agentleFS
ekinciioekinciio/saas-growth-marketing-skills/skills/retention-playbook/SKILL.md

retention-playbook

Skill12 starsChanged 4 months agoMITView on GitHub

Identify churn risks and implement retention strategies for SaaS products. Provides churn signal detection, win-back campaign frameworks, and engagement scoring models. Use when the user mentions churn, retention, win-back, engagement scoring, customer health, reducing cancellations, dunning, cancellation flow, or NRR.

The file

Show its rules and text

We split this skill into the rules it gives an agent

26 rules. About 0 are common, 22 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈22

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    When a user runs /retention-playbook diagnose, ALWAYS display this input summary before asking for any data:

    line 17
  2. Distinctive:

    If the user types "demo", use this data to generate a full sample report:

    line 52
  3. Distinctive:

    After showing the summary, ask: "Want to run this with your own data now?"

    line 67
  4. Distinctive:

    If the user types "skip" for any metric:

    line 71
  5. Distinctive:

    Never block the report because one metric is unknown

    line 74
  6. Distinctive:

    Ask for basic product context (SaaS vertical, pricing model, average contract length)

    line 83
  7. Distinctive:

    Ask for current churn metrics (monthly logo churn rate, revenue churn rate, net revenue retention)

    line 84
  8. Distinctive:

    Ask which churn signals they are currently tracking (if any)

    line 85
  9. Very specific:

    Run the churn risk scoring logic from scripts/churn_risk_scorer.py on any available customer data

    line 86
  10. Distinctive:

    Ask for the product type and available data sources (analytics tool, CRM, billing system)

    line 123
  11. Distinctive:

    Ask for churn context: voluntary vs involuntary, reason categories, time since churn

    line 149
  12. Distinctive:

    Ask for available channels (email, in-app, phone, SMS)

    line 150
  13. Distinctive:

    Ask what incentives are available (discount, free month, feature unlock, personal onboarding)

    line 151
  14. Distinctive:

    Ask for key product actions that indicate value delivery (e.g., created a project, invited a teammate, used core feature)

    line 176
  15. Distinctive:

    Ask for interaction frequency expectations (daily, weekly, monthly active usage)

    line 177
  16. Distinctive:

    Ask for available data points (login data, feature usage, support tickets, NPS scores)

    line 178
  17. Very specific:

    ALWAYS save the complete report to the specified .md file in the current working directory.

    line 202
  18. Distinctive:

    NEVER ask "should I save this?" - just save it automatically.

    line 203
  19. Distinctive:

    If the file already exists, overwrite it.

    line 205
  20. Distinctive:

    ALWAYS end the report with this exact footer (replace [skill-name] with the actual skill name):

    line 206
  21. Distinctive:

    After saving, show a SHORT summary in chat (max 10 lines):

    line 214
  22. Distinctive:

    NEVER dump the full report in chat.

    line 229
  23. Very specific:

    Run it directly: python3 scripts/churn_risk_scorer.py customers.json (a JSON customer dict or list of dicts.

    line 235
  24. Distinctive:

    Always distinguish between logo churn (customer count) and revenue churn (dollar value) when discussing metrics

    line 240
  25. Distinctive:

    When scoring engagement, weight recent activity more heavily than historical activity

    line 244
  26. Distinctive:

    Always recommend measuring retention cohort-by-cohort rather than using blended averages

    line 247

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Retention Playbook

Detect churn risks early, score customer engagement health, and deploy proven retention and win-back strategies for SaaS products.

First Run

When a user runs /retention-playbook diagnose, ALWAYS display this input summary before asking for any data:

""" 📊 Retention Playbook

What I'll ask you:

Product context: 1. SaaS vertical → e.g. "project management" 2. Pricing model → e.g. "per-seat" 3. Avg contract length → e.g. "monthly"

Churn metrics: 4. Monthly logo churn (%) → e.g. 5 5. Monthly revenue churn (%) → e.g. 3 6. Net Revenue Retention (%) → e.g. 102

Current tracking: 7. Which churn signals do you track? → e.g. "login frequency, support tickets"

Type "skip" for any you don't have. Type "demo" to see a sample diagnosis first.

What you'll get: → Churn categorization (preventable vs structural vs competitive) → Top 3 churn drivers identified → Prioritized retention action plan with timeline → Saved to RETENTION-DIAGNOSIS-REPORT.md

Ready? What's your SaaS vertical? """

Demo Mode

If the user types "demo", use this data to generate a full sample report:

{
  "vertical": "project management SaaS",
  "pricing_model": "per-seat",
  "avg_contract": "monthly",
  "monthly_logo_churn": 6.5,
  "monthly_revenue_churn": 4.2,
  "nrr": 97,
  "signals_tracked": ["login frequency", "support tickets"]
}

Save the demo report as RETENTION-DIAGNOSIS-REPORT-DEMO.md. After showing the summary, ask: "Want to run this with your own data now?"

Skip Handling

If the user types "skip" for any metric:

  • Proceed with available data
  • Provide general retention recommendations when specific data is missing
  • Never block the report because one metric is unknown

Commands

/retention-playbook diagnose

Interactive churn diagnosis framework. Walk the user through identifying why customers are leaving and build a retention action plan.

