Latest / Startup Stories with Fexingo: Conversations About Founders, Funding, and Building Companies from Zero / How Stripe Became the Operating System for Internet Commerce
Transcript
- Lucas: It's hard to remember now, but in 2010, if you wanted to accept payments on a website, your options were basically PayPal, which everyone hated, or setting up a merchant account with a bank, which took weeks and required faxing forms. Luna: Faxing. In 2010. Lucas: Right. That was the reality. And then two brothers from Ireland, Patrick and John Collison, decided the whole thing should be seven lines of code. That's the origin of Stripe, and today it processes over a trillion dollars a year in payment volume. Luna: A trillion with a T. And they're still not a household name the way PayPal is, even though they power Amazon, Lyft, DoorDash, and millions of smaller businesses. Lucas: That's the fascinating part. Stripe deliberately stayed behind the scenes. They built for developers, not for consumers. Their pitch was never 'Stripe is the best way to pay.' It was 'Stripe is the best way to build payments.' That developer-first approach created a network effect that traditional payment companies couldn't replicate. Luna: Let's go back to the beginning. Patrick Collison was a math prodigy who dropped out of MIT. John was even younger, also dropped out. They sold their first startup for five million dollars as teenagers. Lucas: Yeah, that was Auctomatic, a tool for eBay sellers. They made a few million, then started working on what became Stripe. But the real insight came from trying to build other products. Patrick said that every time they tried to launch something new, the hardest part was not the product itself — it was setting up payment processing. Luna: So they built the solution for themselves first, and then realized everyone else had the same problem. Lucas: Exactly. And they didn't just make it slightly easier. They made it drastically easier. Before Stripe, integrating payments required a merchant account, a payment gateway, PCI compliance paperwork, and usually a week of engineering work. Stripe collapsed that into a single API call and a few lines of JavaScript. Luna: That's the 'seven lines of code' pitch. I remember reading that their first documentation was literally a single page with a curl example. And it worked. Lucas: That minimalism was intentional. The Collisons understood that developers are lazy in the best way — they want the path of least resistance. So Stripe made the path of least resistance also the path to using Stripe. And once a developer integrated Stripe into a product, it became very hard to switch because you'd have to rewrite all that payment logic. Luna: That's a classic lock-in, but it's earned lock-in. They actually built a better experience, not just a contractual one. Lucas: Right. And they kept improving. In 2016 they launched Stripe Atlas, which lets anyone incorporate a Delaware company and open a US bank account from anywhere in the world. That was huge for international founders. Then they added Stripe Connect for marketplaces, Stripe Capital for lending, Stripe Issuing for virtual cards. They basically became the operating system for internet commerce. Luna: And all of that runs on their own infrastructure. They don't outsource processing to a legacy bank. They built their own payment processing network from scratch, which gives them better control over reliability and fraud detection. Lucas: That's a key differentiator. Most payment companies are just a layer on top of existing bank rails. Stripe built its own connection to the card networks — Visa, Mastercard, American Express — directly. That means they can offer better uptime, faster settlement, and more granular data. When you see a dashboard that shows you exactly why a payment failed, that's Stripe's data model at work. Luna: And that data has another use: fraud detection. Stripe Radar uses machine learning on that massive dataset to block fraudulent transactions in real time. They claim it saves merchants billions annually. Lucas: It's a flywheel. More volume means better fraud models, which means lower fraud rates, which attracts more volume. That kind of network effect is very hard to compete against. It's why even though there are dozens of payment startups now, Stripe's market share keeps growing. Luna: Okay, but they're not public yet. There's been talk of an IPO for years. In 2025 they did a private market transaction that valued them at around 27 and a half billion dollars. That's down from their 2021 peak of 95 billion. What happened? Lucas: Two things. First, the broader fintech correction. In 2021, every payments company was overvalued. After the Fed started raising rates, growth stocks got hammered. Stripe's revenue growth slowed from over 50 percent to the mid 20s. That compression affected their valuation. Luna: And second, they took a deliberate stance. The Collisons didn't want to go public at the peak. They wanted to go public when they felt the company was ready, not when the market was frothy. So they did a tender offer in 2025 at that lower valuation to give employees liquidity, but they didn't file for an IPO. Lucas: That's a very founder-friendly move. Most boards would have pushed for an IPO at the peak to cash out early investors. But the Collisons control the company through dual-class shares. They've prioritized long-term independence over short-term market timing. Luna: And that independence lets them make other contrarian bets. Like