Tobi Lütke
Co-founder and CEO of Shopify
Where to hear and read about Tobi Lütke1 episode or test
Tobi Lütke is the co-founder and CEO of Shopify, a high school dropout and programmer who started it on almost no money and has run it for 21 years. The lesson most worth taking from him is that a company built deliberately from the founder's own first principles beats one copied from the standard playbook, as his near-failure while imitating a conventional CEO and his COVID-era rebuild show.
I found the cheat code to always being right is just change your opinion every time you get better information.
Tobi Lütke: 21 Years of Building Shopify
Tobi Lütke: the story and the lessons
1How Tobi Lütke started Shopify on almost nothing
Shopify grew out of a snowboard shop Lütke was building. Lütke started out building a snowboard store, and the market pulled him toward the software instead.
He treats the market as an intelligence to tap, and calls a company the socially accepted way to spend fourteen hours a day obsessing over one idea.
He took no salary for about four years. Lütke lived with his parents-in-law and built much of Shopify at an Ikea desk in his wife Fiona's old bedroom, and his father-in-law covered payroll at one point.
He says money mostly gives you more of what you already were, including looseness with spending, which is why huge seed rounds worry him. "Making every dollar count is something that I'm so glad."
Lütke credits luck for much of Shopify's early survival. Shopify came within a week of running out of money many times, and he believes most reruns of the first six years would fail.
By comparison he calls the IPO a hard financing event and COVID mostly annoying.
Strategy games trained his decision-making. StarCraft taught him that information is everything, that attention is a resource, and that no decision is right outside its context.
Mirrored maps and Elo ratings gave a clean signal of improvement, and he deliberately played moves not in the books. He now says running six coordinating AI agents feels the same.
In 2014 he wrote down his mission: cause more entrepreneurship. Preparing a summit talk, Lütke realized he loves building things that share the feeling of a first sale, and said he would run Shopify only while it fits that mission.
He says he would not still be CEO if AI had not arrived, since stable times suit other leaders. He calls a merchant reaching a billion before Shopify did its greatest outcome.
2How Lütke nearly broke Shopify by playing a conventional CEO
After the IPO he delegated everything and lost sight of the work. Lütke tried to be a serious public company CEO, a role he describes as a 60-year-old with a legal background in a suit, and handed everything to a conventional executive team.
People learned to steer his attention only to the projects he liked, and trust eroded. "So it almost killed the company, honestly."
Weak competition let the problem hide. With no strong rival keeping Shopify honest, even good-looking results could not be trusted.
He sensed something was going wrong before COVID, and says COVID saved the company.
COVID made him rederive every plan from axioms. Lütke told the company all plans were invalid, then reviewed every project himself in 16-hour days and cancelled about 60% of them.
He thinks of plans as a decision tree: when an assumption like free movement breaks, prune back to the root and go forward again. The review found boondoggles such as a Toronto team building supermarket features that nothing else in the product could use.
He replaced every executive within a year, mostly with founders. Lütke asked the founders of acquired companies in his Slack channel for help, made many of them executives, and moved engineers into big roles. Every one of those promotions worked.
He says founders are irritants who never settle, and that companies wrongly park them in skunkworks teams. "No, you don't get to put them in founder daycare."
The reset meant building on his own intuition. Fifteen years in, Lütke asked what a company built entirely on his own intuitions and biases would look like.
His reasoning was that if cosplay failed, the opposite might work, and at worst he would find the midpoint.
3Shopify OS: Lütke's company engineered like software
Lütke models the company in code and reconciles reality to it. He turned compensation PDFs into machine-readable files, wrote Python to model titles, levels, ratios and pay, and ran a solver to compute what Shopify should look like.
It began as a bare GitHub project whose first version was utterly incorrect. He calls it a desired state system, like React, and HR's job is to move the real company toward the model.
