Part IThe Story
An Airport Without a Title
In April 2025, John Collison gave a journalist from The Currency a morning tour of Weston Airport, a small private airfield in Kildare that he and a consortium of investors had bought in 2021 from the Galway builder Brian Conneely. He walked the place the way a man walks property he expects to keep. Over coffee, the journalist asked a simple question. What, exactly, was his role here?
"Do I have a title?" Collison said. "Owner? Maybe I need to get some Weston business cards."
It is a throwaway line, the kind of self-deprecation that Irish people deploy almost by reflex when asked to account for themselves. It also contains most of what is interesting about him. Collison was then thirty-four, the president and co-founder of a company that processed $1.4 trillion in payments the previous year and was valued at $91.5 billion in a tender offer that February. Stripe is routinely described as the most valuable private company in the world. He has run it alongside his older brother Patrick for fifteen years without a public listing, without a dual-class drama, without the ritual autobiography. When he started at Stripe he did not have a desk. For years he sat beside the reception desk in the company's Mission District office so he could watch visitors come and go. And here he was on an airfield in the Irish midlands, shrugging off the idea that he needed a job title for something he had paid to own.
Later that afternoon, at an event for the Kildare Chamber of Commerce at the K Club, he took questions from the floor without seeing them in advance. He answered at length. Someone asked the question he hears every year, the one Fortune's newsletter editors called "our collective obsession": when will Stripe go public? He did not give the usual deflection. He gave a worldview.
— John Collison, Kildare Chamber of Commerce, 2025 (via The Currency)Why would you want to float it? It's like – 'oh, so you can sell your shares, obviously'. But it's like, no, we're working on this for decades to come, it is our life's work, and we really enjoy it. I feel like there's an assumed worldview behind the question of flipping something and moving on to the next thing, which is absolutely not how we view the world.
The journalist used a word for it that he admitted was unfashionable in modern capitalism: steward. Of Stripe, certainly. Of Weston. And, by the time this profile was written, of a 1,120-acre eighteenth-century estate in County Laois, a Victorian house that had been a ruin, a computer-science programme at the University of Limerick, a think tank, and a running public argument with the Irish state about why it can no longer build anything.
The question underneath John Collison is not how a teenager from a lakeside village with no internet connection came to co-own the plumbing of the online economy. That story has been told, in Forbes and Bloomberg and on 60 Minutes, until it went smooth. The more interesting question is the one the airport poses. What does a person do with a fortune that arrived before he was old enough to rent a car, in a culture that assumes the point of building something is to sell it? Collison's answer, assembled over two decades and never quite stated, is that you don't sell it. You keep it, you maintain it, you keep the runway open. You remain, in some essential way, the owner without the business cards.
By the Numbers
The Steward's Ledger
$1.9TTotal payment volume processed by Stripe in 2025, up 34% year over year
$159BStripe valuation in the February 2026 employee tender offer
50Users Stripe had roughly two years after the first line of code
$5MSale price of Auctomatic in March 2008, when John was 17
8 A1sJohn's Leaving Certificate result in 2009, one of eleven students nationally
5 millionBusinesses whose transactions Stripe processed in 2025
~25%Share of new Delaware corporations started through Stripe Atlas as of mid-2026
Satellite Weather
Dromineer sits on the eastern shore of Lough Derg, in north Tipperary, a village of a harbour, a ruined castle, and a sailing club. It is pretty in the way Irish lakeside places are pretty, which is to say it is pretty when it isn't raining. When the Collison boys were growing up there in the 1990s, it was too remote to get an ordinary internet connection. The only way to get decent broadband was an expensive satellite link. The future arrived at the Collison house from the sky, at a premium, and probably dropped out when the weather turned.
That detail is worth holding on to. It is the earliest instance of a pattern that runs through the rest of the story: infrastructure that should be simple is absurdly hard, and the Collisons pay whatever it costs to route around it.
The parents were scientists. Lily Collison is a microbiologist; Denis Collison an electronic engineer. They were, in their son's understated phrase, "pretty open minded." There were three boys. Patrick, born on 9 September 1988, was the eldest, the one who took his first computer course at the University of Limerick at eight and began programming at ten. John was born in August 1990. Tommy, the youngest, entered the Young Scientist exhibition in 2010 with a project on blogging and became, by 2011, what one Irish tech journalist called "an ardent tech blogger and public speaker." The household seems to have run on a kind of ambient technical enthusiasm. When the boys started blogging and tweeting, their mother asked one day if John would set her up with a blog. He installed WordPress on her machine. She started writing that same day, and the parents joined Twitter.
Patrick was the prodigy who made the newspapers first. In January 2004, at fifteen, he entered the 40th Young Scientist and Technology Exhibition with an artificial intelligence project he'd nicknamed "Isaac," after Isaac Newton, and finished as individual runner-up. That same year, at the EU Young Scientist competition, he met a Swiss teenager named Silvana Konermann who beat him; eighteen years later he married her. He came back to the Irish contest in 2005 with Croma, a programming language in the Lisp family, and on 14 January 2005, aged sixteen, won. President Mary McAleese handed him a cheque and a trophy of Waterford Crystal.
It's tempting to cast John as the younger brother in the prodigy's shadow, and early press coverage did exactly that. The facts don't cooperate. By seven, according to a family member quoted by The Irish Times, John was attending courses in journalism and medicine. He debated competitively. He played piano. He went to a physics Olympiad. "Learning is food to John," the relative said. "He was never a swot, he just doesn't see any boundaries between school and life – it's all learning and it's all good."
That last sentence is the closest thing in the record to a description of his temperament, and it holds up across two decades of interviews. The boundary between work and curiosity, between the classroom and the startup, between Stripe and everything else, never seems to have been drawn.
In April 2021, Forbes published a profile of the brothers claiming they had "escaped" from Limerick, which it described as a "warzone," the "murder capital" of Europe, where "shootings, pipe bomb attacks, and stabbings" happened every night and some neighbourhoods were "walled off by a dirty graffitied 10-foot-high barrier, like the Berlin Wall." The reaction in Ireland was swift and contemptuous. Two Limerick TDs demanded an apology. Patrick O'Donovan offered to host the writer personally: "Please let me know when suits to visit." The article came down on 9 April. John's verdict on Twitter was one word: "daft." Patrick wrote the sentence that matters.
— Patrick Collison, on Twitter, April 2021Not only mistaken about Limerick but the idea of 'overcoming' anything is crazy. We are who we are because we grew up where we did.
The American magazine wanted an escape narrative because escape narratives are how Silicon Valley explains its immigrants. The Collisons rejected it, and not out of politeness. Their whole account of themselves runs the other way. They did not overcome Ireland. Ireland, with its satellite dish and its exam system and its parents who let a sixteen-year-old fly to San Francisco, made them. What Ireland would not do, in 2007, was give them any money.
