There’s a detail buried in the Stripe success story that most people skip over.
Patrick and John Collison grew up in Dromineer, County Tipperary. Population: around 100. No tech scene. No startup culture. Their dad ran a hospitality business. Their mom was a microbiologist who once ran for the European Parliament. By most measures, this is not the origin story you’d write for two people who would go on to process $1.4 trillion in payments a year.
But here’s the thing about where you grow up when you’re a builder. It doesn’t matter that much. What matters is that Patrick taught himself to code at 10, sold his first company at 16, and by the time he was in college had already decided that interesting problems were more valuable than degrees. John was the same.
They sold a company called Auctomatic for $5M before either of them was legally old enough to drink in the US. Then they went back to college, got bored, and started poking at a problem that everyone else in Silicon Valley had apparently decided was too annoying to fix.
Accepting money on the internet was, in 2009, an absolute mess.
The Problem Was Hiding in Plain Sight
This is one of those things that’s obvious once someone explains it, but somehow nobody had fixed it. To accept payments online in 2009, you needed a merchant account. Getting one meant paperwork, waiting periods, faxing documents, and praying a bank liked you. Once you had one, you had to integrate a legacy API that felt like it was designed by people who genuinely disliked developers.
PayPal existed and was the default, but anyone who worked with it knew the developer experience was bad. Not “could be better” bad. Actually bad. The kind of bad where you spent more time fighting the integration than building your actual product.
Patrick and John’s read on this was different from most people’s. Most people looked at this situation and thought “payments are complicated, that’s just how it is.” The Collisons looked at it and thought “this is a code problem. The plumbing is bad. We can fix the plumbing.”
That reframe is where the whole company came from.
Their original product, which they built while hacking away in Buenos Aires after getting into Y Combinator in 2010, let you accept payments in 7 lines of code. That was the entire pitch. Not 7 lines and then three days of configuration. Seven lines and you were live.
They named it /dev/payments. Had to rename it Stripe later for trademark reasons. The name changed. The idea didn’t.
How Stripe Got Its First Customers
The early customer acquisition story is one of the most instructive parts of the payments startup playbook that Stripe built without really meaning to.
During YC, Patrick and John did something that Paul Graham later turned into a thing he talked about in essays. When another YC founder agreed to try Stripe, the Collisons didn’t say “great, we’ll send you a link.” They said “hand me your laptop.” They would sit down, right there, and get the founder live on Stripe before they left the room.
Graham called this the “Collison installation.” The broader lesson is one of the most valuable ones in early-stage startups: don’t optimize for scalability when you’re trying to get your first 100 customers. Optimize for getting them actually using the product. Do things that don’t scale. Get in the room.
Beyond the hands-on installs, Stripe got their next wave of customers through a single Hacker News post by YC partner Garry Tan. Something along the lines of: if you need to accept payments, email these people. That post landed them somewhere between 300 and 500 signups. They had to implement a waitlist because account setup was still manual at that point.
They also sent swag kits to every person who processed their first transaction. Not just a welcome email. Actual physical swag, shipped to the address they had on file from signup. Small thing. The kind of thing that makes someone tell three people about it.
None of this is a payments startup playbook you’d find in a textbook. It’s just being genuinely good at making people feel like they found something special.
The Funding Story Nobody Makes a Big Deal About
Stripe’s seed round was under $100K from YC and Paul Graham. For a company that would go on to be valued at $95B at its peak, that’s a funny place to start.
The $2M round came after Patrick got a meeting with Peter Thiel at a YC dinner. The story goes that the brothers spent the meeting explaining in detail every structural flaw in how PayPal had been built. Thiel, who co-founded PayPal, found this compelling rather than annoying. He wrote a $200K check at the first meeting. Sequoia, Andreessen Horowitz, and SV Angel joined the round.
Series A was $18M from Sequoia in early 2012. By that point, Stripe was already processing significant volume and growing mostly on word of mouth inside developer communities.
From there the cap table reads like someone just invited every top-tier firm in the valley: Series B at $80M in 2014, Series D at $150M in 2016, Series E at $245M in 2018, all the way to a $600M round in 2021 at a $95B valuation.
