Retained Lovable subscribers were more likely to hold another paid AI subscription
Share with a paid Claude, ChatGPT, Perplexity, Cursor, or Replit receipt in May–August 2026, among Lovable subscribers formed September 2025–January 2026
Share holding another paid AI subscription
Source: YipitData U.S. e-receipt panel. Subscribers without a May–August 2026 Lovable receipt n=1,552; subscribers with one n=168. The smaller later-receipt cohort makes its exact rate less precise.
Chart data
| Series | Share holding another paid AI subscription |
|---|---|
| Retained Lovable subscribers | 56.5% |
| Lovable lapsers | 34.1% |
How often did Lovable lapsers pay Cursor or Replit?
Only 2.6% of Lovable lapsers had a Cursor or Replit receipt in May–August 2026, compared with 7.7% of retained Lovable subscribers. Roughly two-thirds of lapsers had no observed paid AI receipt outside Lovable. The tracked set does not include every coding tool, so this is not proof that every lapser left paid AI.
Which paid AI subscription had the largest gap between lapsers and retained users?
Claude appeared among 29.4% of Lovable lapsers and 49.4% of retained subscribers, a 20.0-percentage-point gap. The broader foundation-model set of Claude, ChatGPT, and Perplexity reached 33.1% of lapsers and 54.8% of retained users. These rates describe co-payment, not which tool a customer used more.
Does the comparison prove that paid AI subscriptions improve Lovable retention?
No. The comparison shows that retained Lovable subscribers were more likely to hold another paid AI subscription, but it does not establish that the second subscription caused retention. People who use AI more intensively may be more likely to keep several tools. Credit-based or annual billing can also make an irregular Lovable buyer appear lapsed.
Assumptions & Methodologies
We analyzed 1,720 established U.S. e-receipt panelists with a Lovable receipt from September 2025–January 2026 and continued panel activity through August 2026. Retained users had a Lovable receipt from May–August 2026; lapsers did not. Other paid AI covered direct receipts from Anthropic, OpenAI, Perplexity, Cursor, and Replit during the same status window. A tighter September–November 2025 formation cohort reproduced the direction. Receipt capture can miss other tools and app-store billing, while credit or annual cadence can misclassify irregular buyers. Results describe observed co-payment and do not establish causality or reacquisition potential.