The Shop Nobody at Home Would Fund
The first company was called Shuppa, a phonetic rendering of siopa, the Irish word for shop. The brothers set it up in Limerick in 2007, when Patrick was eighteen and John was still in Transition Year, the optional fourth year of Irish secondary school designed for work experience and self-discovery. Most Transition Year students do a placement at a local solicitor's or bake a cake for a mini-company competition. John co-founded a software firm.
They went looking for money. Enterprise Ireland, the state body charged with funding exactly this kind of thing, declined. No private Irish investor stepped in either. The Guardian, reporting the outcome a year later, was explicit: the brothers "were unable to find funding from either Enterprise Ireland, the state investment body, or any private Irish investors."
Y Combinator, the Mountain View accelerator, said yes. It also put the Collisons in touch with two other founders working on the same problem. Harjeet and Kulveer Taggar were British brothers, Oxford graduates, building tools for people who sold in bulk on eBay. The two pairs merged, and Shuppa became Auctomatic. The Taggars are worth a sentence each because their later careers rhyme with the Collisons'. Harj went on to become Y Combinator's first non-founding partner in 2010, then co-founded Triplebyte. Kulveer built Zeus, a corporate housing company, to more than $450 million in revenue before selling it in late 2023. Two sets of immigrant brothers, both from the edge of Europe, both of whom found that the place that would fund their teenage ambitions was six thousand miles from home.
John went with them. "Our parents are pretty open minded," he told Silicon Republic in 2011, "most evidenced by – but still can't figure out how it happened – me going off to San Francisco when I was 16." Patrick deferred his place at MIT. Auctomatic raised more than $400,000, with checks from Paul Buchheit, the engineer who built Gmail, and Chris Sacca. Its product was unglamorous and, according to an early reviewer who had been an eBay PowerSeller himself, essential: listing templates, scheduling, bulk listing, image hosting, inventory management.
Years later, onstage at Stripe Sessions in 2024, Patrick described what was wrong with it. Auctomatic was over-engineered in a way only very clever teenagers can manage. "Everything was so elegantly and carefully built. We wrote all the software in Smalltalk." Fidji Simo, the Instacart CEO moderating, asked what that meant. "It's a language," Patrick said. John cut in: "It's an obscure programming language, even to those who appreciate obscure programming languages." They had built their own distributed object data store rather than use "any silly mainstream, relational database." They had a system for saving a virtual machine image so that a crashed session could be restored to its exact state.
It is a funny exchange and an instructive one. Patrick tells the engineering story. John supplies the punchline that makes it legible to an audience. That division of labour shows up again and again.
The other lesson of Auctomatic was the one that mattered. In building it, Patrick said, "we started to get exposed directly ourselves and also for the first-time hearing stories from others about the difficulties with accepting online payments. So it's where we discovered the actual Stripe use case."
On 26 March 2008, a Vancouver domain-name company called Communicate.com announced that it was buying Auctomatic for $5 million in cash and stock and renaming itself Live Current Media. Its chief executive, Geoffrey Hampson, owned roughly 800 domains (perfume.com, cricket.com, brazil.com) and had declined an offer of $6 million for cricket.com earlier that year. Hampson's plan was to build branded shopping properties on the domains, with Auctomatic's engine underneath. Michael Arrington, writing in TechCrunch, was unconvinced: "Will this work? I don't know and I'm not even sure I'll remember to check in on them." He hoped the founders, who were contractually obliged to move to Vancouver, would "be able to pry themselves loose at some point in the future and start another company."
The deal closed on Good Friday. Two large American internet companies had reportedly bid against Live Current, which drove the price up. Patrick, nineteen, moved to Vancouver as director of engineering. John, seventeen, went home to Limerick to sit his exams. Irish television put them on a rich list at Christmas. They went on The Late Late Show. Their family later said the attention left John "a little media-shy."
The coverage got the most important fact backwards. Ireland read the sale as an arrival. John apparently read it as a liability. A source close to the family told The Irish Times in 2009: "John does not want to be defined by Auctomatic. He would consider the whole venture a failure if he was stuck with that tag for the rest of his days."
A $5 million exit at seventeen, considered by its beneficiary a potential failure if it turned out to be the high point. You can call that arrogance. It reads more like a scale problem. The boy had seen how big the thing could be and knew this wasn't it.
Ten As and No Airstrip
In August 2009 the Leaving Certificate results came out, and the teenage millionaire made the news again. John Collison of Monaleen, Limerick, had scored eight A1s and two A2s, ten subjects at the top grade bands, including maths and physics. He was one of eleven students in the country to get eight A1s. Martin Wallace, the principal of Castletroy College, gave the papers a quote that has aged well: "Very often when you imagine the type of person who gets results like this you think of somebody with their head stuck in a book all the time but John is a very sociable guy." Wallace also called him "a very level headed student who has a wide range of interests and in particular enjoys debating."
The same report noted, almost as an afterthought, that he was a student pilot with more than forty hours of flying time.
This is where the airport in the opening scene begins. Collison learned to fly at seventeen, the same year he sold his first company. It is a hobby that demands, more than almost any other, the habits a payments company later required of him: checklists, redundancy, paranoia about failure modes, an understanding that the boring infrastructure of a runway and a control tower is what makes the romantic part possible. He has said nothing in public, so far as the record shows, about why he took it up. It isn't hard to guess what appeals to an analytical mind about a discipline where carelessness is fatal and process is a form of respect.
He went to Harvard in September 2009 and had to ground the flying, because there was no airstrip. He joined the rowing team within a week. He was considering physics and statistics with a minor in Chinese. A friend who had travelled around Europe with him reported that he ran the trip "with military precision, and is not the type to wake up on a beach in Naxos." Before term began, Harvard ran a dorm clean-up where students could work on campus for minimum wage. John was "very anxious" to be there, according to a source, "because he felt he would meet people who were the best craic." The same close friend added the line that best predicts the next fifteen years: "John doesn't want what he describes as a selfish career."
His first-year course list, as he recited it to a student journalist from University College Dublin in 2012, was maths, physics, Chinese, writing, statistics, and astrophysics. "It's completely á la carte," he said, "which I really liked."
It was too à la carte to survive contact with what came next. In October 2009, a month into his first semester, he and Patrick (now at MIT, a few miles down the river in Cambridge) wrote the first lines of code for a payments product. John later described the year as a gradual inversion.
— John Collison, The Irish Times, 2015It started off with Stripe being this fun little side thing when we weren't working in class and probably by the end of the year class was this fun little side thing when we weren't working on Stripe.
