Tobi Lütke: Empowering a World of Rebels | Knowledge Project Podcast

KEYPOINTS

• 01

  In chaos, protect the North Star, not the plan. Roadmaps are overrated.

  Shopify became a symbol of the pandemic shift to digital, then of the pullback. Lütke says no one in a company works on the stock, and if people start to, something is wrong. What matters is the company's purpose: making entrepreneurship simpler. When shelter-in-place closed stores, the move was obvious: get retailers' catalogs online and ship curbside pickup. His alternative to roadmaps is a clear view of what merchants need, an honest model of your capabilities, and teams ready to pick the best next task. When reality changed, Shopify sorted all its work into 'helpful' and 'not helpful' and stopped about 60% of it.

• 02

  A founder's most valuable power is subtraction, funded by 'legitimacy'.

  Non-founder companies are like river stones: smooth, strong at measurable operations, and run through stakeholder management and plans that carry legitimacy. Founder companies are spiky volcanic rock. Lütke calls legitimacy one of the most underappreciated resources and credits an essay by Vitalik Buterin. A company's founding story deposits it into a kind of bank account, and only a founder still present can draw on it. 'If you say no to a thing, you say no to one thing. If you say yes, you say no to every other thing.' Adding is easy, removing takes courage. Satya Nadella pulled off subtraction at Microsoft, but it's 'ten or a hundred times harder' for a non-founder.

• 03

  Best practices are just average with lower risk. Not taking risks is the real career risk.

  Companies revert to the industry mean, and industry quality is low. Doing what everyone else does is the only option with no risk, which is exactly why it's average. Innovation requires being different, which means sometimes underperforming. At Shopify a miss is 'the successful discovery of something that didn't work', and people get coached for not aiming high. He wouldn't be proud of a company that only maintains its products and buys all its innovation. Even HR should innovate. Shopify built a new compensation system from first principles.

• 04

  What gets called bureaucracy is often sediment layers of individually good decisions.

  People rarely call processes that work 'bureaucracy', so the word flags places where first-principles thinking can streamline a lot. But some checks and balances only look bureaucratic. A person patches a hole in a wall, someone with more perspective knows the floor is being remodeled, and someone else knows the company is going 'digital by design' with no offices. Ideas should come from everywhere, while strategic decisions come from wherever the right perspective sits. He gives executives Parkinson's Law, an 80-page book on how organizations get silly.

• 05

  Compensation shouldn't depend on what the stock did the week grants were issued.

  Stock options and RSUs piled up as tax optimizations and turned into complex instruments with wildly varying outcomes. Lütke asks why an employee's lifetime earnings should depend on whether Russia invaded Ukraine the week before a grant date. Shopify now gives each person one annual number, then lets them use sliders to choose the mix of cash and stock. A low share price buys more shares. He'd like to add options such as a charity bucket. The aim is to 'treat people like adults'.

• 06

  Business should be re-derived from programming principles.

  He argues computing sits between language (abstraction) and math, a practical 'applied computational philosophy' that early universities stuck in basements. Engineers spend their formative years building systems that survive non-deterministic chaos like packet loss, so their systems design is '100 years ahead' of management books. A compensation team looking up comparables and percentiles is really processing data. Make it an executable, auditable function, change one file, and watch the company model rebuild. As Nash took lessons from poker for game theory, business should take lessons from engineering.

• 07

  Decisions are easy. Finding the inputs is hard. Rerun the function when the inputs change.

  Most disagreements disappear once you state unstated assumptions. Judge a decision by its process, not its outcome ('resulting'). His test: was there 20 minutes I could have spent to get a missing input? Shopify built great offices. When the input 'can people go there?' flipped, and then where staff lived changed, it called 'digital by design' early, around May 2020, to remove ambiguity for staff and recruits. Second-order effects often outweigh first-order ones, he argues: anti-nuclear sentiment in Green movements kept coal in use and contributed to climate change.

• 08

  Your value isn't your title. Most people should choose range over specialization.

  Anxiety often comes from tying self-worth to a role or a credential. At Shopify, roles are 'lures to great people', and hiring asks whether the company will be better with you. True specialists know who they are. For everyone else, the Pareto principle means you can get to eight out of ten in five fields, and learning speed compounds. He calls 'jack of all trades, master of none' 'one of the dumbest sentences I've ever heard'. A longtime colleague told him his high opinion of people leaves him constantly disappointed. Lütke would rather help people see their potential.

• 09

  Play infinite games, and pick finite games that serve them.

  From James Carse's Finite and Infinite Games: a finite game like tennis has rules and a winner, and dies if conditions change, say gravity at 7G. An infinite game, like fitness or education, exists only to keep playing, so change is welcome. School turns education into a fractal of finite games of tests and grades and forgets the point. Lütke's infinite game is giving people superpowers through technology. The most successful people, he says, choose finite games that do 'dual duty' for their own infinite one.