo Passoni, Partner at Valor Capital Group.

 This Week’s Number

 4X

Brazil’s audio message volume versus the global WhatsApp average. Any AI product entering this market without native voice handling is entering the wrong market.

Trends I Am Watching

1. 24/7 isn’t a startup thing anymore. It’s coming for your doctor, your lawyer, and your accountant.

Healthcare. Legal. Financial advice. Education.

These sectors have operated on appointment-based, office-hours models for decades.

You want to see a doctor? Book two weeks out.

You want legal advice? Call during business hours.

You want to talk to your accountant? Leave a voicemail.

AI agents are about to make that feel as outdated as a fax machine.

The incumbents aren’t ready. The startups that move first will eat their lunch.

2. The one-person company is no longer a lifestyle business.

Matthew Gallagher built Medvi — a GLP-1 telehealth company — with $20,000 and a suite of AI tools.

First full year revenue: $401 million.

Net profit margin: 16.2%.

Full-time employees: two.

Sam Altman once ran a betting pool with fellow tech CEOs on when the first one-person billion-dollar company would arrive. Dario Amodei put his money on 2026.

Medvi just made it a case study.

If one person can generate $400M in revenue with margins that would make most public companies envious, the entire framework of how we think about team size as a proxy for ambition, fundability, and scale needs an upgrade.

3. Creator and media acquisitions are about to have a moment.

OpenAI just acquired TBPN, a daily tech and business podcast.

It won’t be the last move like this.

In the age of AI, distribution and audience trust are the scarcest assets — and they can’t be generated by a model. For companies with billions in capital and a product that needs to reach people, acquiring media for hundreds of millions of dollars isn’t a content strategy.

It’s the fastest way to buy what takes years to build.

Expect more of this. A lot more.

What I’m Loving 

“Sam Altman May Control Our Future — Can He Be Trusted?” — Ronan Farrow & Andrew Marantz, The New Yorker

A year and a half of investigative reporting. Hundreds of pages of memos from people inside OpenAI — including Dario Amodei. The conclusion, per a former board member: “He has a relentless will to power. He’s unconstrained by truth.” Whether you think this is a hit piece or a necessary reckoning, it’s the most important profile written about anyone in tech in years. Required reading.

→ Read it here

“TBPN and the Rise of the Tech-Friendly Talk Show” — The New Yorker

OpenAI just acquired TBPN for hundreds of millions of dollars. The New Yorker’s piece on why is sharper than most coverage — it’s not really about a podcast. It’s about who controls the narrative around AI and why tech companies are now buying the microphone instead of talking into someone else’s. Connects directly to this week’s trends.

→ Read it here

20VC: Andrew Dudum, Founder of Hims & Hers — Harry Stebbings

Hims & Hers was down 66% six months ago. Now it’s at a $4.3BN market cap on $2.3BN revenue and just acquired its largest global competitor for $1.5BN. Andrew Dudum’s conversation with Harry Stebbings on what happened, why brand beats performance marketing, and how AI is reshaping customer acquisition is one of the best founder conversations I’ve heard this year. The Medvi story makes a lot more sense after you listen to this.

→ Listen on Spotify

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.