"At first we were just messing around a bit," he told the UCD interviewer, "and then it got more serious and we got a few customers. We decided to move out to Palo Alto for the summer to work on it full-time. We had no plans as to what we'd do at the end of the summer – we hadn't thought it through, but by this time the team was bigger and we had raised some money, so we decided that we'd take some time off just to see where it went and we're still just taking time off to see where it goes."
He was twenty-one when he said that. As far as anyone knows, he is still on leave.
Two Years, Fifty Users
The problem the Collisons set out to solve in 2009 was simple to describe and miserable to fix. Taking money online meant obtaining a merchant account through a bank, negotiating with card networks, filling in long contracts, waiting days for approval, paying high fees, and integrating with gateway software that looked like it had been designed to discourage use. The Dutch indie hacker Pieter Levels, interviewed by John on his podcast in 2025, described signing up for Worldpay as a twelve-year-old in the Netherlands. His father had to sign a contract making him liable for damages of up to $100 million. "What am I signing for?" his father asked. They faxed it to America anyway. He never sold anything, "luckily, maybe."
John's diagnosis, given to The Irish Times in 2015, was that the brothers were the target customer. "We had been building online businesses and we had run into the fact that the hardest part of running an online business was not creating a product that had lots of demand. It was actually capturing that demand and building out the business side of it and being able to accept money for it."
What they built, after a fashion, was a handful of lines of code a developer could paste into a website to accept a card payment. Bloomberg Businessweek later put "seven lines of code" in a headline, and the number became lore. Pricing was simple: 2.9 percent plus thirty cents per transaction, no setup fee, no contract to fax.
It took much longer than they'd expected. "I remember right at the beginning when we were starting it I said to Patrick, 'Yeah let's do it. How hard can it be?'" John told a Stanford audience in February 2015. "Which gives you a sense of our mindset. And the answer was: two years of difficulty. We had not predicted that."
The first customer came three months after the first line of code. Then a long grind. Payments is a regulated industry layered over a set of decades-old bank networks, and two young Irish brothers with a beautiful API had to negotiate their way into it. "There was a long period where we were writing code to support a small handful of users," John told the Limerick Post, "and navigating a relatively unfamiliar industry." In May 2011, TechCrunch reported a $2 million investment from Peter Thiel, Elon Musk, Sequoia Capital, Andreessen Horowitz, and SV Angel, with a smaller check from the Cork-born manufacturing entrepreneur Liam Casey. It was a striking roster for an unlaunched product, and the irony wasn't lost on anyone: PayPal's founders were backing a company built to make PayPal look clumsy.
Stripe launched publicly at the end of September 2011, almost two years after the first commit, with ten people on the team. By John's count it had about fifty users.
"When you spend two years getting 50 users," he said at Stanford, "it doesn't feel like a whole lot of progress." Then he added the observation that a whole generation of founders has since written on whiteboards: "If you're working on a startup that's a bad idea, it's going to feel like slow-going. But if you're working on a startup that's a good idea, it may feel like slow-going too."
What the brothers did with those fifty users became the company's genetic code. Whenever a user hit a bug, one of them got an email or a call. Minutes later the user would get a note saying it was fixed, and people, as John put it, had their minds blown. There was a chat room on Campfire that any customer could join to message the founders at any hour. Feedback submitted through the dashboard got a reply within ten minutes. If a user was in the Bay Area, the brothers would invite them to the office and do the integration for them.
"Since we were spending so much time on each one of those users; since we were hyper-focused on building a great product; and since we weren't dealing with problems of scale yet, that allowed us to build the product that we wanted," John said. "Part of the culture that set in really early on was taking abnormally good care of those early users."
The result was growth through word of mouth, he said, "even to this day."
There is a phrase John uses for this that makes Stripe's product managers wince a little, because it is a reproach to most of their industry. He calls it talking up to the user. In a 2024 podcast with his brother he described what he considers the default posture of consumer software: "You have users by the wheelbarrow. You treat them in aggregate. You talk about MAUs. Everything is statistical and everything – like LinkedIn with the dark patterns and complete your profile – is nudging you towards some business outcome." Stripe's users, he insisted, were "people starting a business who are generally really talented. And you can just see it if you go talk to any of them." Every week, one customer is invited to the company's internal fireside chat to explain how they use the product and what they'd like done better.
The fifty-user years also set the hiring pace. "We hired extremely slowly at the beginning," John has said. "It took us a year to get to four people." He told Wired in 2017 that the habit persisted long after it was financially necessary: "If you're building a team to solve a problem that might take a decade, it would be really bad short-term optimisation to hire the person who doesn't seem quite right. It means you can't scale as quickly, but we're perfectly fine with that."
The word decade turns up in that sentence when Collison was twenty-six. It turns up again, as decades, at the K Club at thirty-four. The horizon never shortened.
The Younger Brother's Job
In 2015 a reporter from The Irish Times visited Stripe's office in a refurbished luggage factory in the Mission District and found bicycles leaning against each other in reception, staff on sofas with laptops, and a crowd squeezed into a glass conference room called Manatee. The ground-floor rooms were named after aquatic mammals. Across 18th Street, in a former laundry, the second-floor rooms were named after animals that climb trees. John arrived in Jaguar ahead of his brother, who was running late.
Until three months before that visit, John had had no desk of his own. He had been sitting next to the receptionist "where he could watch visitors come and go." Patrick's workspace, according to a 2016 profile, was "a little work area packed between a wall and the desk used by the admin he shares with his sibling."
The brothers share a lot. They do not share a temperament, or at least not a public one. Patrick is the CEO, the reader with the famous online bookshelf, the co-author with Tyler Cowen of the 2019 Atlantic essay calling for a discipline of "Progress Studies," the co-founder (with Konermann and Patrick Hsu) of the Arc Institute, a nonprofit biomedical research lab. His intellectual range is real and occasionally a little forbidding. John is the president, and by most accounts the one who ran partnerships and sales in the early years while Patrick handled product design and served as the company's public face. He is also, unmistakably, the funnier one.
The comedy is a recurring feature of the company's public life and it's worth taking seriously. At Stripe Sessions in 2023, before John's fireside chat with Sam Altman, a video bit had Patrick appear to hijack his brother's slot. "Hmm, I think John has a bug," Patrick said. "Get out of here, Patrick," John replied. "You had your own talk this morning. Sorry, Sam. Patrick's always deepfaking his way into my talks." At Sessions 2025, the opening sketch had Patrick hunting for his brother backstage and finding him "vibe coding." "Patrick, you're so TradFi," John said. "Accept all and deploy. What are you waiting around for?"
