# What Percent Of Paywalls Have A Visible Dismiss Option?

Source: Lazyweb Research
Published: 2026-07-07
Sample size: n=2708
Tags: paywall, monetization, ux-patterns, mobile, design
HTML: https://www.lazyweb.com/research/what-percent-of-paywalls-have-a-visible-dismiss-option
Markdown: https://www.lazyweb.com/research/what-percent-of-paywalls-have-a-visible-dismiss-option.md

**Answer.** Across 2,708 mobile paywall screens tracked by Lazyweb Research, 710 (26.2%) show a visible exit affordance — an X/close control, a 'Maybe later'-style link, or a dismiss-role button [1]. The remaining 1,998 screens (73.8%) show no visible exit in extracted data, a hard-paywall proxy [1]. Because 39% of CTA rows are unlabeled and vision descriptions can miss close controls, this is a lower bound on how many paywalls can actually be dismissed [1].

> Only 26.2% of 2,708 mobile paywall screens (710) show a visible dismiss affordance — Lazyweb Research, July 2026.

## The finding

Of 2,708 distinct paywall screens with extracted CTAs, 710 (26.2%) carry at least one visible way out — a dismiss-role CTA, dismissal copy such as 'not now' or 'maybe later', an X/close text, or a close control described in the screen's vision description [1]. The other 1,998 screens (73.8%) expose no visible exit in the extracted data [1]. In other words, roughly three in four captured paywalls read as hard gates.

## Why this is a lower bound

Two things pull the true dismissal rate above 26.2% [1]. First, 39% of the 4,406 CTA rows carry role='unknown', so some genuine dismiss buttons are simply unlabeled [2]. Second, OS-level back gestures and unlabeled close controls are invisible to extraction. Treat 26.2% as the floor on dismissible paywalls, not a point estimate.

## How to apply it

If you are deciding whether to ship a hard or soft paywall, know that you are in the minority (26.2%) if you add a visible exit [1] — but that most competitors run at least one soft variant somewhere (see companies-level cut). Use the hard-default as context for A/B design, not as proof that hard converts better: this corpus records what shipped, never measured lift [1].

## The numbers

| Stat | Computed from |
| --- | --- |
| 710 of 2,708 screens (26.2%); 1,998 (73.8%) no visible exit | screens_with_any_visible_exit: 710/2708 = 26.2% |
| 39% of 4,406 CTA rows are role='unknown' | cta_role_distribution: 1,727 unknown of 4,406 |

## Methodology

Universe is 2,708 distinct paywall screens with extracted CTAs across 252 companies in the ~800-tracked-app mobile corpus. A screen counts as having a visible exit if extraction found a dismiss-role CTA, dismissal copy, X/close text, or a close control in its vision description; all figures are lower bounds because 39% of CTAs are unlabeled.

## Sources & citations

- [1] Lazyweb Research analysis of 2,708 paywall screens (mobile app corpus, 252 companies), July 2026. Visible-exit detector combines dismiss-role CTAs, dismissal-copy regex, and vision-description close-control mentions; all prevalence figures are lower bounds.
- [2] Lazyweb Research analysis of 4,406 paywall CTA rows (mobile app corpus, 252 companies), July 2026. Role distribution across all CTA rows; 2,679 rows are role-labeled.

## Related questions

- [Hard Vs Soft Paywall: How Common Is Each?](https://www.lazyweb.com/research/hard-vs-soft-paywall-how-common-is-each)
- [How Do Apps Let Users Exit The Paywall — X Button, 'Maybe Later', Or Not At All?](https://www.lazyweb.com/research/how-do-apps-let-users-exit-the-paywall)
- [What Share Of Companies Run A Soft Paywall Variant?](https://www.lazyweb.com/research/what-share-of-companies-run-a-soft-paywall-variant)