Bon Appétit and Epicurious subscribers had the highest NYT All Access overlap
Share with a paid Cooking-inclusive NYT All Access receipt, January 2023–August 2026
Share with NYT All Access receipt
Source: YipitData U.S. e-receipt panel. Bon Appétit or Epicurious n=507; other recipe apps n=6,198; general digital subscribers n=329,579. The primary cohort was Apple-billed.
Chart data
| Series | Share with NYT All Access receipt |
|---|---|
| Bon Appétit or Epicurious | 21.1% |
| Other paid recipe apps | 8.5% |
| General digital subscribers | 3.2% |
Which subscription usually appeared first?
Among 107 dual subscribers, the first observed paid receipt was NYT for 48 users and Bon Appétit or Epicurious for 56. The near-even ordering does not support a claim that one service systematically acquired the customer from the other.
Which service remained when only one had a recent receipt?
Twenty-three dual subscribers had only a recent Bon Appétit or Epicurious receipt, while 15 had only a recent NYT receipt; 19 had both and 50 had neither. The single-service groups are below 30 users, so this is directional and does not establish which product users considered primary.
Does NYT All Access prove that subscribers used NYT Cooking?
No. The receipt language identifies a bundle that includes Cooking access, not confirmed recipe usage. Direct or web-billed Condé Nast subscriptions were also outside the measured cohort, so the 21.1% rate describes the Apple-billed segment rather than the full Bon Appétit and Epicurious subscriber base.
Assumptions & Methodologies
We analyzed YipitData’s U.S. e-receipt panel from January 2023 through August 2026. The primary cohort included 507 Apple-billed subscriptions explicitly naming Bon Appétit or Epicurious; generic uses of “bon appétit” were excluded. NYT overlap required a paid receipt with All Access or Cooking-inclusive product language. Benchmarks were 6,198 other paid recipe-app subscribers and 329,579 general digital subscribers. Direct Condé Nast billing was not captured, bundle access does not prove Cooking use, and the 107-person overlap yields directional sequence subgroups.