These are corporate skits and they're a little corny, and that's the point. The brothers have chosen to present the most powerful sibling partnership in technology as a running joke between two people who have known each other all their lives. When a Fortune newsletter writer floated the idea of a cage match between the Collisons and the Winklevoss twins in 2024, John played along. The levity is a way of signalling that the arrangement is stable, which, in a company where the two founders together hold a controlling stake, is not a trivial thing for employees and investors to believe.
John has been explicit about why it works. Of starting a company with a co-founder you've known forever, he said: "All the meta issues of working style and how we work together were out of the way and we could focus squarely on the task at hand." He has also been explicit about the costs. Starting a company with someone, he said, "is often seven-day weeks and 14-hour days again and again and again and going through a lot of uncertainty, a lot of setbacks."
There is a story John tells about his brother that captures the family dynamic better than any analysis. Talking to Pieter Levels about entrepreneurs who conscript their parents into schemes they shouldn't, he recalled that Patrick, as a boy, once filled out a form requesting a demo unit of a Sun Microsystems workstation, the high-end machines of the era. "And he just filled out the form and got them, and they came with a truck and they unloaded it." The older brother orders the workstation. The younger brother tells the story, and gets the laugh.
Mike Moritz of Sequoia Capital, who sat on Stripe's board and had backed PayPal, Yahoo, and Google, called the Collisons some of the sharpest people he had ever funded. Patrick O'Shaughnessy, introducing them on his Invest Like the Best podcast after they shared the Singleton Prize with John Malone, called them "two of the smarter people that I've ever encountered" and admitted he'd gone in with "tons of topics to cover, only to get through close to zero of them." The prize itself is named for Henry Singleton, the founder of Teledyne, the conglomerate builder whose capital allocation became a cult among investors. It was an odd honour for two software founders, and a telling one. The foundation saw in them not a pair of product geniuses but a pair of allocators.
Seven Lines and a Trillion
In January 2014, the Limerick Post ran a headline that would have read as satire if it hadn't been sincere: "Limerick brothers have no plans to cash in on billion euro firm." Stripe had just raised fresh money at a €1.3 billion valuation. It had 83 employees in San Francisco. Patrick told the paper: "We've no interest in selling the company. John and I are very lucky to work with the people we do on building something for the long term, and we wouldn't like to change that." John gave the line that, read now, sounds like a plan for the next twelve years: "It's still early days for the payment systems that power the internet. It's outrageously difficult and expensive to move money between countries. Fraud and storing credentials still cause merchants hassle. So online payments is a great place to go hunting for problems worth solving."
They went hunting. In 2012, Stripe Connect let software platforms embed payments inside their own products, which is how Lyft, Instacart, Shopify, and Kickstarter came to run on Stripe. In February 2016 came Atlas, a service that let a founder anywhere in the world incorporate a Delaware company, open a US bank account, and start accepting payments. It was invite-only at first; in March 2016 it added Cuba. Atlas is the most personal product Stripe has built. It solves, for strangers, exactly the problem that sent two Irish teenagers to California: the place you live won't give you the infrastructure to start. By mid-2026, by Patrick's count, around a quarter of new Delaware corporations were being started through it.
In April 2016, Inc. put the brothers on its cover as "The Money Men." The Inc. April 2016 Patrick & John Collison The Money Men issue caught them at an inflection: Stripe was valued at $5 billion, still largely a startup's tool, and about to discover that the giants wanted it too. Later that year a Stripe estimate held that 27 percent of Americans would buy something through Stripe in 2016. In November, an investment from CapitalG and General Catalyst valued the company at $9.2 billion and made the brothers, at 26 and 28, the world's youngest self-made billionaires. John told Wired he had not seen it coming in the way the press assumed: "What we didn't realise was that the problem was actually much broader. One, the problem was applicable all around the world. Two, it was relevant not just to new companies starting out, but also to the Salesforces, the Walmarts and the Twitters."
Radar for fraud and Billing for subscriptions came in 2018, Issuing for cards, Terminal for physical readers in 2019, Capital for merchant advances, Treasury for embedded bank accounts in 2020. Paystack in Nigeria was bought in October 2020 for more than $200 million. The valuation went to $20 billion in 2018, $35 billion in 2019, and $95 billion in March 2021, the last raise including Ireland's National Treasury Management Agency, which made the Irish state, fourteen years after Enterprise Ireland passed on Shuppa, a shareholder in the brothers' second company.
Then it went down. In July 2022 the Wall Street Journal reported that Stripe's internal share price had dropped, cutting its implied valuation from $95 billion to $74 billion. In November the company laid off 14 percent of its staff. In March 2023 it raised more than $6.5 billion in a Series I at a $50 billion valuation, not because it needed operating capital (the company said it didn't) but to give liquidity to current and former employees and cover the tax bills on their equity. Nearly half the paper value had evaporated in two years. In June 2023 the Sunday Times ran an interview with John under the headline "I've just lost $4bn — and I'm more ambitious than ever."
The headline is the right summary of what happened next. Stripe processed $1 trillion in 2023 and announced it was cash-flow positive. "We've been driving a lot of [efficiency] over the course of the last few years," John told Fortune in March 2024, "which is how we've ended up cash flow positive." He did the arithmetic for the reporter with the cheerfulness of someone reframing a cliff as a foothill: "We're a trillion dollars in payment volume, we're very proud to have passed that milestone. But, depending on how you define digital payments, it's $50 trillion-plus in total payment volume. So, we're less than 2% of the overall digital payments market."
The valuation rose to $65 billion in February 2024, $91.5 billion in February 2025, and $159 billion in February 2026. Payment volume reached $1.4 trillion in 2024 and $1.9 trillion in 2025. Stripe's Revenue and Finance Automation suite passed $500 million in annual run-rate revenue in January 2025, driven largely by AI companies like OpenAI, Anthropic, and Perplexity that needed to price usage in ways older software couldn't. In February 2025 Stripe closed its $1.1 billion acquisition of the stablecoin platform Bridge. In September it launched Instant Checkout inside ChatGPT and co-published the Agentic Commerce Protocol with OpenAI. On 15 July 2026, Stripe and the private equity firm Advent International announced a $53 billion joint bid for PayPal, the company whose founders had written Stripe's first checks. In August, Stripe agreed to buy the AI routing startup OpenRouter for $7.5 billion.
John's account of all this deal-making is almost comically unstrategic. "We've always got acquisitions on the brain," he told Fortune. "We're always looking for companies, but it's also very hard to set a goal of 'I'm going to acquire one company a quarter, roughly of this size.' So, it's opportunistic."