What’s worth noting is that the Collisons were not chasing the highest valuation at every round. They consistently chose investors they felt understood the long-term bet. Building financial infrastructure for the internet is a 20-year project. You want investors who think in those timescales.
The Stripe Success Story Is Really a Platform Story
Here is where most writeups get the Stripe success story wrong. They tell it as a payments story. Better product, better developer experience, better pricing transparency than the incumbents. All of that is true and also misses the point.
Stripe’s real success is that they used payments as a beachhead to build an entirely different thing: a financial operating system for internet businesses.
Watch how the product line expanded:
Stripe Connect launched to handle payouts in marketplace businesses. If you’re building something like Lyft or a freelance marketplace, you don’t just need to accept money. You need to split it, hold it, route it to different parties, and comply with regulations in multiple countries while doing all of that. Connect solved that. One integration, massive complexity handled underneath it.
Stripe Billing launched for subscription management. Not just charging a card on a recurring basis. The whole stack: proration when a customer upgrades mid-cycle, dunning when a card fails, invoice generation, revenue recognition support. The kind of stuff that SaaS companies were building in-house and hating every minute of.
Stripe Radar launched for fraud detection. The interesting thing about Radar is that it gets better as more businesses use Stripe. Every transaction across the entire Stripe network improves the model. It’s a data flywheel built into the product. Smaller competitors can’t replicate this because they don’t have the transaction volume to train against.
Stripe Atlas launched to let founders incorporate a US company, open a bank account, and issue stock in a single afternoon. This is not a payments product at all. But it puts Stripe in the room at the exact moment a company is born. The first thing a newly incorporated company needs to do is figure out how to accept money. Stripe is already there.
Stripe Capital launched to offer loans to businesses based on their transaction history. If you process $3M a year through Stripe, they know more about the health of your business than any bank does. They used that knowledge to underwrite credit that businesses actually needed.
Each of these products deepens the same relationship. None of them require Stripe to become a different company. They all sit on top of the same infrastructure layer. That’s the strategy. You build the layer once, then you build a lot of things on top of it.
By 2024, the billing and subscription suite alone was tracking toward $1B in annual run rate. Stripe Capital generated $420M in interest revenue. These are not payments numbers. These are platform numbers.
How Stripe Actually Sold to Developers (and Why It Worked Later)
This is the part of the payments startup playbook that takes years to pay off, which is why most companies don’t do it.
Stripe spent years selling almost exclusively to developers, who at the time had limited or zero budget authority. A developer at a startup would integrate Stripe because it was the best tool for the job. No procurement process. No enterprise contract. Just a developer who wanted to build something and found a product that didn’t make them want to throw their laptop out the window.
The reason this mattered later is compounding familiarity.
That developer became a senior engineer. Then an engineering manager. Then a CTO. By the time they had a real budget and were evaluating enterprise payment infrastructure for a company processing hundreds of millions of dollars a year, they had used Stripe for a decade. Their teams were fluent in it. The mental switching cost was real.
Stripe’s developer NPS has consistently been among the highest in fintech. That’s not a vanity metric. It’s a measure of how embedded the product is in the professional identity of the people making technical decisions at companies with real money.
Most B2B companies would look at this strategy and dismiss it because the timeline is too long. Sell to people who don’t have budget now and hope they have budget later? That’s a hard pitch to a board that wants revenue next quarter.
But the companies that play this game well end up with moats that are genuinely hard to attack. By the time a competitor shows up with a comparable product, the people they’re selling against have spent years with Stripe. Switching has real cost.
Going Upmarket (and Why the Timing Was Right)
For roughly the first six years, Stripe didn’t have a meaningful enterprise sales motion. The product sold itself to startups and developers, and that was enough to build a very large business.
In 2016, they changed that.
They hired Jeanne DeWitt Grossman as the first Head of Revenue for North America. What she did when she arrived is the kind of thing that sounds obvious but almost never happens in practice. She spent her first couple of months working as an account executive, closing deals herself on the front lines. Not observing. Not doing listening tours. Actually selling.
