Ivan Zhao tested thirty different shades of warm white on Notion’s office walls before choosing one.
Not three. Not “let’s look at a few options.” Thirty.
His reasoning? “It hurts the eyes if you just see ugly things. You have more visual pain from the eyes. If the chair looks ugly, the wall looks ugly, it hurts.”¹
I can already hear every CFO reading this: “We’ve got a burn rate crisis and this guy’s worried about whether the white is warm enough?”
This is founder mode.
A year ago, Paul Graham broke the internet with his essay naming the phenomenon. Graham described that founder mode exists. But he openly admits that nobody’s quite nailed down what it is.
So I did what any reasonably obsessive person would do: spent an unreasonable amount of time analysing conversations with founders who supposedly “have it.” Ivan Zhao from Notion. Tobi Lütke from Shopify. Brian Chesky from Airbnb. Gaurav Misra from Captions. The usual suspects who’ve built things that made them very rich (and the rest of us very jealous). I even reflected on my time reporting to Co-founder and CEO Eugenio Pace in the very early days of Auth0.
Here’s the pattern: Founder mode isn’t some mystical cognitive ability bestowed upon the chosen few. It’s not a special gene or secret handshake or whatever Silicon Valley wants you to believe.
It’s the willingness to care deeply about things that make you look obsessive, weird, or dumb to everyone else.
That’s it. I reckon that’s the secret.
This only looks dumb until it works. Then it gets repackaged as “visionary genius who just knew.” But while it’s happening? You look like you’re wasting everyone’s time on the wrong things.
Why 30 Shades Matters
Back to Ivan’s walls.
Most CEOs would’ve said, “Get me three options, I’ll pick one, let’s move on.” Ivan spent personal time on this because he connected dots others couldn’t — or couldn’t be bothered — to see.
Here’s what he saw: office aesthetic affects creativity, which affects product quality, which affects Notion’s success. It’s a cascade of second and third-order effects most people can’t see because they’ve been trained to focus only on “important” things.
He also implemented a no-shoes policy. One communal table for meals. Obsessed over details that would make most CEOs delegate faster than you can say “ah… I have a board meeting.”
The principle: Founders see second and third-order effects of “small” decisions that others miss entirely.
That willingness to spend capital (time, attention, credibility) on things that make people question your priorities? That’s the mode.
The Tobi Tornado
At Shopify, there’s a phenomenon people call the “Tobi Tornado.”
Tobi Lütke walks into projects — projects people have worked on for weeks or months — and says, “This isn’t going to work, let’s stop it now.”
Tornado. Destruction. Everyone scrambling.
Most managers would rather eat glass than do this. It feels terrible. Creates conflict. Makes you look indecisive. (“But wait, you approved this three weeks ago?”)
But Tobi’s willing to look like the bad guy who kills projects and doesn’t value people’s work. Why? Because compressing time to the correct path matters more than looking consistent or kind.
Not that I’m jealous of his ability to just… stop things. (I am absolutely jealous. The number of times I’ve let something limp along because I didn’t want to be the tornado? Too many to count. I’ve written before about courage being a bottleneck for businesses.)
The principle: Founders optimise for being right, not for looking smart or kind.
Manager mode optimises for smooth team dynamics, respecting sunk costs, appearing decisive, and being consistent.
Founder mode optimises for getting to the right answer faster, even if it looks like you don’t know what you’re doing, even if it pisses people off, even if you contradict last week’s you.
The CEO Who Wrote the Docs
In 2015, I joined Auth0, and one of the first things that struck me as odd? My boss’s obsession with documentation. Eugenio Pace spent his early days at Auth0 obsessing over documentation.
Not “reviewing” documentation. Not “approving” documentation. Writing it. Himself. The co-founder and CEO.
Eugenio was grinding through developer guides, code samples, API references. Making sure every endpoint was explained. Every error message was clear. Every integration path was documented.
I could just imagine the board: “We need you closing deals, not writing docs. That’s what we hire technical writers for.”
Here’s what everyone missed: Auth0 was selling to developers. And developers don’t trust sales pitches — they trust documentation. Good docs weren’t just marketing. They were the product moat.
