What CTA Copy Do Hard Paywalls Use Most?
On no-exit (hard) paywall screens tracked by Lazyweb Research, the top primary CTA is 'continue' (276 of 1,342 primary CTAs), followed by 'subscribe' (65), 'start free trial' (58), 'try for free' (49) and 'subscribe now' (42) [1]. 'Continue' is about 3.8x more common on hard paywalls than on soft ones in absolute terms (276 vs 73) [1]. Hard paywalls lean on a neutral next-step verb, not the offer [2].
'Continue' is the top hard-paywall CTA at 276 of 1,342 primary CTAs — 3.8x its count on soft paywalls — Lazyweb Research, July 2026.
Top hard-paywall CTAs
| Primary CTA (no-exit) | Count |
|---|---|
| continue | 276 |
| subscribe | 65 |
| start free trial | 58 |
| try for free | 49 |
| subscribe now | 42 |
Counts are primary-role CTA texts on no-exit screens with frequency >= 8 [1]. 'Continue' dwarfs every other option.
Contrast with soft paywalls
On exit-visible screens the leaders reorder toward the offer: 'continue' (73), 'try for free' (25), 'start free trial' (25), 'subscribe' (24), 'try 3 days free' (19) [1]. Aggregating copy types, only 14.8% of soft primary CTAs are neutral 'Continue/Next' versus 24.7% of hard ones, while free/trial copy is 38.4% soft vs 31.7% hard [2]. Hard paywalls minimize commitment language; soft paywalls sell the trial [2].
How to apply it and caveats
If your paywall is a hard gate, 'Continue' is the corpus-default primary verb — it reframes a forced subscription as a step forward rather than a purchase [1][2]. If you soften the gate, expect to swap toward explicit trial copy like 'try for free' to keep conversion up [2]. Caveats: these are frequency counts on what shipped, pattern-matched, and never tied to measured lift [1][2].
The numbers
| Stat | Computed from |
|---|---|
| no-exit: continue 276, subscribe 65, start free trial 58, try for free 49, subscribe now 42; exit continue 73 | top_primary_cta_no_exit |
| neutral 'Continue' 24.7% no-exit vs 14.8% exit; free/trial 31.7% no-exit vs 38.4% exit | primary_cta_copy_by_exit |
Sources & citations
- [1] Lazyweb Research analysis of 1,342 primary paywall CTAs on hard (no-exit) screens (mobile app corpus), July 2026. Top primary CTA texts on no-exit screens (frequency >= 8); frequency counts, not lift. ↩
- [2] Lazyweb Research analysis of 1,884 primary paywall CTAs (mobile app corpus), July 2026. Copy-type shares split by exit presence; association not causal lift. ↩
Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.