How Common Is a 'Restore' Link on Mobile Paywalls, and How Is It Worded?

Across 597 role-labeled secondary paywall CTAs from ~800 tracked apps, restore-style links account for 18.9% (113 CTAs) and appear across 43 companies — the most broadly adopted secondary affordance in the corpus [1]. The wording splits across 'restore' (32), 'restore purchases' (22), 'restore subscription' (21) and 'restore purchase' (18) [2]. No single restore phrasing dominates, but 'restore purchases' spans the most companies (19) [2].

Restore links make up 18.9% of 597 secondary paywall CTAs and appear across 43 companies — the widest-adopted secondary affordance in Lazyweb Research's corpus, July 2026.

Lazyweb Research · n=597 · Published 2026-07-07

paywallmonetizationmobileux-patternscancellation

The finding

Restore is the most company-broad secondary link: 113 of 597 secondary CTAs (18.9%) are restore-family, spread across 43 distinct companies [1]. That breadth beats the plan-comparison family (24 companies) and legal links (28 companies), even though plan-comparison ties restore on raw share (18.1%) [1].

How teams word it

There is no consensus string — the restore label fragments four ways [2]:

Restore wordingInstancesCompanies
restore3214
restore purchases2219
restore subscription212
restore purchase189

'restore subscription' looks common by instance count but comes from just 2 companies, so it is an app-level artifact rather than a broad pattern; 'restore purchases' (plural) is the most widely adopted across companies [2].

How to apply it

A restore link is effectively table stakes — Apple's guidelines require a restore path for subscriptions, and 43 companies in the corpus surface it right on the paywall. If you are choosing wording, 'Restore purchases' (plural, 19 companies) is the safest default. Keep it as a low-emphasis text link so it never competes with your primary CTA [1][2].

Caveats

This counts only visible on-paywall restore links among role-labeled secondary CTAs; apps that bury restore in settings are not captured, so 18.9% is a lower bound on how many products offer restore at all [3]. 39% of the 4,406-CTA corpus is role='unknown' and excluded [3].

The numbers

StatComputed from
restore family = 18.9% (113/597), 43 companies; view-plans family = 18.1% (108), 24 companies; legal links = 14.2% (85), 28 companiessecondary_families
restore=32 (14 cos); restore purchases=22 (19); restore subscription=21 (2); restore purchase=18 (9)secondary_leaderboard
597 secondary of 2,679 role-labeled; 1,727/4,406 (39%) role='unknown'role_distribution / smallSampleWarnings
Methodology. Universe: 113 restore-family secondary CTAs among 597 labeled secondary CTAs (of 4,406 total) from ~800 tracked apps, Supabase pull July 2026. Text lowercased/trimmed; restore family matched on the 'restore%' prefix. Caveat: only on-paywall restore links are counted, so prevalence is a lower bound.

Sources & citations

  1. [1] Lazyweb Research analysis of 597 secondary paywall CTAs (~800 tracked apps), July 2026. Secondary CTA family shares with distinct-company counts.
  2. [2] Lazyweb Research analysis of 113 restore-family secondary CTAs (~800 tracked apps), July 2026. Restore wording variants; instance and distinct-company counts.
  3. [3] Lazyweb Research analysis of 4,406 paywall CTAs (~800 tracked apps), July 2026. 39% role='unknown'; restore prevalence is a lower bound because settings-only restore links are not captured.

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