Not every bet worked. Stripe led a $20 million Series A in a one-click checkout startup called Fast in March 2020 and a $102 million Series B in January 2021. Fast shut down in April 2022. In January 2025, when Stripe laid off about 300 people, the company sent departing employees a cartoon of a duck, and its chief people officer had to apologise. The record of a steward includes the things that broke on his watch.
Through all of it, the IPO question. "People love speculating on this," John told Fortune. "We have no news to share, and when it's time to share, we'll share it." To Silicon Republic, in 2022: "We're very happy as a private company. We'll let you know if that changes." The brothers made the same point in 2014 at €1.3 billion that they made in 2025 at $91.5 billion. The scale changed by a factor of about seventy. The answer did not change at all.
A Library Inside a Payments Company
In February 2025, a Fortune reporter came to Stripe's San Francisco offices to talk about the billing business and ended up spending, in her words, "far too long looking at books." There is a nook in the office called The Library. Dickens, Tom Wolfe, Charlie Munger. Collison agreed with her that Wolfe's best work transcends The Electric Kool-Aid Acid Test. He sat down with an 82-year-old Dickens compilation from 1943. She picked up a Stripe Press edition of M. Mitchell Waldrop's The Dream Machine, the history of the computing visionary J.C.R. Licklider.
Then she told him she had a confession: a friend of hers used Stripe Billing and had sent along a review. Collison stood up, she wrote, "both (I think) as a small act of physical comedy and mild genuine nerves," and began returning the books to the shelf. "Just putting everything back on the shelf!" He exhaled when she told him her friend loved the product.
A man with a fortune measured in billions, visibly nervous about one customer's opinion of an invoicing tool. It's the fifty-user years, still running underneath the trillion-dollar company.
Stripe Press, launched in 2018, is the company's publishing house, and the most baffling line item in fintech. It publishes, in John's own description of one of its titles, "interesting left-brained books in beautiful, right-brain-friendly editions." It reissued Richard Hamming's The Art of Doing Science and Engineering. In 2023 it republished Poor Charlie's Almanack, with a foreword by John. "His essays extol the virtues of free enterprise, yes," he wrote of Munger, "but also of doing business the right way, with integrity and rigor. Of taking your work very seriously, but never yourself." Munger died that November. Stripe's annual letter for 2023 leaned on him heavily.
The reading is not ornamental. Collison's intellectual method, to the extent he has described one, is to study what others have done at the stage he is entering. "One thing we try to do is just spend a lot of time looking at all the other companies and what they've done," he said in 2024. "Not that you want to blindly emulate them, but you should at least understand them." He offered Apple and Amazon as the canonical pair: "You cannot imagine two companies that work more differently than the two of those. Yet they're both really successful models." He told a story about Magnus Carlsen winning a chess trivia contest: "I don't think it's a coincidence that the world's number one player has also studied the most about chess history." To Wired, the same principle in its bluntest form: "People tend to pay too little attention to history – the history of Silicon Valley and American business – and think they're the first people to come across a problem. For Stripe, being inventive is just about applying the right solutions from other areas."
In 2020, on Invest Like the Best, an episode that became for a long time the show's most downloaded, Patrick O'Shaughnessy opened by asking him about industrial conglomerates. That is where Collison wanted to start. The conversation ranged over GAAP accounting, clear internal writing, the tradeoffs between size and speed, and the claim that board members are like Pokémon characters, each with a different power. A software founder talking about the corporate structures of mid-century American manufacturing is not making small talk. Stripe today, with payments, billing, tax, identity, cards, lending, a stablecoin network, a publishing house, and a carbon-removal coalition (Frontier, a $925 million advance market commitment started in 2022 with Alphabet, Shopify, Meta, and McKinsey), looks more like a conglomerate than any company its founders were compared to in 2011.
In 2025 he started a podcast. It is called A Cheeky Pint, and it is recorded in a pub built inside Stripe's offices, where Collison pours a beer for the guest and they talk. Ben Thompson, the Stratechery writer, came on in January 2026 and spent the opening minutes discussing why Taiwan is the most convenient place to live and how 7-Eleven food there descends from Japanese colonial urban planning. The guests range from Waymo's Dmitri Dolgov to Pieter Levels. The format is almost aggressively Irish: a pint, a chat, a conversation that drifts. That the pub is inside the headquarters of a $159 billion financial infrastructure company is the joke, and also not a joke at all.
Paranoid About Complacency
The money Ireland wouldn't give him in 2007, Collison has been giving back to Ireland with some insistence ever since, along with a good deal of unsolicited advice.
In January 2022 he went to the University of Limerick to hand out prizes for TECS, a new competition for secondary students run through UL's Immersive Software Engineering programme, which Stripe backs alongside Intercom, Zalando, and other firms. The top prize of €2,000 went to Abutalha Alam, a sixth-year at Castletroy College, John's old school, for a sign-language translator. Castletroy got another €1,000. "We want to continue to grow the tech ecosystem selfishly," Collison told RTÉ. "It's what we want at Stripe, but also it's what everyone needs." He said the programme could become "the best computer science programme here in Europe, bar none." He also noted, with a bureaucratic dryness that would sharpen considerably over the next three years, that hiring international workers in Ireland was "a bit slow at present."
In an interview with The Currency, he laid out the problem as he saw it. He opened, without being asked, with Intel's decision to build a plant in Magdeburg rather than expand in Ireland, where it already had a large operation. "I feel like if you are Intel, there is a lot for doubling down on a country where they have lots of operations already, which has been very good for them, that is a great country to do business in. Germany is a new thing for them. It is not generally regarded as the world's easiest country in which to do business." What bothered him was not the decision itself but the reaction. "To my eye, it did not seem to lead to a huge amount of introspection and handwringing about 'Oh crap, what did we do wrong?' Instead, it was about the fact we don't like data centres, which, to my mind, is a fairly loopy way of looking at things because it is such a fixed mindset rather than an expansionary, ambitious, optimistic mindset."
Then the sentence that gives this section its title: "I guess the general worry I have is, is there some complacency that we need to be quite paranoid about?"
He reached for T.K. Whitaker, the civil servant whose economic programme in the late 1950s pried Ireland open to foreign investment and is credited, in Irish political memory, with ending decades of stagnation. "TK Whitaker's programmes were radical for the time. The EU has always been a set of quite radical ideas. We all now take it for granted, and there are nice blue signs on the roads. But it was a collection of pretty radical ideas for its time." The blue signs on the roads are the European Union's funding plaques on Irish infrastructure, so familiar now that nobody reads them. Collison's point is that the plaques are the residue of an ambition the country has since forgotten it had. "I feel like now there is some kind of fear of failure in the air where everything becomes a protracted scandal: the Children's Hospital becomes a scandal, the Dublin Airport cap becomes a scandal. Everything becomes a scandal."