The reason this mattered is that selling payments is genuinely different from selling most software. The value proposition compounds with transaction volume. The pricing model means the customer’s costs scale with their success. The technical integration is sticky in ways that typical SaaS isn’t. Understanding this at the level of detail required to actually close enterprise deals required being in the room when those deals were happening.
Only once she understood what the enterprise motion actually looked like did she build a team around it.
The results are visible in the numbers. Half of the Fortune 100 now processes payments through Stripe. NVIDIA. Pepsi. Ford. News Corp. Comcast. These are not companies that stumbled onto Stripe through a Hacker News post. They were sold.
The timing of the upmarket push also matters. Stripe waited until the product was genuinely enterprise-ready before putting sales resources behind it. Going upmarket too early, before the product can handle the complexity and compliance requirements of large organizations, is how you get churned enterprise contracts and a reputation as a startup tool that doesn’t scale. Stripe avoided that by being patient.
What the R&D Obsession Actually Means
In Stripe’s 2024 annual letter, Patrick and John Collison made a point of noting that in each of the last six years, Stripe had reinvested a higher proportion of earnings into R&D than any comparable company.
For a company processing $1.4 trillion a year, this is a deliberate choice. Most companies at that scale are optimizing margins, buying back stock, running lean. Stripe is still acting like they’re behind.
The R&D investment is showing up in a few places. Their fraud detection is significantly better than most alternatives because of years of model training on real transaction data. Their global infrastructure, covering 135+ currencies and most countries on earth, required years of investment in local banking relationships and compliance frameworks that competitors simply haven’t matched.
And the AI infrastructure is becoming a real business. Over 700 AI agent startups launched on Stripe in 2024. Stripe has built a toolkit specifically for developers building AI agents that need to autonomously manage financial transactions. Subscriptions managed by agents. Refunds processed without human involvement. ElevenLabs used it to let a voice agent handle subscriptions and refunds on its own.
If agentic AI becomes as embedded in business software as people expect, Stripe wants to be the financial layer underneath it. That’s the same bet they made with SaaS in 2012. Build the infrastructure layer before the market demands it, and be ready when it does.
The Stablecoin Angle
The longer-term bet is stablecoins, and it’s worth taking seriously.
Stripe has been investing in stablecoin infrastructure quietly for a while. The argument is that if programmable money becomes a real part of how businesses move funds globally, the company best positioned to build the API layer for it is the one that already built the API layer for the current system.
Cross-border payments are still genuinely painful in most of the world. Expensive, slow, opaque. Stablecoins, if the regulatory environment allows it, could make this dramatically better. Stripe has been positioning to be in the middle of that transition rather than reacting to it.
Nothing is certain here. Stablecoins have been three years away from mainstream adoption for about a decade now. But Stripe’s pattern has always been to invest before the market demands it. The R&D obsession. The Atlas product. The early agentic toolkit. This is the same instinct applied to a longer timeframe.
The Lessons Other Companies Keep Missing
Strip away the mythology and the Stripe success story comes down to a handful of things that are easy to describe and genuinely hard to execute:
They found a problem where the existing solutions were legitimately bad. Not “could be improved” bad. Actually bad. That gave them a quality bar that was easy to clear and a customer frustration level that made switching easy to justify.
They sold to the people who would have budget later instead of the people who had budget now. Took years to pay off. Paid off enormously.
They built a platform instead of a product. Every new thing they launched deepened the customer relationship rather than just adding features. The ceiling on a platform is much higher than the ceiling on a product.
They waited to go upmarket until they deserved enterprise customers. Patience is underrated.
They kept reinvesting in R&D when the natural instinct was to optimize margins. The compounding on that is showing up now.
And Patrick and John are still running the company. The people who understood the original insight are still making the decisions. That’s not nothing.
Where This Ends Up
Stripe processed $1.4T in 2024. That’s roughly the GDP of Australia flowing through an API that started as 7 lines of code written by two bored teenagers in South America.
The next version of the story is about AI agents, stablecoins, and whether Stripe can do to the financial infrastructure of the next internet what it did to the financial infrastructure of this one.
If you’re betting against it, you’re betting against the same founders, the same product instinct, and the same R&D discipline that got it here.
That’s a tough bet to make.
Leave a Reply