churn-prevention

Overview of churn-prevention skill

What the churn-prevention skill is for

The churn-prevention skill helps AI agents design practical SaaS retention interventions for voluntary churn, such as cancel flows, exit surveys, save offers, and win-back messaging, plus involuntary churn recovery through dunning and payment retry strategy. It is best for Growth, lifecycle marketing, product marketing, revenue operations, and founders who need an actionable churn-prevention guide rather than a generic “reduce churn” brainstorm.

Best-fit use cases and users

Use this skill when you are building or auditing a cancellation experience, deciding which save offers to show by cancellation reason, writing payment failure emails, or estimating the revenue impact of better save and recovery rates. It is especially useful for subscription products with measurable MRR, failed-payment data, cancellation reasons, and a billing system such as Stripe, Paddle, Chargebee, or Recurly.

What makes it different from a generic prompt

A generic prompt may suggest discounts or emails. This churn-prevention skill adds more operating structure: it separates voluntary and involuntary churn, asks for current-state context, includes a cancel-flow playbook, includes a dunning guide with decline-code logic, and provides a scripts/churn_impact_calculator.py helper for modeling recovered MRR. That makes the output more implementation-oriented and easier to prioritize.

When this skill is not the right fit

Do not install churn-prevention if your main need is customer health scoring, account expansion, renewal forecasting, or customer success playbooks. It is also a weak fit if you cannot access churn rate, cancellation reasons, failed-payment volume, or pricing context; the skill can still produce draft flows, but its recommendations will be less defensible.

How to Use churn-prevention skill

churn-prevention install and repository path

Install the skill from the repository path:

npx skills add alirezarezvani/claude-skills --skill churn-prevention

The source lives at:

marketing-skill/skills/churn-prevention

After install, read SKILL.md first, then open references/cancel-flow-playbook.md, references/dunning-guide.md, and scripts/churn_impact_calculator.py. The references are not decorative; they contain the decision trees, copy patterns, payment-failure taxonomy, and revenue math that make the skill more useful than a short prompt.

Inputs to prepare before using the skill

For strong churn-prevention usage, provide business and billing context up front. Include:

• Product type, audience, pricing tiers, contract model, and trial status
• Current monthly churn rate and voluntary versus involuntary split, if known
• Current cancel path: instant self-serve, support-led, survey-first, or no flow
• Cancellation reasons and any exit survey data
• Payment processor, retry schedule, card updater support, and failed-payment emails
• Current save rate, recovery rate, average customer MRR, and target improvement
• Constraints such as legal requirements, brand tone, engineering effort, and discount policy

If .claude/product-marketing-context.md exists in your workspace, let the agent read it first so it does not ask for context already documented.

Turning a rough goal into a complete prompt

Weak prompt: “Help reduce churn.”

Stronger prompt:

“Use the churn-prevention skill to audit our SaaS cancel flow and dunning setup. We have $80k MRR, 4.2% monthly churn, about 70% voluntary and 30% involuntary. Current cancel flow is instant self-serve with one optional reason dropdown. Top reasons are price, missing feature, and not using it enough. Stripe retries failed payments 3 times over 7 days; we send one generic email. Average customer MRR is $120. We do not want aggressive dark patterns, annual-contract pressure, or discounts longer than 3 months. Produce a prioritized plan, recommended exit survey, reason-based save offers, dunning sequence, and metrics to track.”

This works because it gives the skill segmentation, financial stakes, operational limits, and enough data to recommend tradeoffs instead of generic retention tactics.

Suggested workflow for Growth teams

Start with diagnosis, not copywriting. Ask the skill to classify churn into voluntary and involuntary buckets, identify missing data, and estimate upside with scripts/churn_impact_calculator.py. Then ask for separate plans: a cancel-flow design, an exit-survey reason map, save offers by segment, and a dunning sequence by failure mode. Finally, ask for implementation tickets and measurement events, such as cancel_started, cancel_reason_selected, save_offer_viewed, save_offer_accepted, payment_failed, payment_recovered, and subscription_canceled.

churn-prevention skill FAQ

Is churn-prevention suitable for early-stage SaaS?

Yes, if you have enough cancellations or payment failures to learn from. For very early products, use the churn-prevention skill to design a clean exit survey, basic cancel confirmation, and humane payment recovery emails. Avoid over-optimizing save offers until you know whether churn is caused by price, activation failure, missing features, poor fit, or failed payments.

Can it handle both voluntary and involuntary churn?

Yes. That is the main reason to use this skill. Voluntary churn work focuses on cancel flow design, cancellation reasons, save offers, and post-cancel messaging. Involuntary churn work focuses on failed-payment recovery, retry timing, decline-code handling, card updates, and dunning emails. Keeping those separate prevents the common mistake of treating all churn as a messaging problem.

How is this different from lifecycle email automation?

Lifecycle automation tools send messages; this skill helps decide what should happen and why. It can produce email copy, but its higher-value use is mapping churn reasons to interventions, designing a cancellation decision tree, choosing offer rules, and aligning payment recovery with billing failure types. You still need your ESP, billing system, or product team to implement the flow.

What should I check before installing it?

Check whether your team can act on the recommendations. The skill is most valuable when you can edit cancellation UX, add or change exit-survey reasons, configure payment retries, update dunning emails, and measure save or recovery rates. If all cancellation and billing logic is locked behind a third-party process you cannot change, the output may become a strategy memo rather than an implementation plan.

How to Improve churn-prevention skill

Improve churn-prevention outputs with better segmentation

The biggest quality lever is segmentation. Provide different churn patterns by plan, customer size, lifecycle stage, acquisition channel, geography, or use case. A price objection from a small self-serve customer should not trigger the same save offer as a missing-feature complaint from a high-value account. Ask the skill to produce separate recommendations for each segment only when the segment changes the decision.

Avoid common failure modes

Watch for three issues. First, over-discounting: ask for non-discount saves such as pause, downgrade, onboarding help, feature education, or billing cadence changes. Second, dark-pattern cancellation UX: explicitly require transparent confirmation and easy final cancellation. Third, dunning spam: ask the skill to vary messages by failure type and retry stage instead of sending the same urgent email repeatedly.

Iterate after the first output

After the first plan, ask for a sharper version using constraints: “Reduce this to a two-week implementation plan,” “Separate no-code changes from engineering work,” “Rewrite the cancel page in our brand voice,” or “Prioritize by expected recovered MRR versus effort.” Then compare recommendations against your analytics, support tickets, and billing logs before shipping.

Add local context to make the skill stronger

For repeated churn-prevention usage, keep a short internal context file with pricing, cancellation policy, current metrics, approved discount limits, brand voice, and billing provider details. The skill already checks for product-marketing context when available, so maintaining that file reduces repeated setup and helps future outputs stay consistent with your business rules.