What paywall CTA copy do market leaders prefer?

Neutral 'Continue' framing. 21.4% of leaders' primary paywall CTAs are the plain word 'Continue' (138 of 645), versus 11.5% for challengers (78 of 678) — leaders lean on neutral copy nearly twice as often.[1] They are also less trial-led: 35.2% of leader primary CTAs mention free/trial versus 43.4% for challengers.[2] The pattern is consistent across 645 and 678 primary CTAs, so it is well-powered.[1][2]

21.4% of market leaders' primary paywall CTAs are plain 'Continue' vs 11.5% for challengers (645 vs 678 CTAs) — Lazyweb Research, July 2026.

Lazyweb Research · n=1323 · Published 2026-07-07

paywallmonetizationcheckoutpricingtrialsupsell

The finding

Leaders and challengers diverge on button copy. On the primary paywall CTA:

Copy styleLeaders (n=645)Challengers (n=678)
Plain 'Continue'21.4% (138)11.5% (78)
Free / trial-led35.2%43.4%
Price in the button22.9%25.2%

Leaders lean neutral: 'Continue' is nearly 2× as common in their primary CTAs, while explicit trial framing and price-in-button are both slightly lower than challengers.[1][2]

The leader CTA vocabulary

Among leader primary CTAs appearing at least 4 times, the top texts are:

CTA textCount
continue138
start free trial61
start trial20
subscribe18
try free & subscribe17
try gold for free today13
get youtube premium12

'Continue' dominates, with explicit trial CTAs as the clear second tier.[3] Leaders still run trial copy heavily — they just don't default to it as reflexively as challengers do.

How to apply it

If your primary paywall button is always 'Start Free Trial', you're following the challenger majority. Leaders more often reduce the button to a neutral, low-friction 'Continue' and let the plan card, price, and trial terms carry the commitment framing above the button. That works best when a strong brand and clear plan details make the neutral verb feel safe rather than vague — test 'Continue' against your trial-led control rather than swapping blindly.

Caveats

These are correlations with market leadership, not proven conversion lifts — a neutral 'Continue' likely leans on brand trust a challenger may not have. Shares are computed on primary-role CTAs only. The direction is well-powered (645 vs 678 CTAs), but treat 'Continue' as a leader tendency to A/B test, not a guaranteed win.[1]

The numbers

StatComputed from
Leaders: 138 of 645 primary CTAs are 'continue' (21.4%); challengers 78 of 678 (11.5%)cta_copy_continue_share_leaders_vs_rest
Free/trial-led primary CTAs: leaders 35.2% (n=645) vs challengers 43.4% (n=678)cta_free_trial_and_price_share_leaders_vs_rest
Price-in-CTA: leaders 22.9% vs challengers 25.2%cta_free_trial_and_price_share_leaders_vs_rest
Top leader primary CTAs: continue 138, start free trial 61, start trial 20, subscribe 18, try free & subscribe 17, try gold for free today 13, get youtube premium 12leader_cta_copy_pattern
Methodology. Universe: Lazyweb ~800-app mobile corpus. Method: classified primary paywall CTA text (cta_role='primary') by copy style, split by leader flag, July 2026. Caveat: shares reflect the labeled primary-CTA subset and describe correlation with leadership, not measured conversion lift.

Sources & citations

  1. [1] Lazyweb Research analysis of 1,323 primary paywall CTAs (leaders vs challengers), July 2026. 645 leader primary CTAs vs 678 challenger primary CTAs; 'continue' share computed on cta_role='primary'. Joined to companies.market_leader on lower(company_name).
  2. [2] Lazyweb Research analysis of 1,323 primary paywall CTAs (leaders vs challengers), July 2026. Free/trial share = CTA text ILIKE '%free%' or '%trial%'; price-in-CTA = contains a currency symbol or digit; cta_role='primary'.
  3. [3] Lazyweb Research analysis of leaders' primary paywall CTA texts (n>=4), July 2026. Most common exact primary CTA strings among market_leader='Yes' apps, appearing at least 4 times.

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

Related questions

Explore the underlying screens, flows, and A/B tests inside Lazyweb. More research