What percent of paywall primary CTAs lead with a free trial?
- By Ali Abouelatta
- n = 1,886
- Published
- Updated
- 2 min read
- paywall · trials · monetization · saas · ux-patterns
Short answer
33.7% of 1,886 primary paywall CTAs (635 buttons) mention 'free' or 'trial' in the button text itself.source[1] Trial-led wording is the single most common framing lever after the generic 'Continue' verb, but two-thirds of primary buttons say something else. Treat a trial-led button as the default worth beating in an experiment, not a guaranteed win.
33.7% of 1,886 primary paywall CTAs (635 buttons) mention 'free' or 'trial' in the button text — Lazyweb Research, July 2026.
The finding
Across 1,886 primary paywall CTAs extracted from ~800 tracked apps, 635 (33.7%) put 'free' or 'trial' directly in the button text.source[1] The pattern is a case-insensitive match on the CTA text for %free% or %trial%. This is a button-copy measurement — it counts what the tappable primary button says, not whether a trial exists anywhere on the screen.
For context, screen-level trial language is broader: 752 of 2,708 paywall screenshots (27.8%) carry free/trial wording somewhere in the extracted CTA-adjacent copy, and 131 of 252 companies use it at least once.source[2] So trial framing is common at the company level even when the button itself stays neutral.
How to apply it
| Framing | Share of primary CTAs | What it signals |
|---|---|---|
| Trial/free in the button | 33.7% (635/1,886) | Removes cost objection at the tap |
| Everything else | 66.3% (1,251/1,886) | Neutral verb, plan-led, or price-led |
If your button already leads with a trial, you are aligned with the largest single copy pattern — but you are also indistinguishable from a third of the market. The decision is whether the trial belongs in the button (33.7% do this) or above it in the offer copy while the button stays a neutral verb like 'Continue'. Both are common; test which converts for your price point.source[1]
Caveats
Role-based cuts use only the 1,886 CTAs labeled cta_role='primary'; 39% of all extracted rows are cta_role='unknown' and are excluded.source[1] The 'free/trial' match is a keyword regex over button text only, so a paywall that offers a trial but labels its button 'Continue' or 'Subscribe' is not counted here — the screen-level figure (27.8%) captures those.source[2] All figures are from a July 2026 pull.
Real in-market screens
Screens from real, in-market apps, tracked by Lazyweb Research.
Nyt Games1206 × 2622opens the full image 
Calm1170 × 2532opens the full image 
Replit1206 × 2622opens the full image 
Justfit1206 × 2622opens the full image
The numbers
| # | Stat | Computed from |
|---|---|---|
| [1] | 33.7% (635/1,886) | trial_led_primary_share: primary CTAs with cta_text ILIKE '%free%' OR '%trial%' |
| [2] | 27.8% screens (752/2,708); 131/252 companies | trial_mention_screens |
Methodology
Universe: 1,886 primary and 2,708 screen-level paywall CTAs across ~800 tracked apps, extracted July 2026 via regex over button text plus CTA-adjacent copy. Caveat: keyword matching on extracted copy, not full-screen OCR; 39% unknown-role rows excluded from primary cuts.
Sources
- [1]
Lazyweb Research analysis of 1,886 primary paywall CTAs (~800 tracked apps), July 2026
Regex on cta_text for %free%/%trial% among cta_role='primary' rows; 39% unknown-role rows excluded.
- [2]
Lazyweb Research analysis of 2,708 paywall screenshots (~800 tracked apps), July 2026
Screen-level bool_or over CTA-adjacent copy for free/trial language; company-level dedupe on lower(company_name).
Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.