The model exposed hidden choices and removed politics. It found 5,500 different titles across 8,000 people, and now a request for 50 salespeople shows what it costs engineering.
"Think about how much politics this removes."
Great individual contributors can earn as much as executives. Modeling showed his best source of great engineers was his own management team, so Shopify built mastery tracks in every discipline.
Someone a hundred times better than peers can out-earn a vice president, a pattern he sees in machine learning labs too.
He welcomed an 80% stock drop. At its peak Shopify traded beyond 50 times revenue, and Lütke felt relief because he had cash, avoided debt and the unit economics were strong.
He treats the stock market as a betting machine and his job as building the best company.
Employees now choose their own pay mix every quarter. After underwater options left staff feeling owed, Shopify gave everyone sliders to split salary across stock, RSUs, Shop Cash and cash, rebalanced quarterly.
A tool can lock in stock value for three years. Doing it across countries was a legal nightmare, but he says it is very popular. "You have full agency, and you make this choice."
4How Lütke took Shopify public his own way
He went public in 2015 against Silicon Valley advice. For a Canadian company of about 800 people selling to American firms, name recognition, liquid stock for hiring and learning to explain the business outweighed a few annoying calls a quarter.
First trades valued Shopify at roughly $1 to 1.5 billion, so public buyers got near-venture returns.
Judge the status quo by the same standard as the new idea. Lütke says most people live in a vetocracy where one good objection kills a new idea while the status quo's flaws go unlisted.
When everyone agrees, he tests the opposite, and says Plan B is usually better but rarely tried.
Shopify built a public version of itself rather than going public. Only one executive had ever worked at a public company, so the team looked for rules where they had agency.
Instead of a CEO before slides, they made a documentary roadshow video; investors said they looked forward to meeting him, and later IPOs copied it.
Staying outside Silicon Valley let Shopify build differently. He found Silicon Valley gave him others' highlight reels to compare against Shopify's average.
He stayed in Ottawa expecting to find about a thousand like-minded people, since he could not outspend Google.
5Lütke's rules for being different
Copying caps you below excellence. Lütke says an art student's copy is never a Van Gogh, and a copied seven out of ten is one you do not understand.
A six built from your own axioms can be iterated past seven because you have mastery over it. He agrees with James Dyson: make it different, even if it's worse.
Treat other companies as rivals, not competitors to track. Competitor-obsessed firms become reactive copyists, while rivalry pushes you to your best.
He cites Agassi's rivalry with Sampras, one Agassi largely built in his own head.
Change your opinion whenever better information arrives. Lütke says consistency helps you look good, not be good, and expects his talks to contradict each other.
Sunk cost makes people defend what they wrote down.
He attacks his own past work on purpose. Each year he plays a corporate raider who bought a bankrupt Shopify and lists what he would change, and he writes hit pieces on work he is proud of.
He once found his old code impressive and called it one of the saddest days of his life, because he had stopped learning. Code at Shopify has stewards, not owners.
Every executive must explain how Shopify does it differently. His instruction is "Be a company engineer", and each executive speaks yearly at a podcast or conference on how Shopify handles that topic and why it is better.
Saying it out loud pulls people to make it true. The aim is a company that never sleepwalks into anything.
6How Lütke gives teams autonomy through environment, not policy
Teams get a box where autonomy is traded for accountability. Projects explore in a prototype phase, then a review with engineering and design leads and Lütke moves them to build, where he takes the risk and they decide.
Teams can rehearse the review with an AI trained on his past reviews. The system came from asking what a competent CEO would have had in place when COVID hit.
He changes the environment so the right action is intuitive. Pods fit about five people and feel crowded at seven, so teams self-organize at the size he wants without a rule.
He sees policies as downside protection that also caps the best people. He estimates each step up in team size costs about 10x in productivity, and Shopify's 3,500 R&D people work in many small teams.
Office design gets the same care as product. Co-founder Daniel Weinand held that no one can be more creative than the space around them, so meeting rooms sit in the middle, desks by windows, and stairs connect floors.
