Skip to content

'Continue' vs 'Start Free Trial': Which Paywall CTA Do Top Apps Actually Use?

  • By Ali Abouelatta
  • n = 1,886
  • Published
  • Updated
  • 2 min read
  • paywall · trials · monetization · ux-patterns · experiments

Short answer

Across 222 companies with a labeled primary paywall CTA, 103 (46.4%) lead with a trial-promise button (try/start + free/trial) versus 69 (31.1%) that lead with a 'Continue'-style button source[1]. So more apps choose a trial CTA at the company level — but 'Continue' is still the most-repeated single string (350 CTAs) source[2]. The two strategies split the market roughly in half; neither is a clear majority.

The quotable stat
Trial-led paywall CTAs are used by 103 of 222 companies (46.4%) vs 69 (31.1%) leading with 'Continue' — Lazyweb Research, 1,886 primary CTAs, July 2026.

Bar chart · Share of primary

Share of primary — Two philosophies, side by side

One block is about 1%

Trial-led (try/start + free/trial)
27.9%
Continue-led (any 'continue…')
21.1%
Exact bare 'continue'
18.6%

The head-to-head

Grouping by intent across the 1,886 primary CTAs: the trial-led family (buttons starting with try/start and mentioning free or trial) is 527 CTAs from 103 companies, while the continue-led family (any 'continue…' variant) is 397 CTAs from 69 companies source[1]number[3]. On CTA count, trial-led (27.9% of primary CTAs) beats continue-led (21.1%); on company adoption, 46.4% vs 31.1% source[1]number[3]. Trial wins both, but not by a landslide.

Two philosophies, side by side

StrategyPrimary CTAsShare of primaryCompaniesCompany share
Trial-led (try/start + free/trial)52727.9%10346.4%
Continue-led (any 'continue…')39721.1%6931.1%
Exact bare 'continue'35018.6%6629.7%

'Continue' defers the ask: the money and plan choice live in the cards, the button just advances. 'Start Free Trial' front-loads the offer in the button itself. Both patterns are mainstream source[1]source[2]number[3].

How to choose

If your plan cards already make the trial and price obvious, 'Continue' keeps the button friction-free — its 29.7% company adoption shows it reads as a safe, low-anxiety default source[2]. If the trial is your strongest lever and cards are dense, put the promise in the button ('Start Free Trial', 84 CTAs / 22 companies; 'Try for Free', 74 / 16) number[4]. This benchmark tells you what is common, not what converts for your funnel — treat it as a prior for an A/B test, not a verdict.

Caveats

Denominator is the 222 companies (or 1,886 primary CTAs) with a labeled primary CTA; 39% of all 4,406 CTAs are role-unknown and excluded number[5]. Family membership is regex-based on lowercased text, so borderline phrasings ('try premium', 'start now' without 'free') land outside both families and are not double-counted.

Real in-market screens

Screens from real, in-market apps, tracked by Lazyweb Research.

  • Apple Fitness Paywall / Subscription screen — real in-market example tracked by Lazyweb Research
    Apple Fitness1179 × 2556opens the full image
  • Reddit paywall screen — real in-market example tracked by Lazyweb Research
    Reddit1125 × 2436opens the full image
  • Shein Checkout upsell screen — real in-market example tracked by Lazyweb Research
    Shein1179 × 2556opens the full image
  • Discord All plans screen — real in-market example tracked by Lazyweb Research
    Discord1170 × 2532opens the full image

The numbers

#StatComputed from
[1]trial-led 527 CTAs (27.9%), 103 companies (46.4% of 222)statpack trial_led_primary + continue_vs_trial_company_level
[2]exact bare 'continue' = 350 CTAs (18.6% of 1886), 66 companies (29.7%)statpack top_primary_cta_overall + exact_continue_companies
[3]continue-led 397 CTAs (21.1%), 69 companies (31.1%)statpack continue_led_primary + continue_vs_trial_company_level
[4]start free trial 84 CTAs/22 cos; try for free 74/16statpack primary_leaderboard_top10
[5]1,727 of 4,406 CTAs role-unknown (39%)statpack role_distribution

Methodology

Universe: 1,886 role-labeled primary paywall CTAs from ~800 tracked mobile apps, July 2026. Trial-led = text matching ^(try|start).*(free|trial); continue-led = text LIKE 'continue%'. Caveat: 39% of CTAs are role-unknown and excluded from role stats.

Sources

  1. [1]

    Lazyweb Research analysis of 1,886 primary paywall CTAs (from ~800 tracked mobile apps), July 2026

    Intent families defined by regex on lower(trim(cta_text)); company counts via COUNT(DISTINCT).

    Back to text

  2. [2]

    Lazyweb Research analysis of 222 companies with a labeled primary paywall CTA, July 2026

    Company-level adoption of continue-led vs trial-led primary CTAs.

    Back to text

Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.

Explore the evidence