Most fintech growth advice revolves around chasing conversion rates and optimizing funnels. But operators who have scaled companies like Nubank, Rappi, and Yuno from zero to tens of millions of users point to a different reality: the distribution channel dictates everything, and payment infrastructure becomes the hidden growth ceiling no one warns you about.
The counterintuitive claim? Going mobile-first isn't about responsive design or app-store optimization. It's about recognizing that acquisition happens where attention lives—and in 2024, that's overwhelmingly on phones. The second surprise: payment processing, the invisible plumbing of fintech, determines whether you scale or stall. Growth teams that ignore infrastructure constraints end up fighting battles they've already lost.
Acquisition Happens on the Device You're Already Holding
Jeremy Goillot, former Head of Growth at Spendesk and now CEO of The Mobile-First Co., learned this the hard way. Running acquisition campaigns for a B2B spend management platform, he noticed something obvious in retrospect but invisible at the time.
The majority of my ads at Spendesk were viewed on mobile. So the term 'mobile first' means recognizing that the way to acquire users today is on mobile, because screen time—where we spend the majority of our time—is on the phone.
— Jeremy Goillot, CEO at The Mobile-First Co.
This wasn't a design philosophy. It was a distribution insight. Goillot watched campaign after campaign deliver impressions on mobile devices, only to send users to desktop-optimized experiences. The friction killed conversions before product quality ever mattered. His new company raised $12 million in seed funding (with Lightspeed Ventures backing) to build telephony infrastructure that assumes mobile as the default—not the afterthought.
The shift from desktop-first to mobile-first acquisition mirrors a broader pattern in fintech: distribution determines product strategy, not the other way around. Companies that build for where users already are—on phones, in messaging apps, embedded in existing workflows—skip the expensive step of changing user behavior.
Payment Declines Are a Growth Problem Disguised as a Tech Problem
Julian Nunez, founder of Yuno, saw the payment infrastructure trap from inside Rappi. When the company launched in Colombia, it connected to a single payment processor. Growth was explosive—20-30% week-over-week—but 30-40% of transactions were declining. The team had no clear explanation why, and it didn't matter at first. There were bigger fires to put out.
When Rappi started in Mexico, we realized the processor we used in Colombia didn't even exist there. We had to connect to another company. And they also declined 30-40% of payments. We went to Brazil—identical story. Argentina, Chile, everywhere.
— Julian Nunez, Founder at Yuno
The pattern repeated in every new market. Different countries meant different processors, each with its own approval rates, compliance requirements, and alternative payment methods. Rappi's solution was brute force: connect to multiple processors in each country, build retry logic, and hire hundreds of people to manage the complexity. The payments team became one of the largest functions in the company.
Nunez realized other high-growth companies were building the exact same infrastructure in parallel. The pain was universal, but each company treated it as a proprietary problem. Yuno's thesis was simple: abstract the entire payments layer into a single API. Connect once, access any processor, any payment method, any fraud tool globally. The company now processes tens of billions of dollars annually.
The lesson for growth operators: payment approval rates aren't just a finance metric. They're a growth lever. A 10-point improvement in approval rates can translate directly to 10% more revenue without changing acquisition spend or product. Most teams don't measure it until it's already a bottleneck.
Product-Market Fit in Fintech Means Communicating the Fit, Not Just Having It
Nubank's expansion into Mexico offers a case study in how product-market fit evolves across markets. The company had 4-5 million customers in Brazil before entering Mexico, but the value proposition that worked in one market didn't automatically translate.
Ivan Canales, who served as Nubank's Country Manager in Mexico, points to a dual challenge: finding the fit, then communicating it.
There are two things. One, having that fit, but equally as important, knowing how to communicate it. Many of the things Nubank has done have genuinely transformed how financial products work in Mexico.
— Ivan Canales, Former Country Manager at Nubank Mexico
Nubank added 10 million customers in Mexico in roughly three years—a pace that required both product iteration and a go-to-market strategy tailored to local context. The playbook from Brazil didn't copy-paste. Regulatory environments differed. Consumer trust in digital-only banks varied. Competitor responses weren't uniform.