Stripe Treasury, which is basically banking as a service. Or Stripe Climate, where they let businesses automatically donate a fraction of revenue to carbon removal. These are long-term plays that public markets might punish. Lucas: The climate thing is interesting. Stripe was one of the first companies to make carbon removal purchases at scale. They committed to spending at least 1 million dollars a year on negative emissions technologies back in 2020. That's not a marketing stunt — it's a real procurement budget. Luna: And they opened it up through Stripe Climate so any business can participate. That's consistent with their strategy: take a capability they built for themselves and make it a product. Lucas: Exactly. That pattern repeats everywhere. They built their own accounting system, then launched Stripe Revenue Recognition. They built their own tax calculation engine, then launched Stripe Tax. They built their own identity verification system, then launched Stripe Identity. Every internal tool becomes a product. Luna: That's a powerful R&D model. You solve your own problems first, and if the solution is good enough, you sell it. It keeps the product team grounded in real needs. Lucas: I think that's a big reason why Stripe has survived while other fintech unicorns have struggled. They didn't go after flashy consumer apps or crypto gambling. They stayed focused on the boring, essential plumbing of online commerce. And boring, essential plumbing is very sticky. Luna: Speaking of sticking around — if these conversations have moved your work forward in some small way, there's a simple way to make sure they keep coming. Lucas: We keep Fexingo ad-free, and the only reason we can do that is because some listeners chip in a couple of dollars a month. It genuinely makes a difference — covers hosting, transcription, the things that keep a small independent show running. Luna: If you've gotten something out of these episodes, buy me a coffee dot com slash fexingo. That's it. No tiers, no rewards, just knowing you helped keep the lights on. Lucas: Yeah. And now back to Stripe — one more thing I want to touch on: their international expansion. Stripe is now live in over 40 countries, but they've been very deliberate about which markets they enter. They didn't just launch everywhere at once. Luna: Right, because payments regulation varies wildly. In Europe you have PSD2 and strong customer authentication. In India you have UPI and the RBI's data localization rules. Stripe had to build local payment methods — like iDEAL in the Netherlands or Alipay in China — while complying with local laws. Lucas: And they did it by acquiring local companies. They bought Paystack in Nigeria for 200 million, which gave them a foothold in Africa. They bought TaxJar for tax automation. They bought Recko for reconciliation. Each acquisition filled a specific gap in their infrastructure. Luna: The Paystack acquisition is a good example. Stripe could have built from scratch in Africa, but it would have taken years. Paystack already had the relationships with African banks and the regulatory approvals. Stripe basically bought a shortcut. Lucas: And they let Paystack operate fairly independently. The founders stayed on, the brand stayed. That's another lesson from Stripe: when you acquire a team, don't smother them. Let them keep doing what they're good at. Luna: So what's the biggest risk Stripe faces now? They have a trillion dollars in volume, but the payments market is still huge and competitors are getting smarter. Adyen is strong in Europe. Square is strong with small businesses. And then there's the possibility that Apple or Google build more aggressive payment products. Lucas: I think the biggest risk is actually regulatory. Stripe is increasingly acting like a bank — offering lending, issuing cards, holding funds. That invites more scrutiny from regulators around the world. If they're classified as a systemically important financial institution, the compliance costs could eat into their margins. Luna: They've been preparing for that. They hired a former FDIC chair to their board. They've been building out their compliance infrastructure. But it's a real headwind. Lucas: Still, if you look at the trajectory, Stripe has built something remarkable. They started with a simple idea — make payments suck less — and turned it into the infrastructure that powers a huge chunk of the internet economy. And they did it while staying private, staying independent, and staying focused on developers. Luna: It's a playbook that a lot of founders study, but very few can execute. The combination of technical excellence, product taste, and patience is rare. Lucas: And it's not over. They're investing heavily in AI tools for payments, like Stripe Sigma which lets businesses query their transaction data with natural language. They're also working on Stripe Connect for embedded finance. The next decade could be even bigger than the last. Luna: So maybe the question isn't whether Stripe will go public, but whether it needs to. It's already the operating system for internet commerce. Going public might just be a formality. Lucas: Or it might never happen. The Collisons have said they want to build a company that lasts a hundred years. An IPO is a tool, not a goal. And as long as they can keep raising private capital and giving employees liquidity, they might stay private indefinitely. Luna: That would be unprecedented for a company of that scale. But Stripe has always done things differently.