Stripe Sold to the One Person Who Couldn't Sign the Check

In 2011, if you wanted to take a credit card on the internet, the hard part was never the code. It was the paperwork: a merchant account application, a gateway contract, a bank that treated you like a suspect, and weeks of waiting before a single dollar moved. Two brothers looked at that and built a prototype they called '/dev/payments'⁵ — and then did the one thing the entire industry had never bothered to do. They aimed it at the engineer instead of the executive.

The official story is that Stripe won because it was easy to use. That's true, and it misses the point entirely. Stripe built a better product and customers came. Stripe built a better product for a customer nobody else was selling to — and that choice of target, not the product itself, is the whole rebellion.

The buyer signed the check. The developer made the decision.

Every payments incumbent before Stripe ran the same playbook: find the business owner or the CFO, put a salesperson in front of them, negotiate rates, sign a contract, and hand the integration off to some poor engineer afterward as an implementation detail. The engineer was the last person consulted and the first person blamed. Stripe flipped the order. It treated the developer — the one writing the checkout code — as the actual decision-maker, because in practice, she always had been. A founder asks her one engineer 'how do we take payments?' and whatever that engineer can get working by Friday is the answer. Stripe simply built the thing that engineer could get working by Friday.

This is the inversion: in a developer-first model, the person who can't sign the purchase order is the person who chooses the vendor, and the person who signs the order merely ratifies a decision that's already been made in code. Legacy sales motions spent their budget on the wrong human. Stripe spent its budget on documentation.

The myth of seven lines — and why it doesn't matter that it's a myth

Stripe's legend hardened around a number: seven lines of code, and you're taking payments. Strip the legend back and the number won't hold. No Stripe filing or official release ever fixed it at seven; the precise line count varied by language and version, and no authoritative tally has been independently verified. The precise figure is marketing mythology laundered into engineering fact. But notice what the myth is reaching for, because that part is true: Stripe took a process that used to require a bank, a lawyer, and a fortnight, and compressed it into something a developer could paste in before lunch. The exact line count is decoration. The collapse of effort is the asset.

When adoption costs almost nothing, the developer doesn't ask permission — she just tries it. And trying it, once, on a real transaction, is how Stripe's distribution actually propagated. The very first transaction ran through a fellow YC company, 280 North, whose founder later joined Stripe as an early employee.¹⁰ That's the pattern in miniature: a builder adopts it, it works, and the builder becomes the channel.

Paul Graham coined a phrase for the brothers' earliest move: the 'Collison installation.' When someone agreed to try Stripe, they didn't send a follow-up link and hope. They said 'give me your laptop' and set it up on the spot.⁶ It looks like a quaint do-things-that-don't-scale anecdote. It's actually the developer-first thesis wearing work clothes — the recognition that the gap between 'I'll try it later' and a working integration is exactly where most payment adoptions died, so you close that gap by hand, one laptop at a time, until the product itself can close it for you.

Run the numbers forward and the inversion compounds into scale that's hard to fathom. Stripe reports it moved $1.4 trillion in 2024, up 38% year over year — about 1.3% of global GDP — and that it was profitable.¹ Net revenue came in around $5.1 billion.² The valuation arc reads like a typo: roughly $20 million at the 2011 seed, $95 billion at the March 2021 Series H, and a February 2026 tender offer pegging the company at $159 billion.⁸⁴ One caution worth stating plainly, since the figures get thrown around as if they were public-market prices: Stripe is private, files no 10-K, and every post-2011 valuation is a private tender or secondary number, not a market quote. The revenue and volume figures come from Stripe's own annual letter — self-reported, not independently audited.¹

Isn't this just 'build a good product'?

The fair objection is that 'developer-first distribution' is a grand label for an unglamorous truth: Stripe simply built a better product, and good products spread. There's something to that — the product genuinely was better, and a clumsy API aimed at developers would have gone nowhere. But the objection misses what was actually scarce. Plenty of companies could build a clean API. What almost nobody did was treat the developer as the buyer rather than the implementer — and that single reframing changed every downstream decision: who the docs were written for, who the support was aimed at, what 'easy' even meant. The product quality and the distribution choice aren't separable. Stripe built quality at the exact point of friction the rest of the industry treated as someone else's problem.

The honest counter is that timing and access helped enormously. The Collisons came up through Y Combinator, which gave them a built-in first market of fellow startups — 280 North among them — primed to adopt a developer tool from people they knew.⁵ A $2 million round from Thiel, Musk, Sequoia, and Andreessen Horowitz before a public launch is not an advantage available to everyone.³ So the position wasn't pure insight; it sat on a network and a moment. But the network only paid off because Stripe pointed it at the right person. A warm intro to a CFO would have produced a contract. A warm intro to a developer produced a default.

Stripe didn't out-sell the payments industry. It refused to sell the way the industry sold. The incumbents spent a decade perfecting the pitch to the person who could sign the contract, while Stripe quietly handed an API key to the person who actually wrote the checkout — and let a trillion dollars a year follow the path of least resistance. The genius was never seven lines of code. It was deciding, against every convention in the category, that the most powerful person in the room was the one who'd never been asked.