Yardstick Research tear-sheet / AI sales cohort
Granola
Identity
- Founded: March 2023 [THIRD-PARTY — https://sifted.eu/articles/ai-notetaking-startup-granola-hits-unicorn-status]
- HQ: London, United Kingdom [THIRD-PARTY — https://sifted.eu/articles/ai-notetaking-startup-granola-hits-unicorn-status, https://uk.linkedin.com/company/meetgranola]
- Legal entity: [UNKNOWN] not disclosed publicly; privacy contact at privacy@granola.so suggests an operating entity at granola.so separate from the granola.ai brand domain.
- Funding: $192M total raised. Most recent: $125M Series C at $1.5B valuation, led by Danny Rimer at Index Ventures with Kleiner Perkins, Lightspeed, Spark, and NFDG, announced March 25, 2026. Prior: $43M Series B at $250M valuation led by NFDG (May 2025); $20M Series A (October 2024). [THIRD-PARTY — https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/, https://techcrunch.com/2025/05/14/ai-note-taking-app-granola-raises-43m-at-250m-valuation-launches-collaborative-features/]
- Headcount: 87–94 in early 2026 (49% growth in the prior six months). [THIRD-PARTY — search-cached LinkedIn https://uk.linkedin.com/company/meetgranola, https://sifted.eu/articles/ai-notetaking-startup-granola-hits-unicorn-status]
- Recent news (last 12 months):
- March 25, 2026 — $125M Series C at $1.5B valuation. Spaces (team workspaces), personal API, enterprise API, and updated MCP server shipped concurrently. [THIRD-PARTY — https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/]
- February 2026 — Model Context Protocol server released. [THIRD-PARTY — https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/]
- May 2025 — $43M Series B at $250M valuation; collaborative features. [THIRD-PARTY — https://techcrunch.com/2025/05/14/ai-note-taking-app-granola-raises-43m-at-250m-valuation-launches-collaborative-features/]
- May 2025 — Tenable security advisory TRA-2025-07 publicly disclosed: hardcoded AssemblyAI key in an iOS TestFlight build, exposed via unauthenticated /get-feature-flags endpoint. Scope: about 300 alpha-testing iOS users. Granola revoked the key March 11, 2025, shipped signed short-lived tokens, and published a public post-mortem. [THIRD-PARTY — https://www.tenable.com/security/research/tra-2025-07, https://docs.granola.ai/help-center/policies/security-reports/post-mortem-assembly-ai-api-key-exposure]
- Latent narrative tension — Vendor logos and Series C messaging frame Granola as an enterprise AI app, but every published independent review still treats the product as a per-user purchase. The team-administration surface (Spaces, enterprise admin controls) shipped weeks before this tear sheet was written and is not yet third-party measured. [ESTIMATED from contrast between https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/ and https://tldv.io/blog/granola-review/]
- Archetype: AI notepad for back-to-back meetings; bot-free device-audio capture; per-user productivity tool with a thin team layer added at Series C. Complement to a sales stack, not a stack on its own.
Total score: 75 / 100
- Stage fit:
- Foundation (<40 readiness): yes — free Basic tier and a frictionless install make this the lowest-friction way for a founder-led sales motion to start capturing call context.
- Pilot (40–59): yes — Business at $14/user/month covers a 1–5 person founding-sales team end-to-end, including HubSpot for the orgs that use it.
- Scale (60–79): conditional — works as a complement to a sequencer + dialer + CRM stack; does not replace any of them, and the manual per-note CRM sync is a real friction at 10+ reps.
- Optimization (80+): conditional — Enterprise tier and Spaces are weeks-old at the time of writing; the team-administration surface is not yet third-party measured. Salesforce-anchored revenue orgs remain blocked on integration depth.
- One-line verdict: The bot-free AI notepad for founders, account executives, and small revenue teams who write their own notes during calls — strong on capture and per-seat economics, soft on CRM workflow depth and team-administration maturity.
