churn-prevention

Trigger phrases

Phrases that activate this skill when typed to Claude Code:

• churn
• cancel flow
• save offer
• failed payment recovery
• people keep canceling

What it does

churn-prevention is a Claude Code skill from Corey Haines’s marketing-skills repo. It turns Claude into a retention specialist who designs the entire cancel and dunning stack — exit survey, dynamic save offers tied to cancel reason, payment-failure recovery, and a churn-risk health score. The skill activates when you mention “cancel flow”, “save offer”, “dunning”, or “failed payment recovery”.

The output of a session is a cancel-flow specification (survey questions, offer-to-reason mapping, UI patterns), a dunning sequence (smart retry timing, four-email cadence), a churn-risk health score formula, and a list of A/B tests to run on each stage.

When to use it

Reach for it when:

• Cancellation today is instant — no survey, no save offer, no win-back path
• Failed payments silently kill accounts and nobody’s emailing the customer
• You have churn data but no insight into why — and no tested save offers

When not to reach for it:

• The product itself is the problem; no save offer fixes a broken product
• Your cancel rate is already at the floor — diminishing returns

Install

The skill is distributed via Corey Haines’s marketing-skills repo. Install via the repo’s recommended path — copy the churn-prevention SKILL.md into your project’s .claude/skills/churn-prevention/ directory, or use the repo’s plugin install if you’ve set it up.

Once installed, the skill activates on the trigger phrases above. The first time it runs, it will check for .agents/product-marketing-context.md (or .claude/product-marketing-context.md) — populating that file with your product context first dramatically improves output quality across all of Haines’s marketing skills.

What a session looks like

A typical session has three phases:

1. Voluntary stack. Exit survey design with reason categories, then offer-to-reason mapping (“too expensive” → discount or downgrade, “not using it” → pause, “missing feature” → roadmap preview). The pattern is one primary offer + one fallback, not a wall of options.
2. Involuntary stack. Smart retry logic by decline type, four-email dunning cadence (Day 0/3/7/10), card updater services, pre-dunning expiry alerts.
3. Proactive layer. Risk signals (login frequency, feature usage, billing-page visits), a weighted health score, and intervention triggers before a user ever clicks “cancel”.

The discipline that makes it work: matching offer to reason. A flat 30%-off-everything cancel flow saves the wrong customers and trains the rest to cancel for deals.

Receipts

Honest reporting on what churn-prevention produces and where it has limits:

Where it works well:

• The reason→offer mapping is the difference between a save flow and a discount machine
• The dunning playbook recovers a meaningful chunk of involuntary churn — the easiest win in retention
• Proactive health scores surface at-risk accounts before they cancel, not during the exit survey

Where it backfires:

• On B2B with manual contracts the self-serve cancel flow doesn’t fit; the skill leans toward SaaS-self-serve patterns
• Save offers without LTV tracking risk “saving” customers who churn 60 days later anyway

Pattern that works: ship the dunning sequence first — it’s faster to implement, less brand-sensitive, and recovers revenue from customers who want to keep paying.

Source and attribution

Originally written by Corey Haines. The canonical SKILL.md and any supporting files live in the churn-prevention folder of his marketing-skills repository.

License: MIT. You can install, adapt, and redistribute the skill, with attribution preserved.

This page documents the skill from a practitioner’s perspective. For the formal spec and any updates, defer to the source repo.