How Notion's AI Pivot Offers Lessons for Airlines

From Docs to AI Workspace: A Blueprint for Airlines

Notion started as a note-taking app. Now it's a $6 billion revenue business, half of which comes from AI products. The company's journey from a struggling startup to an AI powerhouse is a masterclass in reinvention. But what does this have to do with airlines? Everything, if you're willing to look past the obvious.

Airlines are stuck in a legacy trap. They run on decades-old systems, massive bureaucracies, and processes that haven't changed since the jet age began. Notion's story shows that even a successful company can rebuild itself when the ground shifts. And the ground is shifting for aviation right now—AI is changing how passengers book, how crews are scheduled, how maintenance is predicted, and how loyalty programs work.

Why Airlines Need a 'Kyoto Moment'

Notion's founder, Ivan Zhao, hit a wall early on. The company had no product-market fit, money was running out, and the team was demoralized. His radical move? Cut the team, move to Kyoto, and start over. In a small apartment, with just him and his co-founder, they coded, ate, and coded some more. It wasn't about a new direction—it was about stripping away everything that didn't matter and focusing on the core question: what problem are we really solving?

Airlines need that kind of moment. Not literally moving to Kyoto, but mentally. The industry is obsessed with adding features—better apps, more seat options, flashy lounges—but rarely asks the deeper question: what is the fundamental job we do for travelers? If you started an airline today, with AI as your co-pilot, would you still have check-in counters? Would you still have a call center? Would you still schedule crews the same way?

Most airlines would say no. But they keep patching the old model because it's what they know. Notion's Kyoto moment was about admitting that the old way was dead and being willing to build from first principles.

The GPT-4 Wake-Up Call for Aviation

When GPT-4 came out, Ivan Zhao realized the software world had been interrupted. It wasn't a trend to watch—it was a tectonic shift. Notion had already shipped an AI writing feature before ChatGPT launched, but Zhao wanted more. He wanted agents that could understand context, retrieve information, and execute tasks. That vision took a year and a half of painful experimentation, because early models weren't good enough.

Airlines are in a similar spot. Many have dabbled with chatbots, but they're mostly dumb FAQ bots. The real opportunity is in AI that can handle the messy, context-heavy work: rebooking disrupted passengers, managing irregular operations, predicting maintenance needs, personalizing the entire travel journey. But it won't happen overnight. The technology is still maturing, and airlines need to accept that they'll have to experiment, fail, and iterate—just like Notion did.

From Toolbox to AI Workspace: What Airlines Can Build

Notion's pivot was from a toolbox (docs, notes, databases) to an AI workspace where AI has access to all your context. For airlines, that context is huge: booking data, flight status, customer history, crew schedules, maintenance logs, and real-time operational data. The airline that can wire AI into that context will win.

Think about what that could look like. An AI that knows you're on a delayed flight, automatically rebooks you, sends your new boarding pass, and updates your hotel and car rental—without you lifting a finger. An AI that helps gate agents make split-second decisions during a weather crisis. An AI that predicts which passengers are likely to churn and offers them a proactive upgrade. That's the airline AI workspace.

Organizational Jazz: Breaking the Airline Hierarchy

Ivan Zhao talks about Notion as a jazz band, not a military band. Jazz has structure, but it's flexible—players respond to each other, improvise, and create something new in the moment. Airlines, by contrast, are the ultimate military bands. They run on strict hierarchies, standard operating procedures, and departmental silos. That works when everything is predictable, but it's a liability when the world changes overnight—like, say, a pandemic or an AI revolution.

Notion didn't eliminate hierarchy; it made it more fluid. Design, product, and engineering now work shoulder-to-shoulder, experimenting with models and user feedback in real time. Airlines need to break down the walls between operations, customer service, and commercial teams. A disruption isn't just an ops problem—it's a customer experience problem, a revenue problem, and a brand problem. The airline that can improvise across functions will handle crises better and delight customers more.

Hiring for Taste and Initiative, Not Just Experience

Notion changed its hiring. With AI leveling the playing field on basic skills, the company now looks for taste and initiative. Taste means knowing what good looks like and why. Initiative means actually doing things, not just talking about them. For airlines, this is a wake-up call. The industry is full of people with decades of experience, but how many have true taste in customer experience? How many take the initiative to try something new without waiting for a committee?

Airlines should hire for curiosity, energy, and the ability to work with AI tools. That doesn't mean discarding experience—it means complementing it with fresh perspectives. Notion uses a barbell structure: a few senior architects who provide direction, and a bunch of young, hungry engineers who execute with AI. Airlines could do the same: pair veteran pilots and veteran customer service reps with digital natives who can build AI solutions.

Rethinking Marketing and Sales in the Sky

Notion split its marketing department. One side focuses on storytelling, close to the product; the other on sales support, close to revenue. Airlines have traditionally had marketing as a separate, centralized function—and it shows. Airlines often market at scale, but rarely with the agility of a startup.

For airlines, the storytelling side should be about building a community around the airline—think loyalty beyond miles, like a lifestyle brand. The revenue side should support sales to corporate clients, using AI to personalize proposals and streamline RFPs. And let's not forget direct selling: airlines have been trying to cut out OTAs for years, but they need to make their own channels so good that passengers willingly come back.

One of Notion's mistakes was trying to reinvent enterprise sales from first principles. Zhao eventually admitted that human trust still matters. Airlines should take note: AI can enhance sales, but you still need human relationships for big contracts, especially with corporate clients who care about data security and compliance.

AI Isn't Just About Cutting Costs—It's About Changing the Job

Many airlines see AI as a way to cut costs—fewer call center agents, less paper, more automation. That's short-sighted. Notion's AI transformation wasn't just about efficiency; it was about changing the fundamental nature of knowledge work. For airlines, AI should change the fundamental nature of air travel.

That means moving from reactive to proactive. Instead of waiting for a passenger to complain, AI can anticipate problems and fix them before they escalate. Instead of static loyalty programs, AI can create dynamic offers that adapt to each traveler's behavior. Instead of one-size-fits-all service, AI can tailor the entire journey—from booking to baggage—to individual preferences.

The airlines that thrive will be those that use AI to redefine what an airline is. Maybe it's not just a mode of transport, but a travel companion that handles everything. Maybe it's not just a flight, but an experience that starts months before takeoff.

Practical Steps for Airlines to Start Their AI Rebuild

• Start with a pilot: Pick one operational area—say, disruption management—and build an AI solution that works with your real data.
• Break silos: Form cross-functional teams (ops, IT, customer experience) that work together on AI projects, not in isolation.
• Hire for adaptability: Look for people who can code, but also those who have taste and initiative. Train them on AI tools.
• Invest in data infrastructure: AI is only as good as the data it sees. Ensure your systems are connected and data is clean.
• Don't outsource your AI strategy: Like Zhao says, founders (and CEOs) must get their hands dirty. Understand the technology yourself.

Notion's story proves that even a successful company can reinvent itself. Airlines have the scale, the data, and the customer base. All they need is the courage to rebuild.