In October 2025, the same month Stripe opened its new Dublin headquarters with the Taoiseach, Micheál Martin, at the ribbon, Collison published an essay in The Irish Times arguing that Ireland was going backwards. Too many regulators, departments, agencies, and NGOs had placed too many constraints on large projects; elected politicians had ceded too much power to unelected officials. The government responded within days. A spokesman for the Taoiseach called it "food for thought." The Tánaiste, Simon Harris, posted: "Too many layers, too many structures, processes that take too long. Time for the pendulum to swing back." A Fianna Fáil backbencher said he wanted to invite Collison to address the parliamentary party.
Not everyone was charmed. An Irish Times columnist wondered whether Collison was "a little naive about the price that democracies pay in terms of implementing reform" or whether he had "spent a little too long in Silicon Valley drinking the Kool-Aid," and pointed pointedly at the politics of his early investors. The column conceded that Progress Ireland, the think tank Collison supports and cited, "is not exactly the Heritage Foundation," and that his diagnosis drew on the left-of-centre writer Marc Dunkelman. But the suspicion it voiced (that the billionaire who left had come back to tell the country how to run itself) is the oldest suspicion in Irish public life, and it is not going away.
Collison has handled the political side of his visibility unevenly. In May 2025 he told The Irish Times that he was "baffled by companies doing an about-face on social initiatives," a sentence that put a little daylight between him and some of his Silicon Valley peers. His brother has taken positions that attracted more criticism. John's interventions have stayed mostly on a single theme, which he states the same way in every venue: build things, approve things, stop treating every large project as a scandal waiting to happen. It's the payments thesis translated into planning law. The problem in 2009 was that taking money online was absurdly hard. The problem in 2025, as he sees it, is that building anything in Ireland is absurdly hard. The fix in both cases is not ideology. It is plumbing.
Higher Ground
Abbeyleix is a town of fewer than two thousand people in County Laois, the next county over from where the Collisons grew up. In 2021, John bought Abbeyleix House and Farm, a residence dating to the 1700s set on 1,120 acres, for a reported €20 million. In 2022 he bought the neighbouring Millbrook House, a derelict Victorian building, and said he would spend several million euro over some years restoring it as a family home.
The estate's history is a parable he didn't have to invent. For centuries it belonged to the de Vesci family, whose mark on the town is everywhere. In the eighteenth century, with the old town prone to flooding from the River Nore, the de Vescis made a decision that would be unthinkable under any modern planning regime: they levelled it and rebuilt Abbeyleix on higher ground. A family that owned the land took responsibility for where the town should stand, and moved it.
Five years after Collison arrived, The Irish Times went to Abbeyleix to ask what a tech billionaire does to a small Irish town. Mark Zuckerberg had just bought a castle in Waterford, and people in Youghal were nervous. The answer in Laois was mostly good news. Robbie Quinn, chair of the local Tidy Towns committee, called Collison's arrival "great for the town." "John Collison may be looking to set down roots long term," Quinn said. "He's from the next county. I don't know what Mark Zuckerberg's intentions are long-term." About twenty-five people work on the house and farm, many of them local tradespeople doing the stonemasonry and carpentry that keep an eighteenth-century building standing. The estate works with the Tidy Towns group and with a community bog restoration project that started in 2000 to stop commercial peat harvesting and won a fifty-year lease in 2009 to manage the land for conservation.
Millbrook House did not become a family home, or not only one. It is now Ómós, a boutique restaurant and guest house that Collison part-funded, on the edge of the estate, employing thirty-eight people.
The name is Irish. It means respect, or homage. Nineteen years earlier, the first thing the Collison brothers built was given an Irish name too: Shuppa, from siopa, a shop. It was a business. It was meant to be sold, and it was, on a Good Friday, to a domain-name company in Vancouver, because nobody at home would fund it.
The second Irish name belongs to an old house on higher ground that someone has decided to keep. There is no business card on file for his role there either. There are stonemasons on the walls.
Part IIThe Playbook
What follows is an attempt to extract from John Collison's career the principles that actually explain it, as distinct from the ones that decorate it. Some are about product, some about capital, some about the odd discipline of running a company with a sibling for decades without selling it. All of them rest on evidence from Part I. A few cut against prevailing startup orthodoxy, which is part of why they are worth writing down.
The principles
- 1.Be your own first customer, then go hunting.
- 2.Treat slow traction as a building window, not a verdict.
- 3.Talk up to the user.
- 4.Hire at the speed of conviction, not the speed of capital.
- 5.Make culture operational, not atmospheric.
- 6.Refuse the assumed worldview of the exit.
- 7.Read the history before you claim the invention.
- 8.Let the incumbent's head start become its handicap.
- 9.Reframe the drawdown against the size of the market.
- 10.Divide the partnership by temperament, not just title.
- 11.Route around the infrastructure that failed you, then fix it for others.
- 12.Steward the things you can't monetize.
Principle 1
Be your own first customer, then go hunting
Stripe's founding insight was not the product of market research. It was the product of being annoyed. The brothers had built Auctomatic, encountered the merchant-account maze themselves, and heard the same complaints from other founders. "In a way we were exactly who you would expect to start Stripe," John said in 2015, "because we had been the target customers." Being the customer gave them an unusually precise specification: they knew which steps were insulting, which waits were pointless, which contract clauses existed only to protect the bank.
What separates Collison from the thousands of founders who scratch their own itch is what happened next. Having solved their own problem, the brothers treated payments as a hunting ground. "Online payments is a great place to go hunting for problems worth solving," John said in 2014, and then listed them: cross-border transfers, fraud, credential storage. Each became a product line: Radar, Link, Atlas, Treasury, Tax. The 2026 a16z conversation made the same point about newer bets: Tempo and Atlas both "grew from specific founder complaints, not market-size spreadsheets."
The initial personal pain gives you conviction and a specification. The habit of hunting turns a single tool into a platform. Most founders do the first and stop.
Tactic: Write down the three most irritating steps in your own workflow this month, then ask which of them a thousand strangers share.
Principle 2
Treat slow traction as a building window, not a verdict
Two years and fifty users. Collison's honesty about this number is itself instructive, because the startup world is structured to hide such numbers until they've become a funny anecdote. His reframing is precise: slowness is not diagnostic. "If you're working on a startup that's a bad idea, it's going to feel like slow-going. But if you're working on a startup that's a good idea, it may feel like slow-going too."