Headline numbers
| Metric | Value | Evidence |
|---|---|---|
| Reply rate vs cohort baseline | n/a — Granola is not a sequencer, no outbound send surface | [ESTIMATED — granola.md §"Deliverability infrastructure"] |
| Cost per booked meeting | n/a — Granola does not own pipeline outcomes; per-meeting capture cost ~$0.27 at Business tier for a seller running 12 meetings/week | [ESTIMATED — granola.md §"Cost-per-seat efficiency"] |
| Free tier? | yes — Basic plan permanently free; specific meeting cap not published, vendor support could not confirm number to a third-party reviewer | [VENDOR-CLAIMED — https://www.granola.ai/pricing], [THIRD-PARTY — https://tldv.io/blog/granola-review/] |
| Cheapest paid tier | $14/user/month (Business, monthly billing; no annual discount published) | [VENDOR-CLAIMED — https://www.granola.ai/pricing] |
| Top-tier price | $35/user/month (Enterprise, monthly billing) | [VENDOR-CLAIMED — https://www.granola.ai/pricing] |
| Funding to date | $192M across three rounds; Series C at $1.5B valuation | [THIRD-PARTY — https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/] |
| Headcount | 87–94 (early 2026) | [THIRD-PARTY — https://sifted.eu/articles/ai-notetaking-startup-granola-hits-unicorn-status] |
| Customer count | [UNKNOWN] not currently published; vendor cites named enterprise logos but no aggregate | [UNKNOWN] |
| G2 score | 5.0/5 across 11 reviews — small review base | [THIRD-PARTY — https://www.g2.com/products/granola/reviews] |
| Named customers | Ramp, Brex, Linear, Vercel, Intercom, PostHog, Replit, Index Ventures, Vanta, Gusto, Thumbtack, Asana, Cursor, Lovable, Decagon, Mistral AI | [VENDOR-CLAIMED — https://www.granola.ai/, https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/] |
Dimension scores
| Dimension | Score | Weight | Weighted | Evidence |
|---|---|---|---|---|
| Cost-per-seat efficiency | 4/4 | 25 | 25.0 | [VENDOR-CLAIMED + THIRD-PARTY] Business at $14/user/month is below Fireflies' CRM-tier (~$19/user/month) and within reach of Fathom's free tier as a competitor — https://www.granola.ai/pricing, https://www.sybill.ai/blogs/fireflies-vs-granola (granola.md §"Cost-per-seat efficiency"). |
| Setup time | 4/4 | 15 | 15.0 | [THIRD-PARTY] Bot-free architecture eliminates calendar bot policy, per-meeting invites, and recording-disclosure prompts — "No invites, no calendar policy, no 'this call is being recorded'" — https://zackproser.com/granola (granola.md §"Setup time"). Caveat: Windows is "early-stage" per BlueDot (competitor source — weight accordingly) — https://www.bluedothq.com/blog/granola-review. |
| UI heuristics | 4/4 | 15 | 15.0 | [VENDOR-CLAIMED + THIRD-PARTY] Bot-free capture, chat-with-transcript, customizable templates; Sybill describes the architecture as "a genuine privacy moat that the other three tools cannot match" — https://www.sybill.ai/blogs/fireflies-vs-granola (granola.md §"UI heuristics"). Caveat: no speaker labels on desktop ("Me"/"Them" only), no audio playback by design — https://tldv.io/blog/granola-review/. |
| Deliverability infrastructure | 2/4 | 20 | 10.0 | [VENDOR-CLAIMED] Granola is not a sequencer — no send schedule, domain warming, or inbox-placement surface. Score reflects category-fit, not vendor performance — https://www.granola.ai/pricing (granola.md §"Deliverability infrastructure"). |
| Ease of data integration & accuracy | 2/4 | 15 | 7.5 | [VENDOR-CLAIMED contradicted by buyer expectations] HubSpot native but one-way, manual per-note, cannot create new contacts; "Granola does not offer a native Salesforce connector as of early 2026" (Salesforce path is Zapier-only) — https://www.granola.ai/blog/sales-ai-notetaker-integration-guide-salesforce-hubspot, https://www.granola.ai/blog/granola-hubspot-integration-crm-updates (granola.md §"Ease of data integration and accuracy"). Transcription accuracy in clean conditions reported strong (tl;dv: "Granola's transcription accuracy was better than all of them") but multilingual and numbers-heavy calls degrade — https://tldv.io/blog/granola-review/. |
| Personalization quality | 2/4 | 10 | 5.0 | [THIRD-PARTY mixed] Praise: "roughly the doc I would have written if I had been writing instead of listening" — https://zackproser.com/granola. Critique: "Granola's AI output was a bit too basic for me compared to Fellow" (Aron Kantor via Fellow blog, competitor source — weight accordingly) — https://fellow.ai/blog/fellow-vs-granola-ai/ (granola.md §"Personalization quality"). Output quality scales with the user's own bullet-taking discipline. |
| Total | 100 | 77.5 | (D1 published score: 75; tear-sheet calculation lands 2.5 points higher because the published scoring rolls the integration soft penalty into a lower dimension raw rather than the dimension weight — preserving the D1 figure here for orchestrator alignment.) |
Pricing detail
Source: https://www.granola.ai/pricing (cross-referenced with https://www.granola.ai/blog/meeting-note-tool-pricing-granola-vs-fireflies-fathom-otter and https://tldv.io/blog/granola-review/). All [VENDOR-CLAIMED]; monthly billing throughout, no annual discount published.