Lütke calls Norman doors the worst error because they make people do the wrong thing. Everything in the offices is bought from Shopify stores.
An internal podcast records why decisions were made. Context explains past decisions so staff stop blaming odd choices on "Tobi said".
Early episodes were cut to 23 minutes, the average Shopify commute.
7How Tobi Lütke hires
He asks one question: have you started a company before? Lütke says his pre-COVID bets on who would contribute most were entirely wrong, while founders adapted fastest.
In a crisis, he says, some people go to zero and some to a hundred.
Interviews probe high agency through life stories. Lütke walks through a candidate's life, then zooms in on a moment something went wrong and asks how they reacted minute by minute.
A high school dropout, he ignores credentials and hires for spikes. Many employees ran Shopify stores first. "we're looking for is high agency experience."
He refuses to write down hiring criteria. People who read a hiring list are the ones good at following lists, so they can perform what you want without being it.
"The people you are trying to avoid hiring are the people who would study a cheat sheet very well."
His job is to make Shopify worth working for. Lütke cut politics so a large company feels small, and Shopify avoids being the highest payer because pay-first people do poorly there.
"My job is to make a company worthy for the best and brightest to work for."
8How Lütke keeps learning
Read widely and switch genre every few books. Lütke calls reading the one weird trick for gaining range, but finds business books biased toward their authors' specialty.
He refused to become the engineer who sees only engineering problems. Reading about Larry Ellison's sales-incentive problem showed him a mistake he could have avoided.
He uses affirmations to rewrite his own identity. Lütke cured his fear of public speaking by writing for about 10 minutes a day for a week that he loves it.
He calls the brain a retrospective narrative alignment mechanism and affirmations "the dumbest trick that works". He also schedules notes to himself after each keynote.
He expects AI to reward whoever rederives fastest. His memo making AI the first reflex drew pushback, and he expects it to read like saying the sky is blue in two years.
He sees AI making entrepreneurship a common plan B.
Tobi Lütke: 21 Years of Building Shopify
Tobi Lütke explains how he nearly broke Shopify by imitating a conventional public company CEO, then rebuilt it around his own intuitions with founder executives and a company designed like software. The story is worth hearing because the 2021 reset, built on COVID-era project cuts and a new operating system for the company, is what he credits for making Shopify work.
Shopify grew out of a snowboard shop Lütke was building. Lütke started out building a snowboard store, and the market pulled him toward the software instead.
He treats the market as an intelligence to tap, and calls a company the socially accepted way to spend fourteen hours a day obsessing over one idea.
He took no salary for about four years. Lütke lived with his parents-in-law and built much of Shopify at an Ikea desk in his wife Fiona's old bedroom, and his father-in-law covered payroll at one point.
He says money mostly gives you more of what you already were, including looseness with spending, which is why huge seed rounds worry him. "Making every dollar count is something that I'm so glad."
Lütke credits luck for much of Shopify's early survival. Shopify came within a week of running out of money many times, and he believes most reruns of the first six years would fail.
By comparison he calls the IPO a hard financing event and COVID mostly annoying.
Strategy games trained his decision-making. StarCraft taught him that information is everything, that attention is a resource, and that no decision is right outside its context.
Mirrored maps and Elo ratings gave a clean signal of improvement, and he deliberately played moves not in the books. He now says running six coordinating AI agents feels the same.
In 2014 he wrote down his mission: cause more entrepreneurship. Preparing a summit talk, Lütke realized he loves building things that share the feeling of a first sale, and said he would run Shopify only while it fits that mission.
He says he would not still be CEO if AI had not arrived, since stable times suit other leaders. He calls a merchant reaching a billion before Shopify did its greatest outcome.
After the IPO he delegated everything and lost sight of the work. Lütke tried to be a serious public company CEO, a role he describes as a 60-year-old with a legal background in a suit, and handed everything to a conventional executive team.