Most competitors treated documentation as a back-office cost centre. Something you outsource after product-market fit. Auth0 made it a core value proposition from day one.
The result? Auth0 won adoption from solo developers, startup CTOs, and Fortune 500 enterprises — all because the CEO cared enough about docs to write them himself until he’d built the judgment to know what “great” looked like.
The principle: Founders obsess over what others might dismiss as a “support function” because they see how it compounds into competitive advantage.
By the time Auth0 sold to Okta for $6.5 billion, that “waste of CEO time” had become the reason developers chose Auth0 over every alternative.
Three “Dumb” Moves That Actually Work
It gets weirder. Because it’s not just caring about strange details or killing projects. It’s doing things every piece of conventional wisdom says not to do.
The “Doesn’t Scale” Move
Brian Chesky famously reviewed every single Airbnb listing himself in the early days. Every. Single. One. Photography, descriptions, everything.
Critics said, “That doesn’t scale.”
I mean, they weren’t wrong. It didn’t scale. You can’t personally review millions of listings. That’s just math.
But Brian wasn’t trying to review listings forever. He was trying to understand what made a good listing at a level impossible to achieve any other way.
You can’t A/B test your way to taste. You can’t hire a consultant to build judgment for you. You have to do the unscalable thing — look at hundreds of listings with your own eyeballs — so you build the operating system in your brain that will scale.
Years later, Brian looks at any product decision and knows instantly if it meets the Airbnb standard. Not because he ran a regression analysis. Because he built the judgment through doing the thing everyone said was a waste of time.
The principle: Founders do unscalable things not because they’re naive about scale, but because they’re building the judgment that will make future decisions better.
The Secret Roadmap
Gaurav Misra at Captions has two roadmaps.
Public roadmap: feature requests, things customers ask for, the usual stuff that makes sense.
Then there’s the secret roadmap.
In Gaurav’s mind, these are things nobody has ever asked for. If a user were shown something on it, they might be like, “I don’t need this.”
Example: the eye contact feature that shifts your eyes to look at the camera. Nobody asked for it. If you’d described it beforehand, users would’ve shrugged. “Yeah… nah, don’t need that.”
Now? Copied everywhere. One of those features that once you experience it, you can’t imagine living without it.
Here’s the insight: If you only build what customers ask for, you’re in a feature parity race with every competitor. You’re building a commodity.
This is either visionary or delusional. I guess the only difference is whether it works.
The principle: Founders build things that make them look out of touch until users experience it and can’t live without it.
The Technical Debt Paradox
One more, because this makes engineers physically uncomfortable.
Gaurav: “As a startup, your job is to take on technical debt because that is how you operate faster than a bigger company.”²
Every engineer just winced. Every coding bootcamp drills into people’s heads that technical debt is bad. The thing you avoid. The thing that kills you later.
And here’s Gaurav actively encouraging it.
His reasoning: “If the company fails, that future engineer will never be hired and all this won’t matter anyway. So it’s like financial debt. You can create products you wouldn’t be able to build with a small team by taking on strategic technical debt.”³ (I have since used this very argument with engineers!)
Why does this look dumb? Because everyone’s optimising for code quality, maintainability, and doing it “right.”
But Gaurav’s playing a different game: maximise survival. And in that game, speed trumps perfection. Because if you’re too slow, you’re dead, and your beautiful, debt-free codebase dies with you.
The principle: Founders make bets that look reckless because they’re playing a different game — maximise survival, not maximise quality.
The Deeper Thing: Taste vs. Consensus
So what’s really happening here? What’s underneath all these examples?
Founder mode is what happens when you trust your taste over consensus — even when you can’t fully articulate why.
Tobi Lütke: “The most powerful unquantifiable things in the world of business are fun and delight. If all the metrics are pointing down but everyone says, ‘My God, I’m having so much more fun,’ I think the very next thing that will happen is all metrics will start going up.”⁴
Ivan Zhao’s belief: What is best for the thing you’re building is if you create “a balance… what’s authentic to yourself but also what’s useful for others.”