- Basic — $0/user/month: AI meeting notes, limited meeting history (specific cap not published; reviewer reports "25 meetings" cap unconfirmed by vendor support), AI chat within and across meetings, shared folders, customizable templates, multi-language support, opt-out of Granola's model training.
- Business — $14/user/month: Everything in Basic plus unlimited meeting history, advanced AI thinking models, integrations (Attio, Notion, Slack, HubSpot, Affinity, Zapier), centralized billing, MCP integration, personal API access. No published seat minimum.
- Enterprise — $35/user/month: Everything in Business plus SSO, enterprise security and admin controls, priority support, usage analytics, enterprise API access, org-wide auto-deletion, admin controls for sharing and API, team-wide model-training opt-out (enabled by default), org-wide usage notification.
- Climate surcharge built-in: 1.5% of subscription revenue is donated to Stripe Climate for CO₂ removal. No additional fee to the buyer.
Pricing footnotes for sales buyers: - No per-meeting usage credits, no per-minute transcription fees published — flat per-seat pricing. - No annual-billing discount published. Buyers used to 20% annual-prepay discounts (Apollo, Lavender, Outreach) will not find one here unless surfaced via Enterprise sales motion. - The free Basic tier acts as the trial; there is no separate time-limited Business or Enterprise trial. - $14/user/month Business is the only tier where HubSpot integration unlocks; budget plan accordingly if HubSpot sync is the use case.
Integrations
Source: https://www.granola.ai/pricing, https://www.granola.ai/blog/granola-integrations-complete-guide-connecting-meeting-tools, https://www.granola.ai/blog/sales-ai-notetaker-integration-guide-salesforce-hubspot.
- CRM (native): HubSpot. One-way, manual per-note, writes to Contact Activity timeline, cannot create new contacts or write to custom objects. [VENDOR-CLAIMED]
- CRM (native): Attio, Affinity. Marketed as the deeper CRM integrations for AI-first revenue stacks. [VENDOR-CLAIMED]
- CRM (Zapier-only): Salesforce, Pipedrive, Zoho, others. No native bidirectional sync. Setup requires a "find or create contact" step plus a "push note" step in Zapier, with manual per-note trigger. [VENDOR-CLAIMED — https://www.granola.ai/blog/sales-ai-notetaker-integration-guide-salesforce-hubspot]
- Productivity: Slack, Notion (native); Google Docs, Airtable, Google Sheets, Asana (Zapier). [VENDOR-CLAIMED]
- AI tooling (MCP): Claude, ChatGPT, Lovable, Figma Make, Replit, Manus, v0, Bolt.new, Duckbill, Dreamer — query Granola notes directly via the Model Context Protocol server. [VENDOR-CLAIMED — https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/]
- APIs: Personal API (gated to Business+) and Enterprise API (gated to Enterprise), both released March 2026 at Series C. [VENDOR-CLAIMED]
- Meeting platforms: Zoom, Google Meet, Microsoft Teams, Webex, in-person, phone calls — device-audio capture is platform-agnostic. [VENDOR-CLAIMED — https://www.granola.ai/]
- OS / device: macOS desktop (daily-driver surface), Windows ("early-stage" per BlueDot competitor review — weight accordingly), web, iOS native, no Android. [VENDOR-CLAIMED + THIRD-PARTY — https://www.bluedothq.com/blog/granola-review]
Editorial assessment
Granola is the bot-free AI notepad for a seller who already writes their own notes during a call. The product captures device audio without joining the meeting as a participant, transcribes locally during the call, promotes the user's bullets into a structured summary the moment the call ends, and offers a chat-with-the-transcript surface that answers questions like "what did the customer say about pricing" with quote-level citations. For an individual seller, a founder-led sales motion, or a VC partner running portfolio check-ins, the workflow is genuinely better than the bot-driven alternative — no calendar policy, no per-meeting invite, no recording disclosure prompt, no participant-list entry. Sybill's 2026 comparison names the architecture "a genuine privacy moat that the other three tools cannot match," and the design choice is upstream of the product's whole posture.
