# Churn Prevention — Marketing Skill for AI Agents

Churn splits into voluntary (customers choosing to cancel, typically 50-70% of total) and involuntary (failed payments, 30-50% and often the easiest to fix). For voluntary churn, build a cancel flow — trigger, exit survey, dynamic save offer, confirmation, post-cancel — and match the offer to the stated reason: a discount for 'too expensive,' a pause for 'not using it enough,' a roadmap preview for 'missing a feature.' For involuntary churn, run a dunning stack: pre-dunning card-expiry alerts, smart retries staggered over 7-10 days, and a 4-email escalating sequence. Keep the cancel path visible (no dark patterns), avoid 50%+ discounts that train customers to cancel for deals, and track save rate (target 25-35%) plus dunning recovery (50-60%).

> Maintained by Corey Haines · v2.0.0 · Updated 2026-05-13

Install: `npx skills add coreyhaines31/marketingskills`
Source: https://github.com/coreyhaines31/marketingskills/tree/main/skills/churn-prevention

## What's inside

**The playbook covers:** Before Starting, How This Skill Works, Cancel Flow Design, Churn Prediction & Proactive Retention, Involuntary Churn: Payment Recovery, Metrics & Measurement, Common Mistakes, Tool Integrations

### Reference library

- [Cancel Flow Patterns](https://github.com/coreyhaines31/marketingskills/blob/main/skills/churn-prevention/references/cancel-flow-patterns.md) — Detailed cancel flow patterns by business type, billing provider, and industry.
- [Dunning Playbook](https://github.com/coreyhaines31/marketingskills/blob/main/skills/churn-prevention/references/dunning-playbook.md) — Complete guide to recovering failed payments and reducing involuntary churn.

## Key data

- Example flow for a $29/mo productivity app:**
- New customer (< 30 days):**
- Dormant customer (no login 30+ days):**
- Option A: Full access during grace (recommended for B2B)**
- Option B: Read-only access (recommended for B2C)**

## Example

**Prompt:** Our SaaS product has a 7% monthly churn rate and we need to bring it down. We're a $49/month project management tool with about 2,000 paying customers. Can you help us design a churn prevention strategy?

**Offer-to-reason mapping**

Too expensive        → 20-30% off for 2-3 months  (fallback: downgrade)
Not using it enough  → Pause 1-3 months           (fallback: onboarding session)
Missing a feature    → Roadmap preview + timeline  (fallback: workaround guide)
Switching to competitor → Comparison + discount    (fallback: feedback session)
Technical issues     → Escalate to support now     (fallback: credit + priority fix)
Temporary / seasonal → Pause subscription          (fallback: temporary downgrade)
Business closed      → Skip the offer, let them go gracefully

**Dunning email sequence**

Day 0  — Friendly alert: "Your payment didn't go through. Update your card."
Day 3  — Reminder: "Quick reminder — update your payment to keep access."
Day 7  — Urgency: "Your account will be paused in 3 days."
Day 10 — Final warning: "Last chance to keep your account active."
(Direct link to update page, no login required. Plain text beats designed emails here.)

## FAQ

### What should I show users when they click cancel?

Run a short flow instead of cancelling instantly: an exit survey (one single-select question, 5-8 reasons, plus 'other' with free text), then a dynamic save offer matched to their reason, then a clear confirmation with end-of-billing-period messaging and an easy reactivation path. Keep the 'continue cancelling' option visible the whole way — hidden cancel buttons breed bad reviews, and many jurisdictions require easy cancellation. Even a simple survey plus one offer typically saves 10-15%.

### How do I recover failed payments?

Failed payments cause 30-50% of churn and are the most recoverable. Use the dunning stack: pre-dunning (email 30/15/7 days before a card expires, offer a backup payment method, enable card-updater services that cut hard declines 30-50%), smart retries staggered at roughly 24 hours, day 3, day 5, and day 7, and a 4-email sequence escalating from friendly alert to final warning. Match strategy to decline type — retry soft declines like insufficient funds, but don't retry hard declines; ask for a new card instead.

### 40% of my cancellations say 'too expensive' — what offers should I try?

Try a temporary discount (20-30% for 2-3 months), a downgrade to a cheaper tier, an annual-billing discount, or a pause instead of a cancel, and test which wins on save rate and downstream LTV. But 'too expensive' often masks 'I'm not seeing enough value' — add a follow-up question to the exit survey to separate genuine price sensitivity from low engagement, since the second group needs onboarding help, not a discount. Avoid 50%+ discounts that train customers to cancel for deals.

## Related skills

- **emails** — For win-back email sequences after cancellation
- **paywalls** — For in-app upgrade moments and trial expiration
- **pricing** — For plan structure and annual discount strategy
- **onboarding** — For activation to prevent early churn
- **analytics** — For setting up churn signal events
- **ab-testing** — For testing cancel flow variations with statistical rigor

---
Canonical: https://marketing-skills.com/skills/churn-prevention