The more useful part of the argument is what slowness allowed. With fifty users and no scaling problems, the founders could reply to feedback in ten minutes, fix bugs and email the user before the user had finished complaining, and drive to a customer's office to do the integration for them. That intensity of care created the word-of-mouth engine Stripe still runs on. A company that had grown faster would have had to triage. Stripe could afford to be obsessive.
The discipline is to use a slow period deliberately rather than merely endure it. Collison's metric during those years was not user count. "The passion that the early users had about the product was that leading metric," he has said. Fifty fanatics are a better signal than five thousand indifferent signups.
Tactic: If growth is slow, measure the intensity of your best users' attachment and spend the spare capacity making their experience absurdly good.
Principle 3
Talk up to the user
The phrase sounds like a slogan until you notice what it opposes. Collison's critique of consumer product management is sharp: "users by the wheelbarrow," MAUs, dark patterns, nudges toward "some business outcome." The alternative assumes the user is smart, capable, and engaged in serious work, and designs accordingly. For a developer tool this means clean documentation and honest error messages. For a company it means inviting a customer, every week, to the internal fireside chat to explain what Stripe gets wrong.
The principle extends beyond product. In the Fortune library scene, a billionaire visibly braced for a single user's review of a billing product. The nerves are the principle made physical. Respect for the user means their verdict matters, individually, even at five million businesses.
There is a commercial logic too. Stripe's customers are founders, and founders are disproportionately the people who become CTOs of large companies, start new companies, and recommend tools to other founders. Talking up to them is condescension-proofing a referral network.
Tactic: Remove one "nudge" from your product this quarter and replace it with a direct, honest explanation the user can act on.
Principle 4
Hire at the speed of conviction, not the speed of capital
"It took us a year to get to four people." Stripe had famous investors and could have hired faster. It chose not to, and Collison has defended that choice repeatedly as a long-horizon calculation: "If you're building a team to solve a problem that might take a decade, it would be really bad short-term optimisation to hire the person who doesn't seem quite right."
His reasoning about early hires is almost mathematical: "No batch of 10 people will have as much an influence on the company as those first 10 people." Those people set the norms everyone after them absorbs. He also had a sourcing strategy that suited an unknown company: look for people who were "early in their career or undervalued in some way," and recruit from among users who were already passionate about the product, where "you kind of have an unnatural advantage over other companies."
He has admitted where this caution cost them. "We probably waited too long before we started scaling our international operations in a big way," he told Wired. Slow hiring is a deliberate tax, and Collison is clear-eyed about paying it.
Tactic: Before your next hire, ask whether this person will still look like the right choice in year ten, and if the answer is "probably," keep looking.
Principle 5
Make culture operational, not atmospheric
"A lot of companies talk about 'culture' like the force in Star Wars," Collison said in 2017, "this mysterious thing that floats around. But for us it's very operational. We try to be deliberate about it, write it down, talk to new hires about it on the first day and hold people accountable for it. It will come up in performance reviews. It could be a reason we promote someone."
Patrick, in Elad Gil's High Growth Handbook, made the companion argument: companies are "too precious" and "too apologetic" about culture, describing it in "milquetoast terms" instead of saying plainly that work will be "repeatedly designated as inadequate" and redone. The Collison approach is to make culture legible enough that people can opt out of it before joining.
John adds a counterweight that guards against ossification: "Don't worry about preserving your company's culture. The culture you have at 4 employees is not going to be the culture you have at 40, and that's okay." Writing culture down is not freezing it. It's making it revisable in the open rather than drifting in the dark. His sharpest formulation is also the most practical test: "Culture is what happens when the CEO isn't in the room."
Tactic: Pick one cultural value you claim to hold and name the specific performance-review behavior that would get someone promoted or not for it.
Principle 6
Refuse the assumed worldview of the exit
At seventeen, Collison sold a company and reportedly considered it a potential failure if that sale became his identity. At thirty-four he called the IPO question a symptom of "an assumed worldview" in which the purpose of building is "flipping something and moving on to the next thing." The consistency across seventeen years is the principle.
Staying private has costs the Collisons have managed deliberately: employees need liquidity, so Stripe ran tender offers and the 2023 Series I specifically to pay for it; investors need returns, so Sequoia offered its own limited partners up to $861 million in Stripe shares in 2024. The point is not that public markets are bad. The point is that the decision should follow from the work, not from a convention about what founders do at a given valuation.
S
The Same Answer at Every Scale
Stripe's founders on selling or listing, across a seventy-fold change in valuation.
2008
Auctomatic sold for $5M. John, 17, reportedly does not want to be defined by it.
2014
At a €1.3B valuation, Patrick: "We've no interest in selling the company."
2016
At $5B, Recode headline: "he's in no hurry to IPO."
2022
At ~$95B, John: "We're very happy as a private company. We'll let you know if that changes."
2024
At $65B, John: "We have no news to share, and when it's time to share, we'll share it."
2025
At $91.5B, John: "Why would you want to float it?… it is our life's work."
2026
Valued at $159B; bids $53B with Advent for PayPal. Still private.
Tactic: Write down, in one sentence, what you would do with your company if no one ever offered to buy it, and check whether your current plan serves that sentence.
Principle 7
Read the history before you claim the invention
Collison's curiosity is often presented as a personality trait. He presents it as an operating requirement. The Carlsen anecdote is about the link between depth of historical knowledge and mastery; the Apple-Amazon comparison is about holding multiple successful models in mind without collapsing into imitation. "I don't see how you could be that effective," he said of founders, "without being pretty curious on how to learn the most important mental models from this particular domain or function."
This shows up in Stripe's structure. A founder who opens a podcast interview by discussing industrial conglomerates is a founder thinking about how a multi-product company allocates capital across businesses with different economics. The Singleton Prize recognized exactly that. Stripe Press is the same instinct externalized: a company reissuing Hamming, Munger, and Licklider because the people who built the last generation of infrastructure solved problems Stripe now faces.
The warning embedded in this principle is the one he gave Wired: people "think they're the first people to come across a problem." They almost never are.
Tactic: Before your next major organizational decision, find two companies that faced it a generation ago and read what they actually did.
Principle 8
Let the incumbent's head start become its handicap
Collison calls it "the tyranny of being first": early adopters of a technology become trapped by the installed systems they adopted, accumulating "layers, mazes, and frictions that compound over time." His example is pointed. Startups run on Google Docs; "very large, established companies have giant Excel docs." Stripe's enterprise wave (Hertz, Amazon, Pepsi, Hershey's, the Church of England, the University of Oxford) is the commercial expression of this idea. Hertz, moving off legacy payment systems, saw a four-percentage-point improvement in online payment revenue.
The strategic implication is that an upstart's lack of history is an asset if it builds for the present rather than retrofitting the past. Stripe began with "a useful lack of knowledge about the industry," in John's words. It now sells that lack of legacy to institutions "from, you know, the 1600s."