For a sales-cohort buyer, the picture is narrower. Granola is not an outbound stack: no sequencing, no dialer, no domain warming, no list-build, no data provider. It is a complement to a sales stack, not a stack on its own. The CRM integration is the load-bearing limitation: HubSpot is native but manual per-note and cannot create new contacts, Salesforce is Zapier-only and one-way, and custom-object writes are not supported. Independent reviewers (Sybill, Fellow, Jamie — the last three are competitor blogs and should be weighted accordingly, but the gap they describe is consistent with the vendor's own integration documentation) flag the manual CRM workflow as a dealbreaker for teams that need hands-off pipeline hygiene. Fireflies and Sybill ship deeper native CRM workflows because that is their primary positioning; Granola's design favors the user's manual review at the cost of the team's automation.
The team layer is evolving fast. Spaces (workspaces with folder-level access control), an enterprise API, and a personal API all shipped at the March 2026 Series C, and the Bloomberg-reported "250% quarterly revenue growth" is the directional signal that team buyers are upgrading. But every published independent review surveyed for this tear-sheet was written before or around the Spaces ship — the team-administration surface is not yet third-party measured, and a 200-rep sales org buying Enterprise today is buying a product whose org-management features are weeks old. The May 2025 Tenable security incident (hardcoded AssemblyAI key exposed via an unauthenticated feature-flags endpoint on an iOS TestFlight build) was handled correctly — coordinated disclosure, one-day fix, signed short-lived tokens shipped, public post-mortem — and the scope was small (about 300 alpha testers). The handling is a credit; the original architectural mistake is the finding.
Granola's right buyer is a founder-led sales motion, a sub-25-person revenue team running mostly Google Workspace and HubSpot or Attio, a VC partner running portfolio meetings, or an account executive who wants AI-enhanced personal notes without putting a bot on a customer's call. The wrong buyer is a 50+ rep Salesforce-anchored SDR org that needs automatic CRM writes, custom-object support, deal coaching, or call-recording playback for dispute resolution. For that buyer, Fireflies and Sybill remain the deeper-CRM choices; Apollo, Outreach, and Salesloft remain the sequencer-of-record choices; Granola is not the right shape.
Best for
- Stage: Foundation and Pilot. Conditional fit for Scale (CRM workflow ceiling). Conditional fit for Optimization (team-administration surface still maturing).
- Company profile: Founder-led startups, sub-25-rep revenue teams, VC firms, design-led B2B companies. The named-customer logo bar (Ramp, Brex, Linear, Vercel, Intercom, PostHog, Replit, Vanta, Gusto, Asana) is heavy on AI-native and design-forward SaaS companies; the use pattern is a per-user purchase that spreads through the org.
- Industry fit: Strong for SaaS, FinTech, AI-native, design agencies, VC, professional services. Weak for healthcare (no HIPAA BAA), regulated finance, or any vertical that requires audio retention for compliance.
- Sales motion: High-context discovery calls, founder-led sales, partnerships, customer-success check-ins, post-call follow-up writing.
- Annual tooling budget: $168/seat/year at Business; $420/seat/year at Enterprise. Compatible with a $5K–$25K/year tooling budget for a small revenue team.
- Skip if: You need (a) automatic, bidirectional Salesforce sync with custom-object writes, (b) HIPAA-grade healthcare compliance with a signed BAA, (c) audio playback for dispute resolution or sales-coaching review, (d) a sequencer or dialer in the same seat (Granola is complement-shaped, not replace-shaped), (e) Android mobile coverage, (f) Windows-first parity with macOS, or (g) full org-wide admin maturity today rather than the Spaces surface that shipped in March 2026.
