Vicky Firth
CX Lead at Granola
Granola
AI-powered notepad designed for professionals in back-to-back meetings to capture and enhance notes without using intrusive recording bots.
Industry
SaaS
Support team
5
Monthly Volume
2000
website
www.granola.ai
Granola sustained 10% week-over-week user growth between Series A and Series B, going from 5,000 weekly active users to a $1.5B valuation in fifteen months. Their support team didn't scale the headcount in step. They went the other way, investing in product impact rather than ticket throughput, and built an automation layer that now handles 88% of first responses.
Plain has been the support platform throughout, and somewhere along the way it stopped being a support tool. It became infrastructure: the queryable data layer the team builds around.
Granola is the AI-powered notepad that captures meetings without sending a bot into the call. Plain has been their support platform since early 2025, through 100x user growth, two funding rounds, and a jump from Series A startup to $1.5B unicorn.
The challenge
Before Plain, Granola was running support out of Slack and a shared Google inbox, with engineers and non-technical staff rotating through "support weeks" that pulled them off their core work. Plain replaced all of that in early 2025, just as the company was entering the steepest part of its growth curve.
The team stayed deliberately small while volume climbed, and Plain made that possible by absorbing every inbound channel into one workspace so the team could stay close to the signal:
Slack, email, in-app forms, and Help Center requests in one unified inbox
A Help Center built directly in Plain, with one-click escalation from articles to a human
Custom customer cards built on Plain's API, surfacing user context directly
Company-wide inboxes (privacy@, security@) routed into Plain for shared ownership
Linear integration, linking Plain threads to engineering issues
The automation layer underneath that inbox was built incrementally, over 18 months, and has resulted in 29 workflows that now run across Granola's inbox, layered across five functions:
Noise filters that close spam before it reaches the queue
AI classification rules that label every ticket the moment it arrives
Routing logic that sends security reports and enterprise escalations to the right person automatically
Timed acknowledgments that set customer expectations while the investigation runs
Internal agent that does the investigation itself
When the last pieces clicked into place in March 2026, automated actions jumped from 57% to 88% within 30 days.
"By default, I assume I can do something in Plain somehow. I know that if I want to build a workflow, I can probably do it. And if there's a gap, the API can fill it in."
Vicky Firth
CX Lead at Granola
Plain as infrastructure: where the support team actually works
Granola's internal AI agent reads the full thread context the moment a ticket arrives, queries the help center, checks against a continuously-updated index of Granola's codebase (synced from GitHub), and pulls customer logs.
Plain's API connects all of those sources. The same data layer feeds the agent alongside Amplitude, Stripe, Slack, and Granola's own meeting notes.
A few patterns shaped how the team worked.
Plain data, queryable from anywhere. Granola’s team created an internal AI agent built entirely on Plain's API. Plain feeds the agent alongside every other data source the team needs, and the surface area keeps expanding without rebuilding the support tooling underneath.
Plain replaces the dashboard layer. Vicky queries Plain through the API and MCP to ask what topics are coming through, what's going up or down, what themes correlate with other product changes. No fixed dashboards, no rigid reporting layer.
Plain hands the team direct control over their interface. Custom customer cards are built and modified by the team itself. What they want surfaced on every ticket is one configuration change away, instead of a quarter-long argument with engineering about prioritization.
Results: A shorter loop between problem and fix
Plain threads have become the starting point for product investigation. Support engineers pull tickets into Granola's internal agent, ship pull requests for small bugs, and bring product context directly into peer-level conversations with engineers.
Every ticket reviewed by AI before a human responds
Every thread is classified and labelled by AI the moment it arrives. 10 automated classification workflows sort tickets into bugs, questions, payment issues, security reports, and more before an agent opens the inbox. Threads that surface technical issues get a deeper pass: Granola's AI agent reads the thread, checks the indexed codebase and logs, and hands the agent a ready-made investigation. Agents open tickets with root-cause context already in hand.
5 inbound channels consolidated into a single workspace
Slack, email, in-app forms, and Help Center requests all flow into a single Plain inbox. So do company-wide addresses like privacy@ and security@, which means cross-functional ownership lands in the same place customer support does. The team doesn't context-switch between tools to track a thread.
100x user growth handled on Plain
Granola scaled from around 5,000 weekly active users at Series A to a $1.5B valuation at Series C on the same support platform throughout. No migration, no parallel systems, no rebuilt internal workflows when volume picked up. The Plain workspace they set up early on is the same one running today, through the Series A launch, the Series B raise, and the enterprise push that came with Series C.
"By giving your support team the tools that are usually outside their remit, the loop is shorter. If we can edit our own internal tools, that makes us faster to fix people's problems. If we can fix the bugs ourselves, even better."
Vicky Firth
CX Lead at Granola
Looking ahead
Granola has never re-evaluated Plain against another support tool, and the Plain footprint inside the company keeps expanding. Enterprise customers and new product surfaces like Spaces, the MCP server, and expanded APIs are changing what support sees: different inbound patterns, higher-stakes SLAs, more sensitive customer relationships.
Vicky's focus is on extending what the support team can solve upstream, fixing root causes rather than optimizing the metrics around tickets that shouldn't have existed in the first place. The framing she brings to that work has stayed consistent through every stage of the company's growth.
"I strongly believe a customer support team can leverage so much in the business and bring so much insight. If you see them as central to your business, rather than this kind of outsourced cost center, they can bring so much value to everyone."
Vicky Firth
CX Lead at Granola
