Piece No. 3: Granola
Notes below.
I. Context
Granola was founded in March 2023 by Chris Pedregal and Sam Stephenson. The two had always been excited about the tools we use for work - and while the loose idea for Granola had been in their minds for a while - the launch of ChatGPT in 2022 cemented their belief that something new needed to be built to move towards the AI first world to come.
II. Choices
A | Thoughtfully tie together content and time
Many note taking or documentation apps don’t meaningfully integrate time as a concept into the product - they are surfaces intended for content permanency - inherently implying time matters less
Granola’s focus on meeting notes requires it to pull a calendar concept into the UI. The main surface, though, is neither calendar nor document editor. Instead, the landing experience is a list of upcoming and prior meetings, creating a new paradigm from which users imagine the product experience
Given the value of Granola is in the overlap of notes and calendar, it seems they’ve also chosen to keep the editor, where users can take their own notes, extremely simple. A new page only allows users to write text - adjustments to format or styling of content are not possible
Perhaps this helps users understand that in Granola, AI is meant to do the writing - users need only execute the right actions (e.g. share, summarize, ask)
B | Integrate core user needs into one action-rich surface
Most tools separate out collaborative flows from the AI interaction interface. This is likely a remnant of developing these features sequentially versus a user defining need for the two actions to be distinct - Granola ties both of these experiences into the sidebar panel (which is default open, but can be hidden)
Surface aside, the key Granola value proposition lies in creating a sense of ease when in meetings - so visible nudges of core actions like tagging attendees, suggesting shares, writing follow-up emails make delightful ‘one-clicks’ inevitable
There is also creativity in the implementation of the actions themselves. For example, the LinkedIn URL for meeting attendees is actually an auto populated LinkedIn search that uses the names and email domains from meeting invites
C | Offer easy reference to source material
We’re still in a world where comfort with AI tools is growing - so having access to the transcript to verify or recall live conversations is critical
Users aren’t forced to navigate to a page menu to find these references in Granola - instead they need only hover on each note to understand specific details quickly
III. Conclusions
Granola solves a productivity pain point with thoughtfulness and restraint; its elegance lies in a focused yet powerful product experience
There’s growing industry enthusiasm around AI-enabled workflows - a natural next step would be to explore what sits upstream or downstream of meetings. One could easily imagine Granola adding functionality around CRMs or applicant tracking systems - the operating systems for users with meeting-heavy workflows
However, doing so is also likely to challenge the design discipline Granola has exhibited, and could make differentiation difficult over time



Loved this piece: the "AI does the writing, users just execute actions" framing nailed why Granola sticks where other AI note tools don't. I run my whole knowledge stack on Granola + Notion, and ended up building GranolaSync to push my meeting notes from Granola straight into the right Notion databases (per-class for my MBA, per-project for my startup).
Curious whether you see the same dynamic at Notion when AI features ship.