Canales also noted that Nubank imported Silicon Valley best practices but adapted them. The company's technical rigor around hiring—finding people technical enough to execute but flexible enough to operate differently—created a culture that could move fast without breaking compliance or customer trust.
The growth insight here: market expansion isn't about exporting a fixed product. It's about re-proving product-market fit in each geography and then building the narrative that makes the value legible to new users. Fintech companies that skip the communication step end up with great products no one understands.
Mobile-First Distribution Demands Mobile-First Infrastructure
Goillot's decision to build a new telephony company stems from observing a gap: most business communication tools were built for the desktop era. Google Meet added AI transcription and summaries, but no one had applied that to traditional phone calls—the primary communication channel for millions of businesses.
We've added AI in two ways. First, all your calls are recorded, transcribed by AI, and summarized by AI. You get a ChatGPT version of all your calls so you don't forget anything. People have had this on Google Meet for 2-3 years, but no one had done it for classic telephony.
— Jeremy Goillot, CEO at The Mobile-First Co.
The Mobile-First Co. raised its seed round on a Notion deck, not a PowerPoint. Goillot was more comfortable in Notion, and the depth of thinking mattered more than the format. Lightspeed Ventures and 30 business angels invested $3.5 million before the product launched. The bet wasn't on traction—it was on distribution thesis and founder credibility from scaling Spendesk.
This mirrors a broader trend in fintech infrastructure: the companies winning today are the ones building for the actual behavior patterns of 2024, not the workflows of 2014. Mobile-first isn't a buzzword; it's a constraint. If your acquisition, onboarding, and core product experience don't assume a phone as the primary device, you're building for a shrinking audience.
The Hidden Cost of Building Payment Infrastructure In-House
Rappi's payments team grew to hundreds of people. Yuno's entire business model is predicated on the idea that this shouldn't be necessary. The disagreement isn't about whether payment infrastructure matters—everyone agrees it does. The question is whether it should be a competitive advantage or a utility.
Nunez argues the latter. For most companies, payments are essential but undifferentiated. Building and maintaining integrations with dozens of processors, managing compliance across jurisdictions, and optimizing approval rates is expensive and distracting.
We saw many other companies building very similar infrastructure and trying to solve the same problem in very similar ways. There should be a single infrastructure that connects companies with the entire payments ecosystem—processors, payment methods, anti-fraud tools, everything—so companies don't have to focus on building this, but on building what they're meant to do: their business, delivering value to their customers.
— Julian Nunez, Founder at Yuno
The counterargument, implied by companies like Rappi that did build in-house, is control and customization. When payments are core to your business model—especially in markets with high fraud or complex regulatory environments—outsourcing can introduce risk. The build-versus-buy decision depends on whether payment processing is a strategic differentiator or operational overhead.
For most fintechs, the answer is shifting toward "buy." The ecosystem has matured. Tools like Yuno, Stripe, and others have absorbed enough complexity that building in-house is harder to justify unless you have unique requirements or massive scale.
What Actually Compounds in Fintech Growth
The common thread across these operators: distribution and infrastructure compound faster than product features. Goillot's mobile-first thesis is about owning the acquisition channel. Nunez's payments abstraction is about removing infrastructure drag. Canales' Nubank experience highlights that product-market fit is as much about communication as product quality.
The growth patterns that win in fintech aren't the ones taught in most growth playbooks. They're not about A/B testing button colors or optimizing email subject lines. They're about recognizing where users already are (mobile), removing the invisible bottlenecks that kill conversion (payment declines), and adapting both product and narrative to each new market (localized product-market fit).
Goillot's advice from Andreessen Horowitz applies here: if you don't have a problem you're passionate about solving, join a high-growth company and learn how they execute. The operators scaling fintech today didn't invent new growth tactics. They identified structural advantages—mobile distribution, payment infrastructure, market-specific positioning—and built around them.
The takeaway for growth operators: the next bottleneck isn't in your funnel. It's in the infrastructure you're not measuring and the channels you're not prioritizing. Fintech companies that win are the ones that treat distribution and payments as first-class growth problems, not afterthoughts.