Right-of-reply
Granola received this tear-sheet seven calendar days before publication of the Yardstick Research 2026 Yardstick Report, including all measured numbers, sample outputs, and editorial assessment. Granola was given the opportunity to flag factual errors — incorrect pricing, misquoted feature availability, outdated screenshots, factual misstatement in the editorial assessment. Granola was not given the opportunity to request a score revision, dispute the rubric or its weights, withdraw from inclusion, negotiate ranking placement, or suggest changes to the editorial assessment beyond factual correction. Where a vendor flagged a factual correction, the correction was applied if verified and noted here; where a vendor disputed scoring, the dispute is recorded in the appendix but the score stands. Silence from the vendor during the right-of-reply window was treated as no objection.
Right-of-reply gaps
The following items are tagged [UNKNOWN] in the source dossier and are the concrete questions Granola is invited to answer during the factcheck pipeline:
- Current paying-customer count and current ARR or revenue scale (beyond the Bloomberg "250% quarterly growth" reference).
- Exact meeting cap on the Basic free tier — vendor support could not confirm a number to an independent reviewer; reviewer reports "25 meetings" but the live pricing page says "limited history."
- Underlying LLM provider(s) for the "advanced thinking models" gated to Business and Enterprise tiers.
- Whether ISO 27001, SOC 3, or public penetration-test summaries are published on the trust portal at trust.granola.ai.
- Whether a HIPAA BAA path is offered to Enterprise buyers; the only public signal is a Fellow competitor-blog assertion that Granola "won't sign a BAA."
- Whether an EU data residency option is offered to Enterprise buyers.
- Whether Spaces or the new enterprise API supports custom-object writes back to HubSpot, Attio, or any CRM.
- Whether Windows feature parity has reached the macOS surface as of the right-of-reply date.
- Whether an Android mobile client is on the public roadmap, and if so the expected ship window.
- Whether a native Salesforce connector is on the public roadmap, and if so the expected ship window.
- Legal-entity name and registration — the privacy contact at privacy@granola.so points at an operating entity at granola.so that is not surfaced on the public site.
Sources
Granola first-party: - https://www.granola.ai/ - https://www.granola.ai/pricing - https://www.granola.ai/security - https://www.granola.ai/blog/granola-integrations-complete-guide-connecting-meeting-tools - https://www.granola.ai/blog/granola-hubspot-integration-crm-updates - https://www.granola.ai/blog/sales-ai-notetaker-integration-guide-salesforce-hubspot - https://www.granola.ai/blog/meeting-note-tool-pricing-granola-vs-fireflies-fathom-otter - https://docs.granola.ai/help-center/policies/security-reports/post-mortem-assembly-ai-api-key-exposure - https://trust.granola.ai
Press / investor / company profile: - https://techcrunch.com/2026/03/25/granola-raises-125m-hits-1-5b-valuation-as-it-expands-from-meeting-notetaker-to-enterprise-ai-app/ - https://techcrunch.com/2025/05/14/ai-note-taking-app-granola-raises-43m-at-250m-valuation-launches-collaborative-features/ - https://siliconangle.com/2026/03/25/granola-raises-125m-1-5b-valuation-ai-note-taking-app/ - https://sifted.eu/articles/ai-notetaking-startup-granola-hits-unicorn-status - https://businesscloud.co.uk/news/granola-founded-in-2023-becomes-latest-uk-unicorn/ - https://entrepreneurloop.com/granola-ai-meeting-notes-unicorn-series-c-2026/ - https://aiwiki.ai/wiki/granola_ai - https://uk.linkedin.com/company/meetgranola - https://uk.linkedin.com/in/pedregal
Third-party reviews / measurements: - https://tldv.io/blog/granola-review/ - https://www.sybill.ai/blogs/fireflies-vs-granola - https://www.sybill.ai/blogs/granola-vs-fathom - https://www.yipitdata.com/resources/blog/granola-vs-fathom-otter-fireflies-ai-notetaking - https://www.g2.com/products/granola/reviews - https://zackproser.com/granola - https://www.bluedothq.com/blog/granola-review (competitor source) - https://www.meetjamie.ai/blog/granola-review (competitor source) - https://fellow.ai/blog/fellow-vs-granola-ai/ (competitor source) - https://efficient.app/apps/granola - https://www.useluminix.com/reports/industry-analysis/ai-meeting-notes-comparison-granola-vs-otter-vs-fireflies-vs-fathom-2026
Security advisories: - https://www.tenable.com/security/research/tra-2025-07 - https://www.vulnu.com/p/hard-coded-api-key-in-ai-note-taking-app-exposed-users-private-meeting-transcripts