Summary of Key Points
Notion’s CEO, Ivan Zhao, is a master at “restarting businesses.” The first time this happened was when the company was on the verge of collapse; he laid off everyone and moved to Kyoto to rebuild from scratch, finding the right product-market fit. The second time was when GPT-4 emerged, and he led a team of 500 people in a complete transformation towards AI. The interview explores his two restarts, organizational changes in the AI era (replacing the “marching band” model with the “jazz mode”), the reengineering of talent strategies (focusing on taste and initiative over experience), adjustments to the company structure (disbanding the CMO position), and his unique cultural perspective that views the company as a form of “religion,” all reflecting a deep understanding of technological trends and human nature.
#### I. Two Reboots from the Brink: From Kyoto to Cancun
Ivan’s two restarts were both driven by a sense of urgency, but the approaches were different:
- The First Restart in Kyoto (2015): In the early stages of Notion, the team spent four to five years without finding the right “product-market fit” (PMF), and they ran out of funds. Ivan decisively laid off everyone and moved to Kyoto with co-founder Simon. The reason was simple: Tokyo’s apartments were too small, while Kyoto offered larger, cheaper homes that hadn’t been damaged during World War II. They focused solely on coding in Kyoto, working in a cycle of “code → eat → code,” and eventually, the product took off in 2019. He described this as an “instinctive choice; otherwise, they would have failed.”
- The Second Restart in Cancun (2023): When they gained early access to GPT-4, Ivan felt it was like a profound spiritual experience, realizing that AI could change everything, and they had to transform or be left behind. With a team of 500 people, many internally questioned the move (thinking it was another bubble similar to those in the cryptocurrency market), but Ivan persisted. The initial year and a half of developing AI agent products was unsuccessful until the underlying models improved, after which revenue began to grow again.
The common factor in both restarts was a willingness to make bold decisions for survival, trust in intuition, and focusing on core issues in a quiet environment.
#### II. A New Organizational Model for the AI Era: Jazz Mode Instead of Marching Band
Ivan proposes the “Jazz Mode” as an alternative to the traditional “marching band” model:
- Marching Band: Everyone follows a fixed script, with the conductor making all the decisions, leaving no room for improvisation—similar to hierarchical management in traditional companies where employees can only follow orders.
- Jazz Mode: There is a basic harmonic framework (such as the company’s core goals), but individuals are encouraged to improvise and contribute creativity. This flexibility is essential in the AI era due to rapid technological changes, which cannot be managed with rigid processes.
Ivan says he doesn’t want to act like a “conductor of a marching band”; if everything is delegated, the entrepreneurial spirit is lost. In the AI era, “jazz musicians” (those who are proactive and can improvise) become more valuable—senior engineers set the direction, while junior engineers use AI to execute tasks, resulting in higher efficiency than having multiple senior employees working independently.
#### III. A Major Shift in Talent Strategy: Focus on Taste and Initiative Over Experience
AI has leveled the playing field in terms of capabilities (e.g., coding and writing). Therefore, Notion now prioritizes these two qualities when hiring:
- Taste: Values and a sense of direction—what do you want to bring to the world? In which direction will you move? AI cannot make these decisions for you.
- Initiative: The ability to drive things forward on your own, rather than waiting for instructions.
The team has become “barbell-shaped,” with highly experienced individuals (who manage the architecture and direction, as AI is not good at big-picture thinking) working alongside newly graduated employees (who use AI to execute tasks quickly). For example, a senior architect can lead 2-3 junior engineers, and together, they produce more than several senior employees working alone.
The hiring process has also changed: For sales positions, the first round of interviews does not involve reviewing resumes; candidates are asked to create something using Notion to demonstrate their practical skills. Designers need to be capable of doing PM tasks (working with clients and driving projects), as it’s difficult for PMs to learn design skills, but designers can understand PM logic.
#### IV. Major Organizational Changes: Disbanding the CMO Position
This year, Notion reorganized its marketing department by dissolving the CMO role and dividing it into two teams:
- Content Narrative Team: Closely aligned with the product, as product iterations are too fast for traditional marketing methods. They directly engage with social media and communities (such as Notion’s fan base) to promote the product through content.
- Sales Support Team: Responsible for generating leads and working directly with sales.
The reason for this change is to eliminate middle layers; information used to go through the CMO, but now both teams work directly, improving efficiency. For example, designers are part of the Content Narrative Team, creating creative content without waiting for marketing instructions.
The sales approach has also changed: They tried a new product-driven growth (PLG) model but found it ineffective; corporate clients still prefer to communicate with people. They hired a former CRO from GitHub and paired them with “predatory” sales managers who respected the traditional sales system, which proved successful.
#### V. Company Culture: Essentially a “Religion,” with Regular All-Staff Meetings
Ivan views the company as a form of “religion” with shared values that are reflected in the product. He compares the company to some of the most successful companies in history, such as the Catholic Church (which has remained unchanged since 2000), where Jesus is the founder and Paul is the sales director, using a referral system and viral marketing. The company’s all-staff meetings are like church services, held weekly with great effort to make them impactful, sometimes even bringing tears to employees’ eyes.
He believes people need something to believe in to find meaning in their lives, and the company culture serves as that “system of faith.” For example, he forces himself to hold these meetings (even though he is introverted) because employees need to hear the direction from him to build trust. Now, AI helps with preparing the meeting transcripts.
#### Advice for Stranded Founders
Ivan advises: “Listen to your intuition. If you feel stuck and know it’s time to restart, do so decisively.” In the AI era, market opportunities are abundant, and inaction carries greater risk than taking a chance. For companies with revenue below 50 million or slow growth, it’s better to seize the AI opportunity and start over.
Ivan’s own routine includes waking up at 7 AM to make coffee, thinking and writing at home, exercising at noon, working until midnight, and reading on weekends. He uses a 25-minute schedule to create dedicated time for reflection.
What is most inspiring about this interview is Ivan’s unconventional approach: he doesn’t blindly follow traditional management practices; he dares to restart a large company and combines AI with human nature. His discussions cover not only technology but also culture, history, and human behavior, adding depth to the business conversation.