Our panel saw Granola only in direct Stripe receipts; rivals appeared through more billing channels
Distinct U.S. panelists with a paid receipt from June 2024–August 17, 2026; billing-channel coverage differs by product
Observable paid users
Source: YipitData U.S. e-receipt panel. Granola n=260; Otter n=3,279; Fathom n=5,815; Fireflies n=7,929. Granola is observable only through receipts paid directly through Stripe, while rivals appear through multiple consumer-visible channels.
Chart data
| Series | Observable paid users |
|---|---|
| Fireflies | 7,929 |
| Fathom | 5,815 |
| Otter | 3,279 |
| Granola — paid through Stripe only | 260 |
How often did observable Granola payers also pay an AI-notetaking rival?
Forty of 260 Granola payers, or 15.4%, also had a receipt from Otter, Fathom, or Fireflies. Thirty paid Otter and 14 paid Fathom; some paid both. No Fireflies receipt appeared in the small Granola cohort. These counts are estimates and do not establish which subscription came first or whether the products were used together.
Do the payer counts show that Granola has fewer customers?
No. Granola is identifiable only in receipts paid directly through Stripe, and the panel underrepresents work inboxes and enterprise invoicing. App-store and enterprise-billed Granola seats are not isolated here. The counts compare only payments visible in consumer email receipts, not total subscribers, market share, or demand.
Assumptions & Methodologies
We analyzed YipitData’s U.S. e-receipt panel from Granola’s first observed receipt in June 2024 through August 17, 2026. Granola required a Stripe email with the subject ‘your receipt from Granola’; food and grocery matches were excluded. Otter, Fathom, and Fireflies used validated receipt-text matches. Granola line items were masked, and its app-store, enterprise-invoiced, and work-inbox billing is undercaptured. Rival billing channels also differ. Counts are observable receipt footprints rather than audited subscribers or market share. The 14–30-user overlap cells are small, and co-payment does not establish sequence, switching, use, or causality.