# What percent of cancel flows show a save offer, and which type?

Source: Lazyweb Research
Author: Ali Abouelatta, Lazyweb Research
Published: 2026-07-07
Updated: July 2026
Sample size: n=48
Category: [Cancellation flows and save offers](https://design.lazyweb.com/blog/category/cancellation-flows-and-save-offers)
Tags: cancellation, retention, monetization, upsell, saas
HTML: https://design.lazyweb.com/blog/what-percent-of-cancel-flows-show-a-save-offer
Markdown: https://design.lazyweb.com/blog/what-percent-of-cancel-flows-show-a-save-offer.md

**Answer.** Among 48 apps with genuine cancel-intent screens, 30 (63%) surface discount or offer language on a cancel screen [1]. Pause offers (0 of 48) and downgrade offers (0 of 48) are rarely observed in captured cancel screens — discount dominates by a wide margin [2][3]. These are absence-in-vision-text signals, so treat the pause/downgrade zeros as 'rarely observed,' not 'never used.'

> 30 of 48 apps (63%) show discount/offer language on cancel screens; pause and downgrade near-zero, July 2026.

## The breakdown by offer type

| Save-offer type | Apps | Share of 48 |
| --- | --- | --- |
| Discount / offer language | 30 | 63% [1] |
| Pause | 0 | 0% [2] |
| Downgrade / cheaper plan | 0 | 0% [3] |

Discount is overwhelmingly the observed save tactic. Pause and downgrade are near-absent in captured cancel screens [2][3].

## How to read the zeros

The pause (0/48) and downgrade (0/48) figures reflect **absence in captured vision text**, not proof the product never offers them [2][3]. State them as 'rarely observed on cancel screens' rather than 'never used.' The 48-app denominator is the genuine-cancel-intent subset (screens whose text literally mentions 'cancel'), used because the raw category is polluted with paywall and win-back screenshots [4].

## How to apply it

If you are choosing one save offer to build first, a **discount** is what the field overwhelmingly ships — it is by far the most-observed tactic [1]. Pause and downgrade are open space in this corpus [2][3]: rarer, so potentially differentiating, but validate demand before investing. Note the field also runs discount-plus-loss-aversion win-back plays (e.g. Headway's 33%-off 'trial expired' modal) adjacent to the in-flow cancel screen [5].

## The numbers

| Stat | Computed from |
| --- | --- |
| 30 of 48 apps (63%) with discount/offer language on cancel screens | save_offer_offer_language |
| 0 of 48 apps show pause save-offer on cancel screens | save_offer_pause |
| 0 of 48 apps show downgrade/cheaper-plan save-offer | save_offer_downgrade |
| 401 genuine cancel-intent screenshots across 48 apps | genuine_cancel_intent_subset |
| Headway 'trial expired' modal offering 33% discount to renew | qualitative / headway discount-to-retain |

## Methodology

Universe: 48 apps whose cancel-category screenshots literally mention 'cancel' (401 screenshots), July 2026. Method: company-level presence of tactic keywords in vision descriptions. Caveat: zeros reflect absence in captured text, not confirmed absence in product.

## Visual references

Real, in-market app screens tracked by Lazyweb Research:

- Linkedin: https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/screenshots/business4_linkedin/backlog/2026_01_07_08-19-55_A7F3811C.png
- Any Do: https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/screenshots/prod120_any-do/lazyscreenshots/2024_05_11_23-28-48_B352840D.png
- Reader: https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/screenshots/uploaded_reader/paywall/compare/2026-05-28/1780024840157_00_Reader-20260529-032039-01.png
- Clime: https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/screenshots/uploaded_clime/paywall/compare/2026-05-28/1780021306222_00_Clime-20260529-022145-01.png

## Sources & citations

- [1] Lazyweb Research analysis of 48 apps (genuine cancel-intent screens), July 2026. Company-level presence of discount/offer language on cancel screens.
- [2] Lazyweb Research analysis of 48 apps (genuine cancel-intent screens), July 2026. Pause language absent in captured cancel vision text; absence, not proof of no offer.
- [3] Lazyweb Research analysis of 48 apps (genuine cancel-intent screens), July 2026. Downgrade/cheaper-plan language absent in captured cancel vision text.

## Related questions

- [How many steps should I expect in a subscription cancel flow?](https://design.lazyweb.com/blog/how-many-steps-in-a-subscription-cancel-flow)
- [Are pause and downgrade save offers worth building for cancellation?](https://design.lazyweb.com/blog/are-pause-and-downgrade-save-offers-worth-building)
- [What reasons do cancellation surveys ask about, and how common are they?](https://design.lazyweb.com/blog/what-reasons-do-cancellation-surveys-ask-about)
- [What do click-to-cancel flows look like screen by screen in the field?](https://design.lazyweb.com/blog/what-do-compliant-click-to-cancel-flows-look-like)
- [How is the cancel button styled versus the keep/subscribe path?](https://design.lazyweb.com/blog/how-is-the-cancel-button-styled-in-cancel-flows)
- [How common is an 'are you sure' confirmation step when cancelling?](https://design.lazyweb.com/blog/how-common-is-a-confirmation-step-when-cancelling)
- [How common is 'cancel anytime' reassurance copy on paywalls?](https://design.lazyweb.com/blog/how-common-is-cancel-anytime-reassurance-on-paywalls)
- [Does 'cancel anytime' copy help or hurt paywall conversion?](https://design.lazyweb.com/blog/does-cancel-anytime-copy-help-or-hurt-conversion)

## Explore the evidence

- [Billing toggle experiment kit](https://design.lazyweb.com/blog/billing-toggle-experiment-kit?utm_source=lazyweb&utm_medium=research&utm_campaign=billing-toggle-kit)
- [Cal AI annual-plan and trial presentation](https://design.lazyweb.com/blog/cal-ai-paywall-annual-plan-and-trial)
- [Monthly versus annual billing toggles](https://design.lazyweb.com/blog/monthly-vs-annual-billing-toggle-pricing)
- [Billing cadence versus plan tiers](https://design.lazyweb.com/blog/annual-billing-toggle-vs-plan-tiers)
- [Cal AI trial-introduction screens](https://design.lazyweb.com/flow/cal-ai/onboarding-paywall)

## Continue your research

[Connect Lazyweb to your coding agent](https://www.lazyweb.com/mcp-install)