People learned to steer his attention only to the projects he liked, and trust eroded. "So it almost killed the company, honestly."
Weak competition let the problem hide. With no strong rival keeping Shopify honest, even good-looking results could not be trusted.
He sensed something was going wrong before COVID, and says COVID saved the company.
COVID made him rederive every plan from axioms. Lütke told the company all plans were invalid, then reviewed every project himself in 16-hour days and cancelled about 60% of them.
He thinks of plans as a decision tree: when an assumption like free movement breaks, prune back to the root and go forward again. The review found boondoggles such as a Toronto team building supermarket features that nothing else in the product could use.
He replaced every executive within a year, mostly with founders. Lütke asked the founders of acquired companies in his Slack channel for help, made many of them executives, and moved engineers into big roles. Every one of those promotions worked.
He says founders are irritants who never settle, and that companies wrongly park them in skunkworks teams. "No, you don't get to put them in founder daycare."
The reset meant building on his own intuition. Fifteen years in, Lütke asked what a company built entirely on his own intuitions and biases would look like.
His reasoning was that if cosplay failed, the opposite might work, and at worst he would find the midpoint.
Lütke models the company in code and reconciles reality to it. He turned compensation PDFs into machine-readable files, wrote Python to model titles, levels, ratios and pay, and ran a solver to compute what Shopify should look like.
It began as a bare GitHub project whose first version was utterly incorrect. He calls it a desired state system, like React, and HR's job is to move the real company toward the model.
The model exposed hidden choices and removed politics. It found 5,500 different titles across 8,000 people, and now a request for 50 salespeople shows what it costs engineering.
"Think about how much politics this removes."
Great individual contributors can earn as much as executives. Modeling showed his best source of great engineers was his own management team, so Shopify built mastery tracks in every discipline.
Someone a hundred times better than peers can out-earn a vice president, a pattern he sees in machine learning labs too.
He welcomed an 80% stock drop. At its peak Shopify traded beyond 50 times revenue, and Lütke felt relief because he had cash, avoided debt and the unit economics were strong.
He treats the stock market as a betting machine and his job as building the best company.
Employees now choose their own pay mix every quarter. After underwater options left staff feeling owed, Shopify gave everyone sliders to split salary across stock, RSUs, Shop Cash and cash, rebalanced quarterly.
A tool can lock in stock value for three years. Doing it across countries was a legal nightmare, but he says it is very popular. "You have full agency, and you make this choice."
He went public in 2015 against Silicon Valley advice. For a Canadian company of about 800 people selling to American firms, name recognition, liquid stock for hiring and learning to explain the business outweighed a few annoying calls a quarter.
First trades valued Shopify at roughly $1 to 1.5 billion, so public buyers got near-venture returns.
Judge the status quo by the same standard as the new idea. Lütke says most people live in a vetocracy where one good objection kills a new idea while the status quo's flaws go unlisted.
When everyone agrees, he tests the opposite, and says Plan B is usually better but rarely tried.
Shopify built a public version of itself rather than going public. Only one executive had ever worked at a public company, so the team looked for rules where they had agency.
Instead of a CEO before slides, they made a documentary roadshow video; investors said they looked forward to meeting him, and later IPOs copied it.
Staying outside Silicon Valley let Shopify build differently. He found Silicon Valley gave him others' highlight reels to compare against Shopify's average.
He stayed in Ottawa expecting to find about a thousand like-minded people, since he could not outspend Google.
Copying caps you below excellence. Lütke says an art student's copy is never a Van Gogh, and a copied seven out of ten is one you do not understand.
A six built from your own axioms can be iterated past seven because you have mastery over it. He agrees with James Dyson: make it different, even if it's worse.
Treat other companies as rivals, not competitors to track. Competitor-obsessed firms become reactive copyists, while rivalry pushes you to your best.
