Patrick Collison: Seven Lines of Code That Changed the Internet Seven Lines of Code

In 2010, if you wanted to accept credit cards on your website, you were in for a bad week.

You filled out a merchant account application that read like a mortgage. You got underwritten by a bank that may or may not call you back. You wired paperwork to a processor. You integrated against a gateway whose documentation had last been meaningfully updated during the Bush administration — the first Bush administration. If you were lucky, you were taking live payments by the end of the month. If you were unlucky, your account got frozen on the first big sale and nobody would pick up the phone.

This was not a secret. Every engineer who had ever shipped a product online knew it. The whole industry knew it. The industry had simply decided this was how things were.

And then a 22-year-old from a village of a few hundred people on the shore of Lough Derg opened his documentation with seven lines of code, and the entire premise of the business quietly fell apart.

Dromineer to Dublin to Cambridge

Patrick Collison was born in Limerick in September 1988 and raised in Dromineer, County Tipperary. His father Denis trained as an electronic engineer and ran a hotel. His mother Lily is a microbiologist who ran a training business from the house. Patrick was the oldest of three brothers, all of whom would eventually make headlines of one kind or another.

He took a computer course at the University of Limerick when he was eight. He was programming at ten. At sixteen he built Croma, a Lisp-like programming language, and submitted it to Ireland's Young Scientist and Technology Exhibition, the national research competition that has run every January since 1965. He won. President Mary McAleese handed him the cheque — 7,500 euros — and a Waterford Crystal trophy.

He had sat the SAT at thirteen. By 2006 he was at MIT studying mathematics, having arrived on campus at seventeen. He left within months to work on Auctomatic, an eBay seller tool he was building with his brother John, three years his junior, and two co-founders. In March 2008 Live Current Media bought Auctomatic for a reported $5 million in cash and stock. Patrick was nineteen. John was seventeen.

He went back to MIT. He left again.

The idea that was almost too obvious

The story Collison has told in interviews is that he and John kept trying to spin up small side projects — the kind of weekend experiments that are the lifeblood of software — and kept running into the same wall. Taking money on the internet was unreasonably painful. Not technically interesting. Not an unsolved research problem. Just procedurally awful, in a way that filtered out everyone except operators who had already paid the setup cost once before.

That observation is the entire thesis of Stripe. The brothers went through Y Combinator's Summer 2009 batch under the working name /dev/payments, dropped out of MIT and Harvard respectively, and in 2010 launched the product publicly as Stripe. The pricing was 2.9% plus 30 cents per transaction, posted on the homepage. There was no application. You signed up, got test keys, and started integrating.

And then there were the seven lines.

The original Stripe documentation opened with a code snippet so short it felt like a provocation. Drop a script tag on your page. Call Stripe.createToken with a card number. POST the token to your server. Charge it with one API call. Done. A thing that was supposed to take days or weeks now took an afternoon.

The internal insight was not technical. The APIs under Stripe's hood talked to the same card networks and acquiring banks everyone else talked to. What Stripe had figured out was that the complexity was discretionary. The incumbent gateways and processors had assumed, essentially forever, that it was the merchant's job to understand KYC, PCI compliance, 3DS, chargebacks, tokenization and settlement timing. The Collisons flipped it. All of that was now Stripe's problem. The merchant's problem was shipping their product.

The slow, boring moat

Stripe is often described as a developer-tools company, which is both true and slightly misleading. The reason it worked is that Collison was willing to spend the next decade doing the deeply unglamorous work of hiding complexity that other people had been content to pass through.

That meant hand-negotiating interchange agreements. It meant building out risk and fraud models good enough to underwrite merchants that banks would have ignored. It meant adding payment method after payment method — ACH, SEPA, iDEAL, Alipay, Klarna — each of which required its own integration into card networks or local rails. It meant, eventually, Stripe Atlas, Stripe Issuing, Stripe Capital, Stripe Tax, Stripe Billing. Each new product was a new bucket of mess that Collison had decided Stripe should carry so the user did not have to.

The moat, in other words, is not the seven lines. The moat is the ten thousand person-years of back-office work that make the seven lines possible.

The numbers

Stripe is now one of the most valuable private companies in the world. An employee tender in February 2024 priced it at $65 billion. A follow-on tender in February 2025 marked it at $91.5 billion. Revenue hit $6.8 billion in 2025, up 33% from the year before. The company processes well over a trillion dollars of payment volume a year — a figure Stripe pegged at $1.4 trillion for 2024.

Collison has become, somewhat to his own apparent discomfort, a billionaire several times over. He has also become one of the louder voices in Silicon Valley arguing against going public. His view, articulated plainly in a February 2025 Fortune interview, is that payments growth is 'lumpy' — tied to the health of merchants, consumer spending and cross-border flows — and badly suited to the quarterly reporting rhythm of public markets. Stripe stays private. Employees get liquidity through tenders instead.

What the story is actually about

It is tempting to read Collison as a prodigy story. He is, by any ordinary measure, one — the Young Scientist trophy, the SAT at thirteen, MIT at seventeen, exited company at nineteen. He has even warned Gen Z not to copy the dropout move, pointing out that his own instinct that formal education was a waste was, in his words, a 'poor intuition.'

But the prodigy framing misses the actual lesson. The useful thing to notice about Stripe is that it did not require a prodigy. It required someone willing to look at an industry where every incumbent had agreed the pain was unavoidable, and quietly disagree. The seven lines of code were not a technical trick. They were a product decision — a bet that if someone was willing to absorb the mess, developers would gratefully hand over their payment volume forever.

That template generalises. Almost every category has a Stripe-shaped opportunity sitting in plain view: a legacy workflow everyone has learned to tolerate, surrounded by vendors who have each decided the tolerance is a feature. If you can identify that workflow, and you are willing to do the long, dull work of making it invisible, you have the shape of a company.

Collison's version happened to be payments. Yours is probably something else. But the move is the same: look at the thing everyone agrees must be hard, and ask whether 'must' is doing more work than it should.