What do click-to-cancel flows look like screen by screen in the field?
- By Ali Abouelatta
- n = 48
- Published
- Updated
- 2 min read
- cancellation · retention · ux-patterns · checkout · saas
Short answer
Across 48 apps with genuine cancel-intent screens, the observed building blocks are a confirmation step (12 apps, 25%), a keep/stay retention CTA (10 apps, 21%), a discount offer (30 apps, 63%), and — rarely — a reason survey (2 apps, 4%) source[1]source[2]source[3]number[4]. A common named pattern (Instacart+) stacks loss-aversion warning, reason survey, and a 'remind me later' deferral before the primary cancel button number[5].
Confirmation step in 25% of apps, keep/stay CTA in 21%, discount in 63% of 48 cancel-flow apps, July 2026.
Bar chart · Share
Share — The observed screen components
One block is about 2.3%
- Discount / offer
- 63% [1]
- Confirm / 'are you sure' step
- 25% [2]
- Keep / stay retention CTA
- 21% [3]
- Reason survey
- 4% [4]
The observed screen components
Discount is the most common element; a confirmation step and a keep/stay CTA appear in roughly a quarter and a fifth of apps respectively source[1]source[2]source[3].
A representative sequence
Instacart+'s cancellation confirmation combines three tactics on one screen: a loss-aversion warning ('you'll lose benefits on [date]'), a reason survey (radio options plus free text), and a 'remind me later' deferral, before the primary 'Yes, cancel membership' number[5]. DoorDash shows a survey-first variant: a single-select reason list gates progression number[6]. These illustrate how confirmation, survey, and retention CTA compose into the ~3-screen median depth.
How to apply it (and the compliance angle)
The cancel action itself is typically de-emphasized — across cancel/keep/pause/resubscribe CTAs, 40 are styled dismiss and 18 secondary versus only 7 primary number[7]. If you are aiming for compliance, keep the actual cancel affordance reachable rather than buried in dismiss styling. DoorDash's own A/B made the reason button easier to progress (grayed-out → solid), a completion-friendly direction number[8]. Lead with observed prevalence, not compliance advice — but note the field trend is toward easing, not obstructing, the cancel path.
Real in-market screens
Screens from real, in-market apps, tracked by Lazyweb Research.
Chaton1179 × 2556opens the full image 
Rootd1206 × 2622opens the full image 
Youtube Music1206 × 2622opens the full image 
Copilot1170 × 2532opens the full image
The numbers
| # | Stat | Computed from |
|---|---|---|
| [1] | 30 of 48 apps (63%) show discount/offer language on cancel screens | save_offer_offer_language |
| [2] | 12 of 48 apps (25%) show a confirm/'are you sure' step | confirmation_dialog |
| [3] | 10 of 48 apps (21%) show a keep/stay retention CTA | keep_stay_cta |
| [4] | 2 of 48 apps (4%) show a reason survey | cancellation_survey_prevalence |
| [5] | Instacart+ stacks loss-aversion warning, reason survey, 'remind me later' deferral | qualitative / instacart cancel-flow-screen |
| [6] | DoorDash single-select reason list gates progression | qualitative / doordash cancel-flow-screen |
| [7] | Cancel/keep/pause/resubscribe CTA roles: 40 dismiss, 18 secondary, 7 primary | cancel_flow_cta_roles |
| [8] | DoorDash A/B eased reason button from grayed-out to solid red | qualitative / doordash experiment |
Methodology
Universe: 48 apps with genuine cancel-intent screens (401 screenshots) plus 78 cancel/keep-style CTAs, July 2026. Method: company-level keyword presence and CTA role counts, with named-example descriptions. Caveat: components come from captured vision text and may undercount elements not surfaced in screenshots.
Sources
- [1]
Lazyweb Research analysis of 48 apps (genuine cancel-intent screens), July 2026
Company-level presence of confirmation, keep/stay, discount, and survey elements.
- [2]
Lazyweb Research analysis of 401 cancel-intent screenshots (48 apps), July 2026
Instacart+ and DoorDash named cancel-flow vision descriptions.
- [3]
Lazyweb Research analysis of 78 cancel/keep/pause CTA buttons (paywall universe), July 2026
CTA role split showing cancel actions skew dismiss/secondary.
Source: Lazyweb Research — proprietary analysis of real, in-market app screens. Cite as Lazyweb Research, 2026-07-07.