The principle cuts both ways, and Collison seems aware of it. Stripe is now fifteen years old and itself an incumbent. The bets on stablecoins, Tempo, and agentic commerce read as deliberate attempts to avoid being tyrannized by its own head start.
Tactic: Identify the system your largest competitor adopted first and still depends on, and build the version that assumes it never existed.
Principle 9
Reframe the drawdown against the size of the market
Between 2021 and 2023 Stripe's valuation fell from $95 billion to $50 billion. It cut 14 percent of staff. The headline interview of the period was "I've just lost $4bn — and I'm more ambitious than ever." The response was operational, not rhetorical: efficiency until cash-flow positive, a liquidity raise for employees, and a public framing that measured the company against its total addressable market rather than its own prior peak.
"We're less than 2% of the overall digital payments market," Collison said at $1 trillion in volume. That sentence does real work. It moves attention from a valuation (which the market sets) to a share of a market (which the company earns). Employees whose equity has halved need a reason to believe the ceiling hasn't moved; this was it.
$
The Drawdown and the Recovery
Valuation and volume through Stripe's down round and back.
| Date | Valuation | Signal | Phase |
|---|---|---|---|
| Mar 2021 | $95B | $600M raise incl. Ireland's NTMA | Peak |
| Jul 2022 | $74B | Internal share price cut | Drawdown |
| Nov 2022 | — | 14% workforce reduction | Drawdown |
| Mar 2023 | $50B | $6.5B raise for employee liquidity | Reset |
| Feb 2024 | $65B | $1T volume; cash-flow positive | Rebuild |
| Feb 2025 | $91.5B | $1.4T volume | Rebuild |
| Feb 2026 | $159B | $1.9T volume, up 34% | New peak |
Tactic: In your next hard quarter, report your share of the total market alongside your decline from peak, and make sure both numbers are honest.
Principle 10
Divide the partnership by temperament, not just title
The Collison partnership has survived because it is not symmetrical. Patrick is CEO and the intellectual public face; John historically ran partnerships and sales, translates the engineering into the punchline, hosts the podcast, and walks the airfield. The skits at Sessions (the "deepfaking," the "TradFi" jab) are performed for thousands but encode something real: a relationship in which friction is metabolized as humor rather than escalated as conflict.
John's stated advantage is that the "meta issues of working style" were settled before Stripe existed. Most co-founders spend years discovering how they fight. The Collisons had known since one of them ordered a Sun workstation as a boy. The lesson for non-siblings is not "start a company with your brother." It is that unresolved working-style questions are a tax on every decision, and they should be resolved explicitly and early.
Together the brothers hold a controlling interest. That concentration is what makes the long horizon possible, and it makes the partnership itself load-bearing infrastructure. Its stability is a fiduciary matter.
Tactic: With your co-founder, write down who makes the final call in five recurring decision types, and who is the public voice for each.
Principle 11
Route around the infrastructure that failed you, then fix it for others
The pattern begins with a satellite dish in Dromineer and continues with Enterprise Ireland's refusal, the flight to Y Combinator, and the merchant-account maze. Each time, the Collisons paid to route around broken infrastructure. Then they built the fix for everyone else. Atlas exists so that a founder in Lagos or Havana doesn't have to move to California to incorporate. Stripe's backing of UL's software engineering programme exists so that the next Castletroy student, like the sign-language translator's builder Abutalha Alam, has a better path at home.
The same instinct powers John's Irish political interventions. His critique of planning delays, permitting, and "everything becomes a scandal" is the payments thesis applied to the state: needless friction is a choice, and someone should un-choose it. Critics call this naive about democracy, and the critique deserves a hearing. But the consistency is striking. He diagnoses infrastructure, not ideology.
Tactic: List the three obstacles you personally had to route around to get where you are, and pick one to remove for the people coming after you.
Principle 12
Steward the things you can't monetize
Weston Airport, bought with a consortium in 2021. Abbeyleix House, an eighteenth-century estate whose original owners moved a whole town to higher ground. Millbrook House, a ruin restored into Ómós, with thirty-eight jobs. A bog under a fifty-year community lease that the estate works alongside. None of these is a great investment by the standards of a man whose primary asset compounded 34 percent last year. That is precisely the point.
Stewardship, as The Currency observed, is "an unfashionable word in modern capitalism." Collison's version of it treats long-lived assets (companies, airfields, houses, towns) as things you hold on behalf of future users rather than things you extract value from. The habit cultivated in private, on physical things where neglect is visible, plausibly reinforces the habit at Stripe, where neglect is invisible until it's catastrophic. Payments, like runways and stone walls, are infrastructure: you notice them only when they fail.
Tactic: Take responsibility for one long-lived thing that will never pay you back, and maintain it as carefully as anything that does.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— John Collison, The Irish Times, 2015The first lines of code we wrote were in October 2009. We worked on it through the semester.
— John Collison, Stanford eCorner, February 2015If you're working on a startup that's a bad idea, it's going to feel like slow-going. But if you're working on a startup that's a good idea, it may feel like slow-going too.
— John Collison, Limerick Post, January 2014It's still early days for the payment systems that power the internet. It's outrageously difficult and expensive to move money between countries. Fraud and storing credentials still cause merchants hassle. So online payments is a great place to go hunting for problems worth solving.
— John Collison, foreword to Poor Charlie's Almanack (Stripe Press, 2023)His essays extol the virtues of free enterprise, yes, but also of doing business the right way, with integrity and rigor. Of taking your work very seriously, but never yourself.
— John Collison, The CurrencyI guess the general worry I have is, is there some complacency that we need to be quite paranoid about? Lots of our successes came out of quite radical ideas.
Maxims
- Be the customer first. The most precise product specification is your own irritation, shared by strangers.
- Slow is not a verdict. Good ideas and bad ideas both feel slow early; measure the intensity of your best users instead.
- Talk up to the user. Assume the person on the other side is talented and serious, and design for them, not for the wheelbarrow.
- Hire for the decade. A person who doesn't seem quite right is an expensive way to save a few months.
- Write the culture down. If it isn't in the performance review, it's atmosphere, not culture.
- Question the exit. The assumption that you build things in order to sell them is a worldview, and you are allowed to decline it.
- Read before you invent. Most problems you think are new have been solved, badly or well, by someone a generation ago.
- Beware the head start. Being first to adopt a system can become the reason you can't leave it.
- Measure against the market, not the peak. A halved valuation matters less than a two percent share of a fifty-trillion-dollar market.
- Keep the runway open. Some of the most important things you own will never pay you back; maintain them anyway.
