Which Apps Test Their Paywall Offer But Rarely Change The CTA Text?
- By Ali Abouelatta
- n = 795
- Published
- Updated
- 2 min read
- paywall · checkout · monetization · experiments · pricing · ux-patterns
Short answer
Across 795 detected paywall-CTA experiments (146 companies, July 2026), several high-volume apps rarely touch the button copy: Impulse changed the CTA text in just 1 of 14 experiments, Headway in 3 of 19, and Apple Music in 5 of 19. source[1] These apps iterate the offer and layout while holding the CTA stable — the opposite of YouTube Music's 17-of-21 churn. source[1] These are observed variations, not confirmed A/B tests.
Impulse changed its paywall CTA text in only 1 of 14 detected experiments (July 2026) — the lowest CTA-copy churn among high-volume experimenters.
Bar chart · Text-change rate
Text-change rate — The low-churn cluster
One block is about 1%
- Impulse
- 7% [1]
- Headway
- 16% [1]
- JustFit
- 27% [1]
- Peloton
- 25% [1]
- Apple Music
- 26% [1]
The finding
Across 795 detected paywall-CTA experiments (146 companies), a distinct cluster of apps runs many CTA experiments but almost never rewrites the button text. source[1] Impulse changed it in 1 of 14, Headway in 3 of 19, JustFit in 3 of 11, and Apple Music in 5 of 19. source[1] These apps treat the CTA copy as fixed and iterate everything around it.
The low-churn cluster
For contrast, YouTube Music changed the text in 17 of 21 (81%) and Memrise in 12 of 16 (75%). source[1]
How to apply it
If your paywall CTA copy already performs, this cluster validates leaving it alone and iterating the offer, pricing display, and layout instead. Impulse and Headway are the clearest examples of high experiment volume with a frozen button. The takeaway for test planning: CTA text is one slot, not the only one — apps split roughly into text-churners and offer-churners, and both patterns are well-represented. All figures are detected variations, not proven winners. source[1]
Caveats
All figures are observed CTA variations with LLM-inferred rationale, not company-confirmed A/B tests — no lift is measured. source[1] Only the 30 companies with 8 or more CTA experiments support these rates; the other 116 have too few. source[1]
The numbers
| # | Stat | Computed from |
|---|---|---|
| [1] | Low-churn: Impulse 14/1, Headway 19/3, JustFit 11/3, Peloton 12/3, Apple Music 19/5; high-churn YouTube Music 21/17, Memrise 16/12 | paywall_cta_by_company (impulse 14/1, headway 19/3, justfit 11/3, peloton 12/3, apple-music 19/5, youtube-music 21/17, memrise 16/12) |
Methodology
Universe: 795 detected paywall-CTA experiments across 146 companies, July 2026; analysis limited to the 30 with >=8 experiments. Extraction: total vs. cta_changed=TRUE. Caveat: detected variations only, no lift measured.
Sources
- [1]
Lazyweb Research analysis of 795 paywall CTA experiments (146 companies, ~800-app mobile corpus), July 2026
Per-company total vs. cta_changed=TRUE; only 30 companies have >=8 experiments.
Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.