Notion Plans a 30% Hiring Binge as Ivan Zhao Goes All-In on AI Agents
The Information reports Notion will grow headcount roughly 30% this year, with new roles largely devoted to building and selling AI — the clearest signal yet that the $11B workspace app is becoming an AI company.
In a year when most software companies are trimming headcount and calling it “AI efficiency,” Notion is doing the opposite. On August 29, The Information reported that Notion plans to increase its headcount by roughly 30% this year, and that the new roles will be largely devoted to building AI products and selling them — the concrete workforce behind CEO Ivan Zhao’s all-in bet that the workspace tool becomes an AI-agent company.
The move lands at a moment when the “AI replaces SaaS” narrative has never been louder — and when Notion’s own numbers suggest the bet is already paying off.
What the report actually says
The Information’s piece is short on the surface but sharp in its framing: Notion, the San Francisco-based workspace company, is planning a hiring binge at a scale unusual for its ~1,920-person size (Unify-reported headcount as of April 2026), and the roles are overwhelmingly AI-centric — engineers building agents, and the sales organization required to land them in enterprises.
That is a different kind of announcement than a model launch or a funding round. It is a company allocating its scarcest resource — people — toward one thesis. And the thesis is agentic: Notion’s product line since late 2025 has been Notion Agent (chat that executes tasks), AI Meeting Notes, and above all Custom Agents, the February 24, 2026 launch that lets any team build agents that read and write inside their workspace.
The commercial engine behind that thesis is already visible. Custom Agents run on a credit system introduced May 4, 2026 — $10 per 1,000 Notion credits — while Notion Agent, AI Meeting Notes, and Enterprise Search remain bundled and unlimited on higher tiers. For a company whose legacy business was per-seat subscriptions, a usage-based AI layer on top is the difference between a productivity tool and a platform.
The numbers behind the bet
Three data points explain why Notion can afford to hire into an AI downturn narrative:
Revenue is accelerating, not decelerating. Sacra estimates Notion hit $865M ARR in July 2026, up from $610M at the end of 2025 — roughly +53% year-over-year — with the acceleration attributed to AI attach rates compounding. Back in September 2025, when Notion launched its first agent and crossed $500M in annual revenue, CFO Kasra Kothari noted that the share of customers paying for AI add-ons had climbed from 10–20% the year before to 30–40% earlier in 2025, and “recently” even higher.
Adoption of the agent platform is real. At the Custom Agents launch in February, Notion disclosed that early testers had already built over 21,000 Custom Agents, and that 2,800 agents were running around the clock inside Notion itself. CEO Ivan Zhao has separately described an internal ratio of roughly 700 deployed AI agents against about 1,000 employees — a striking inversion of the usual automation talking point: Notion is not replacing staff with agents, it is hiring more staff to build and sell more agents.
The balance sheet has room. CNBC’s Disruptor 50 profile from May 2026 lists Notion at $330M total funding and an $11B valuation, with a December 2025 report that the company was weighing a $12B tender offer. Zhao still owns at least 30% of the company — an unusually high founder stake at this scale — which means the hiring binge is being funded from a position of control, not desperation.
Why this matters beyond Notion
The interesting tension in this story is what it says about the broader “AI eats SaaS” debate. The bear case for companies like Notion has been that frontier-model providers or autonomous coding agents would simply absorb the workspace layer — why pay for a structured database-of-everything when an agent can generate whatever artifact you need on demand?
Notion’s answer is the opposite bet: the workspace itself is the moat. Agents need context — documents, project history, meeting notes, the structured knowledge of who-did-what — and Notion sits on exactly that context for millions of teams. Every Custom Agent that reads a company’s wiki to draft a brief or chase a deliverable makes the underlying workspace harder to leave, not easier. In that framing, hiring 30% more people to build agents isn’t a defensive move; it’s a land grab for the agentic workflow layer.
There’s also a talent-market signal here. While OpenAI and Anthropic hoard researchers and the infrastructure players hoard capital, the application layer — where most software actually gets used — is quietly re-staffing for AI. Notion hiring aggressively for AI product and AI sales is a template other mid-size SaaS companies are watching: the path to surviving the agent era may run through aggressively re-hiring for it, not shrinking.
The open questions
The bet is not without risk. Usage-based AI pricing is young, and enterprises are still learning to forecast credit consumption; a backlash against unpredictable AI bills is a live possibility. The Custom Agents credit system already drew user debate when it launched in May. Competition is brutal — Microsoft’s Copilot ecosystem, Google’s Gemini-infused Workspace, and a wave of agentic startups all want the same workflow surface. And a 30% headcount jump is operationally hard at any company, let alone one whose founder-CEO famously obsesses over minimal building blocks.
But the direction is unambiguous. Notion spent 2025 proving customers would pay for AI; it is spending 2026 staffing up as if AI is the product. As Zhao put it when describing the shift: building AI enterprise software now feels like building a video game — the surface is simple, and everything underneath is physics. The hiring binge is what it looks like when a company decides to build the physics team.