Notion Shuts Down Its Email App, Bets Users Want AI Agents Instead
Notion is shutting down its Skiff-influenced email client, citing a shift toward AI agents managing inboxes rather than users interacting with email directly. The move signals Notion's full pivot to agent-first workflows for communication.
Original sourceNotion is pulling the plug on its email application — a product it inherited through the 2024 acquisition of privacy-focused productivity suite Skiff — after concluding that the majority of its users have migrated toward AI agents to handle their inboxes. The company is not replacing the email client with another interface; it is replacing it with the premise that an agent should handle most of what a person previously did inside an email app.
The shutdown reflects a broader strategic bet Notion has been making over the past year: that the interface layer for productivity tasks like email triage, scheduling, and document routing is collapsing into an agent layer. Rather than building a better inbox, Notion is arguing the inbox itself is the wrong abstraction. Users who stayed on the email product were a minority, and Notion is choosing to concentrate resources on the agent infrastructure instead.
The Skiff acquisition was framed at the time as a way to bring end-to-end encrypted collaboration into Notion's orbit. The email component was a natural extension of Skiff's privacy-first suite, covering docs, pages, and mail under one roof. Killing the email app this quickly suggests the thesis either didn't hold — users didn't want their email in Notion — or the agent-first alternative proved compelling enough that maintaining a traditional client became a distraction.
For users still on Notion Mail, this is a forced migration back to whatever they used before, or toward Notion's agent tooling if they want to stay in the ecosystem. The practical question is whether Notion's AI agents are actually capable enough to replace an email client today, or whether this is a roadmap decision dressed up as a product retirement.
Panel Takes
The Skeptic
Reality Check
“'Most users use AI agents instead' is doing a lot of work in that headline — Notion should publish the actual retention numbers before framing an acquisition wind-down as a user behavior success story. The more plausible read: they acquired Skiff for the team and the encrypted docs technology, the email product never hit meaningful adoption in the Notion context, and 'agents ate it' is a cleaner narrative than 'we couldn't compete with Gmail.' I'd want to see what percentage of Notion's paying base actually activated the email product before accepting the stated reason at face value.”
The Founder
Business & Market
“Killing a low-adoption product and framing it as a strategic pivot is a legitimate move — the mistake would be keeping a zombie feature alive to justify an acquisition price. The real question is whether Notion's agent infrastructure has the unit economics to replace what the email client was supposed to deliver: daily active engagement, a reason to open Notion for something other than docs. Agents are a stickier wedge than email if they actually work, but Notion is now in direct competition with every AI-native productivity layer, not just email clients.”
The PM
Product Strategy
“The job-to-be-done for email — process communications, decide what needs action — doesn't go away just because Notion is shutting down the interface for it. Notion is betting users will hire an agent for that job instead of hiring a client, which is a coherent product opinion, but only if the agent can handle the full lifecycle: triage, reply, schedule, archive, escalate. Half an agent that handles 70% of email and punts the rest back to a separate inbox is a worse product than a mediocre email client. The completion question is everything here.”
The Futurist
Big Picture
“The falsifiable thesis here is: within three years, the primary interface for asynchronous communication will be an agent that surfaces decisions rather than a client that surfaces messages. Notion is betting early on that trend, and if they're right, shutting down the email client now is the correct move — you don't maintain a legacy interface while trying to normalize the replacement. The second-order effect nobody is discussing: if agents handle inbox triage across your entire user base, Notion becomes a node in communication infrastructure, not just a document tool, which is a fundamentally different kind of leverage over enterprise workflows.”