He cites Agassi's rivalry with Sampras, one Agassi largely built in his own head.
Change your opinion whenever better information arrives. Lütke says consistency helps you look good, not be good, and expects his talks to contradict each other.
Sunk cost makes people defend what they wrote down.
He attacks his own past work on purpose. Each year he plays a corporate raider who bought a bankrupt Shopify and lists what he would change, and he writes hit pieces on work he is proud of.
He once found his old code impressive and called it one of the saddest days of his life, because he had stopped learning. Code at Shopify has stewards, not owners.
Every executive must explain how Shopify does it differently. His instruction is "Be a company engineer", and each executive speaks yearly at a podcast or conference on how Shopify handles that topic and why it is better.
Saying it out loud pulls people to make it true. The aim is a company that never sleepwalks into anything.
Teams get a box where autonomy is traded for accountability. Projects explore in a prototype phase, then a review with engineering and design leads and Lütke moves them to build, where he takes the risk and they decide.
Teams can rehearse the review with an AI trained on his past reviews. The system came from asking what a competent CEO would have had in place when COVID hit.
He changes the environment so the right action is intuitive. Pods fit about five people and feel crowded at seven, so teams self-organize at the size he wants without a rule.
He sees policies as downside protection that also caps the best people. He estimates each step up in team size costs about 10x in productivity, and Shopify's 3,500 R&D people work in many small teams.
Office design gets the same care as product. Co-founder Daniel Weinand held that no one can be more creative than the space around them, so meeting rooms sit in the middle, desks by windows, and stairs connect floors.
Lütke calls Norman doors the worst error because they make people do the wrong thing. Everything in the offices is bought from Shopify stores.
An internal podcast records why decisions were made. Context explains past decisions so staff stop blaming odd choices on "Tobi said".
Early episodes were cut to 23 minutes, the average Shopify commute.
He asks one question: have you started a company before? Lütke says his pre-COVID bets on who would contribute most were entirely wrong, while founders adapted fastest.
In a crisis, he says, some people go to zero and some to a hundred.
Interviews probe high agency through life stories. Lütke walks through a candidate's life, then zooms in on a moment something went wrong and asks how they reacted minute by minute.
A high school dropout, he ignores credentials and hires for spikes. Many employees ran Shopify stores first. "we're looking for is high agency experience."
He refuses to write down hiring criteria. People who read a hiring list are the ones good at following lists, so they can perform what you want without being it.
"The people you are trying to avoid hiring are the people who would study a cheat sheet very well."
His job is to make Shopify worth working for. Lütke cut politics so a large company feels small, and Shopify avoids being the highest payer because pay-first people do poorly there.
"My job is to make a company worthy for the best and brightest to work for."
Read widely and switch genre every few books. Lütke calls reading the one weird trick for gaining range, but finds business books biased toward their authors' specialty.
He refused to become the engineer who sees only engineering problems. Reading about Larry Ellison's sales-incentive problem showed him a mistake he could have avoided.
He uses affirmations to rewrite his own identity. Lütke cured his fear of public speaking by writing for about 10 minutes a day for a week that he loves it.
He calls the brain a retrospective narrative alignment mechanism and affirmations "the dumbest trick that works". He also schedules notes to himself after each keynote.
He expects AI to reward whoever rederives fastest. His memo making AI the first reflex drew pushback, and he expects it to read like saying the sky is blue in two years.
He sees AI making entrepreneurship a common plan B.
More from Shopify
GuestArchie Abrams
VP of Product and Head of Growth at Shopify
Shopify
1 source · 17 tips
GuestTimothy Davis
Head of performance marketing at Shopify; former consultant to Pinterest, LinkedIn, Redfin and Eventbrite
Shopify1 source · 23 tipsLuc Levesque
Chief Growth Officer at Shopify, former Facebook/Meta executive and TripAdvisor VP of Growth
Shopify1 source · 17 tips