So, what’s the pattern? These founders make decisions that look irrational by standard business logic. Obsessing over aesthetics. Killing projects with momentum. Taking on debt. Building features nobody wants.
Why? Because they’re optimising for unquantifiable things — taste, beauty, fun, authenticity — that downstream create quantifiable results.
You can’t put “fun” in a spreadsheet. You can’t A/B test “authentic.” You can’t run a regression on “this white feels warmer.” But these things matter, and founders in founder mode trust they matter even when they can’t prove it.
Here’s the optical illusion:
While it’s happening: Looks obsessive, micromanaging, naive, wasteful and a little bit dumb. After it works: Looks like visionary genius who “just knew” But the truth? The only difference is outcome.
Brian Chesky reviewing every listing looked dumb — until Airbnb became a verb and he became a billionaire.
Ivan testing 30 whites looked dumb — until Notion hit $10 billion valuation.
Eugenio writing documentation looked dumb — until Auth0 sold for $6.5 billion on the strength of developer love.
If you’re never being told “that’s a waste of time” or “that doesn’t scale,” you’re probably not in founder mode. You’re in manager mode — optimising the consensus path, following best practices, doing the reasonable thing.
Which is fine! Most businesses are built this way. But you won’t look like a genius in retrospect. You’ll look like someone who executed a known playbook well.
How to Practice “Looking Dumb”
Three exercises to build the muscle:
Exercise 1: The “Nobody Else Cares” Audit
List five details about your product/company that:
- You care deeply about
- Others think is a waste of time
- You can’t fully articulate why it matters
If you can’t list five, you might be over-indexed on consensus. Too worried about looking dumb to trust your own judgment.
(And if you listed fifty, you might just be stubborn and overly picky. There’s a difference between taste and being difficult!)
Exercise 2: The “Doesn’t Scale” Test
What’s one thing you could do this week that:
- Definitely won’t scale to 1000x
- Would teach you something you can’t learn any other way
- Everyone would say is a waste of time
Now go do it. Actually do it. See what you learn.
Exercise 3: The Tobi Tornado Drill
Pick one current project. Ask yourself:
- If I knew for certain this wouldn’t succeed, would I stop it now?
- Am I keeping it alive because of sunk cost, momentum, or to save face?
- Is there a better approach we haven’t considered?
If the honest answer is “we should stop,” have the courage to be the tornado.
It’ll feel terrible. People will be confused. You’ll look indecisive.
But if you’re right, you just gave everyone back weeks or months to work on something that actually matters.
Permission to Look Dumb
Here’s why “founder mode” became a meme: it named something transgressive.
It gave people permission to trust themselves over the crowd, even when they look obsessive, naive, or dumb.
Manager mode optimises for looking smart, being data-driven, avoiding conflict, respecting sunk costs, following best practices, maintaining team morale, and appearing consistent.
Founder mode optimises for being right, building something great, compressing time, trusting taste over data, ignoring “best practices,” being willing to piss people off, and contradicting yourself when you learn something new.
Yeah, you’re going to get it wrong sometimes. You’ll test 30 shades and still pick the wrong one. You’ll kill a project that had potential. You’ll build a feature nobody wants and it’ll stay that way.
But the alternative — always doing the reasonable thing, always following best practices, always optimising to look smart — won’t get you there either. It’ll just make you look… well, reasonable, at best.
And, you know… reasonable is kinda good. But I reckon there’s greater reward in looking a little dumb.
¹ Ivan Zhao, Lenny’s Newsletter, “Inside Notion: Ivan Zhao (CEO and co-founder),” (March 5, 2025.) ² ³ Gaurav Misra,“How to Win in the AI Era: Ship a Feature Every Week, Embrace Technical Debt, Ruthlessly Cut Scope, and Create Magic Your Competitors Can’t Copy,” Lenny’s Podcast (March 26, 2025). ⁴ Tobi Lütke, Tobi Lütke’s leadership playbook: Playing infinite games, operating from first principles, and maximizing human potential (founder and CEO of Shopify), Lenny’s Podcast (Feb 2, 2025).