The Notion Founder Story: Noodles, Kyoto, and a Rewrite That Saved the Company

By 2015, Ivan Zhao and Simon Last had removed from their lives almost everything that wasn't writing code and eating noodles. Three years into building their startup, they did the thing founders are most afraid to do: they threw away the code that ran their product. They were nearly out of money. They had just laid off their only colleagues. And to stretch the cash that was left, they had moved from San Francisco to Kyoto, Japan — a city less than half as expensive to live in. The company that would become a $10-billion-plus workspace used by more than 100 million people was, at that exact moment, two people in a small apartment deciding whether to start over or quit.

This is the story of how Notion almost died, and what the founders did instead.

The Idea: LEGO for Software

Notion didn't begin as a note-taking app. It began as a philosophy. Ivan Zhao had a background in cognitive science and design, and an obsession that long predated the company: the belief, inherited from computing pioneers like Doug Engelbart and Alan Kay, that computers were supposed to augment human intellect — to make ordinary people more capable — and that somewhere along the way software had instead become a pile of rigid, disconnected apps that bent users to their shape rather than the other way around.

His frustration was specific. To get work done you needed a document tool, and a separate spreadsheet, and a separate database, and a separate wiki, and a separate task tracker — none of which talked to each other, all of which locked their power behind code that normal people couldn't write. Zhao's vision was to collapse all of that into one flexible system: a kind of LEGO for software, where a non-technical person could snap together their own tools without writing a single line of code. In 2013 he and Simon Last founded Notion Labs in San Francisco and raised roughly $2 million from angels — the classic mix of friends, family, and friends of friends — to chase it.

The First Version Failed

The first version of Notion was, by the founders' own account, too ambitious and too fragile. It tried to do everything, and the technology underneath couldn't hold the weight. It crashed. It was hard to use. The grand vision of "build any tool you want" turned out to be paralyzing rather than liberating for the people who tried it — given infinite flexibility and no starting point, most users just bounced.

The market's verdict was indifference. The product didn't resonate, the funding was running thin, competitors were everywhere, and the team was staring at the end. This is the part of the founder story that the highlight reels skip: not a clean pivot, but the slow, grinding realization that the thing you've poured three years into isn't working, and the money to fix it is almost gone.

Kyoto: The Reboot

Most teams in that position do one of two things — raise a desperate round on a story they no longer believe, or shut down. Zhao and Last did something stranger and braver. They let go of their only employees, shrank the company back down to just the two of them, and moved to Kyoto, where the rent and the cost of living were a fraction of San Francisco's. Then they deleted the codebase and started over.

The Kyoto period has become a piece of startup folklore for a reason. The two founders coded for something like eighteen hours a day, lived cheaply, and stripped the product back to its essentials. The runway was so thin that Zhao later admitted he borrowed bridge money from his own mother to keep the lights on — "thanks mom for the bridge," as he put it. There was no team to manage, no investors to perform for, no office politics. Just two people, a hard problem, and the freedom that comes from having almost nothing left to lose.

What came out of that apartment was the idea that makes Notion Notion: the block. Instead of separate apps for documents, tables, and lists, everything would be built from one universal unit — a block — that could be text, an image, a database row, a to-do, a whole embedded page. Snap blocks together and you get a document. Arrange them differently and you get a database, a wiki, a planner. The LEGO metaphor finally had its LEGO brick. The block-based architecture born in Kyoto is still the foundation the entire product stands on today.

The Launch That Looked Like Success — And the Years That Weren't

Notion 1.0 launched in March 2016 and immediately did something remarkable: it became the number-one product on Product Hunt not just for the day or the week, but for the month. After years of silence, the reboot finally got applause.

And then — this is the honest part — it still wasn't enough. The 2016 product was built on a tech stack that couldn't keep up; it crashed, it frustrated people, and the viral launch didn't convert into the kind of durable growth that builds a real company. Zhao has been candid that real traction didn't arrive until 2018, with the launch of Notion 2.0. The near-death in Kyoto wasn't the climax of the story. It was the moment the company earned the right to spend two more unglamorous years grinding the product into something people couldn't put down.

What the Story Actually Teaches

It's tempting to flatten this into a clean parable — "believe in your vision and it'll work out." The real lessons are sharper and less comfortable:

• Shrinking can be the brave move. Laying off your only employees and going back to two people is humiliating in the moment and rational in hindsight. Zhao and Last bought themselves time and focus by getting smaller, not bigger. Burn is the enemy of second chances.

• The vision was right; the first execution was wrong. They didn't abandon "LEGO for software." They abandoned the code and the bloat, and kept the idea. Knowing which to throw away is the whole skill.

• A viral launch is not traction. Being number one on Product Hunt felt like winning. Two more hard years separated that moment from an actual business. Founders routinely mistake a spike of attention for a foundation of retention.

• Constraints manufacture focus. The cheap Kyoto apartment wasn't just a way to save money — it was a way to remove everything except the work. Sometimes the best thing you can do for a struggling product is take away every option except finishing it.

Notion is now valued in the tens of billions, used by over a hundred million people, and held up as one of the great examples of community-led growth in software. But the part worth remembering isn't the valuation. It's the apartment in Kyoto, the deleted codebase, the bridge loan from mom, and two founders deciding that the right response to almost dying was to start over and build the thing properly. The company that sells "augmenting human intellect" was forged by two humans who refused to quit at the exact moment quitting would have been reasonable.