Churn Prevention

• You have no cancel flow today — customers click cancel and are immediately removed, leaving potential saves on the table.
• Your churn rate is climbing and you need a structured diagnosis: voluntary vs. involuntary breakdown, cancel reason analysis, and targeted interventions.
• Failed payments are silently canceling accounts and you need smart retry logic plus a dunning email sequence.
• You want proactive retention — building health score models and triggered interventions before customers ever reach the cancel button.

• You need post-cancel win-back email sequences — use the email-sequence skill instead, which is purpose-built for that flow.
• You're building in-app upgrade paywalls or trial expiration screens — use the paywall-upgrade-cro skill for those conversion moments.
• You have no subscribers yet — this skill is most valuable once you have active paying customers and measurable churn data.

Guides you through a five-step cancel flow: trigger → exit survey → dynamic save offer → confirmation → post-cancel path. Crucially, save offers are matched to cancellation reasons — a pause for seasonal users, a discount for price-sensitive ones, a roadmap preview for those missing a feature — rather than applying a blanket discount to everyone.

Provides a ready-to-use set of cancellation reason categories with guidance on survey copy, option ordering, and 'help us improve' framing that reduces friction. Includes best practices like limiting to 5-8 options, single-select with optional free text, and quarterly review of reason distribution.

Covers the full dunning stack: pre-dunning card expiry alerts, smart retry logic by decline type (soft vs. hard vs. authentication-required), a four-email dunning sequence with timing and tone guidance, and recovery benchmarks to measure against. Includes provider-specific notes for Stripe, Chargebee, Paddle, Recurly, and Braintree.

Defines high-signal leading indicators (login drop, key feature stoppage, billing page visits, data export initiation) and a weighted health score formula (0–100) that maps to action tiers. Includes a proactive intervention table pairing specific triggers to specific outreach actions before customers reach the cancel screen.

Defines the key churn metrics to track (save rate, offer acceptance rate, pause reactivation rate, dunning recovery rate) with benchmark targets. Provides a structured A/B test menu for cancel flow variables — discount depth, pause duration, survey placement, offer presentation — and recommends PostHog feature flags for implementation.

Maps leading retention platforms (Churnkey, ProsperStack, Raaft, Chargebee Retention) to use cases, and lists billing provider dunning capabilities side-by-side. Also covers analytics and event-routing tools (Customer.io, PostHog, Mixpanel, Segment) for health scoring and campaign execution.

A self-serve SaaS product with instant cancellation uses this skill to design a survey → save offer → confirmation sequence. The skill provides the UI wireframe, reason category list, offer-to-reason mapping table, and copy guidance, so the team can ship a working flow without starting from a blank page.

A subscription business notices 40% of cancellations come from failed payments. The skill walks through classifying decline types, configuring smart retry schedules, drafting a four-email dunning sequence with direct payment-update links, and setting up card expiry alerts — turning silent failures into recoverable events.

A B2B SaaS team wants to intervene before customers cancel. Using the health score model and risk signal table, they instrument usage events, build a scoring pipeline, and set up triggered campaigns (e.g., re-engagement email on 14-day login absence, CS escalation for NPS detractors) that catch churn risk weeks before the cancel button is clicked.

A team with an existing cancel flow has a 12% save rate and wants to improve it. The skill's A/B test menu, offer-acceptance benchmarks, and cohort segmentation framework help them identify whether the issue is offer type, discount depth, survey design, or copy tone — and run statistically rigorous tests to find the fix.