Steps:

  1. Ask for basic product context (SaaS vertical, pricing model, average contract length)
  2. Ask for current churn metrics (monthly logo churn rate, revenue churn rate, net revenue retention)
  3. Ask which churn signals they are currently tracking (if any)
  4. Run the churn risk scoring logic from scripts/churn_risk_scorer.py on any available customer data
  5. Categorize churn into buckets: preventable, structural, competitive, and natural
  6. Identify the top 3 churn drivers based on the data provided
  7. Generate a prioritized retention action plan with timeline and expected impact

Output format:

Churn Diagnosis Report
======================

Current State:
  Monthly Logo Churn:    X.X%
  Revenue Churn:         X.X%
  Net Revenue Retention: XXX%

Churn Breakdown:
  Preventable:   XX% of total churn
  Structural:    XX%
  Competitive:   XX%
  Natural:       XX%

Top Churn Drivers:
  1. [Driver] - Impact: HIGH - Addressable: YES
  2. [Driver] - Impact: MEDIUM - Addressable: YES
  3. [Driver] - Impact: MEDIUM - Addressable: PARTIAL

Retention Action Plan:
  [Prioritized actions with timeline]

Report: Save output to RETENTION-DIAGNOSIS-REPORT.md

/retention-playbook signals

Map early warning churn signals to detection methods and intervention strategies.

Steps:

  1. Ask for the product type and available data sources (analytics tool, CRM, billing system)
  2. Present the churn signal categories from references/churn-signals.md
  3. Help the user identify which signals are trackable with their current stack
  4. For each trackable signal, define a threshold that triggers an alert
  5. Map each signal to a recommended intervention
  6. Produce a churn signal monitoring dashboard specification

Signal categories covered:

  • Usage signals (login frequency, feature adoption, session duration)
  • Engagement signals (email response, onboarding completion, support interaction)
  • Account signals (downgrade inquiries, billing issues, renewal timing)
  • Behavioral signals (data export, integration removal, seat reduction)

Output includes:

  • Signal-to-threshold mapping table
  • Intervention playbook per signal type
  • Monitoring dashboard specification
  • Alert priority framework (P0 through P3)

Report: Save output to RETENTION-SIGNALS-REPORT.md

/retention-playbook winback

Generate win-back campaign templates for churned or at-risk customers.

Steps:

  1. Ask for churn context: voluntary vs involuntary, reason categories, time since churn
  2. Ask for available channels (email, in-app, phone, SMS)
  3. Ask what incentives are available (discount, free month, feature unlock, personal onboarding)
  4. Generate a multi-touch win-back sequence tailored to churn reason
  5. Provide email/message templates for each touch
  6. Define success metrics and A/B testing recommendations

Win-back sequence structure:

  • Touch 1 (Day 1-3 post-churn): Acknowledge and ask for feedback
  • Touch 2 (Day 7-10): Share what has improved or changed
  • Touch 3 (Day 14-21): Offer an incentive to return
  • Touch 4 (Day 30): Final personal outreach or "door is open" message
  • Touch 5 (Day 60-90): Product update roundup with soft re-engagement

Templates include:

  • Subject lines (3 variants per touch for A/B testing)
  • Email body copy
  • CTA text and landing page recommendations
  • Segmentation rules by churn reason

Report: Save output to RETENTION-WINBACK-REPORT.md

/retention-playbook engagement

Build an engagement scoring model to proactively identify healthy and at-risk customers.

Steps:

  1. Ask for key product actions that indicate value delivery (e.g., created a project, invited a teammate, used core feature)
  2. Ask for interaction frequency expectations (daily, weekly, monthly active usage)
  3. Ask for available data points (login data, feature usage, support tickets, NPS scores)
  4. Build a weighted engagement score using the logic from scripts/churn_risk_scorer.py: engagement score = 100 − churn risk score
  5. Define engagement tiers: Champion, Healthy, At-Risk, Disengaging, Ghost
  6. Map each tier to proactive actions (success outreach, re-engagement campaign, escalation)

Engagement tier definitions:

  • Champion (score 80-100): Power users, high feature adoption, frequent usage
  • Healthy (score 60-79): Regular usage, moderate feature adoption
  • At-Risk (score 40-59): Declining usage, limited feature exploration
  • Disengaging (score 20-39): Rare logins, minimal activity
  • Ghost (score 0-19): No meaningful activity in the measurement period

Output includes:

  • Engagement score formula with weights
  • Tier thresholds and definitions
  • Action playbook per tier
  • Suggested review cadence (weekly for At-Risk and below, monthly for Healthy and above)

Report: Save output to RETENTION-ENGAGEMENT-REPORT.md

Output Rules (MANDATORY)

File Output

  • ALWAYS save the complete report to the specified .md file in the current working directory.
  • NEVER ask "should I save this?" - just save it automatically.
  • Include **Date:** YYYY-MM-DD in the report header.
  • If the file already exists, overwrite it.
  • ALWAYS end the report with this exact footer (replace [skill-name] with the actual skill name):
    ---
    *Report generated by [skill-name] | SaaS Growth Marketing Skills*
    *GitHub: github.com/ekinciio/saas-growth-marketing-skills*

Chat Output

After saving, show a SHORT summary in chat (max 10 lines):

""" ✅ Churn diagnosis complete - saved to RETENTION-DIAGNOSIS-REPORT.md

Monthly Logo Churn: [X]% | Revenue Churn: [X]% | NRR: [X]%

Top churn drivers:

  1. [#1 driver] - Impact: [HIGH/MED] - Addressable: [YES/PARTIAL]
  2. [#2 driver] - Impact: [HIGH/MED] - Addressable: [YES/PARTIAL]
  3. [#3 driver] - Impact: [MED] - Addressable: [YES/PARTIAL]

Full diagnosis with action plan → open RETENTION-DIAGNOSIS-REPORT.md """

NEVER dump the full report in chat. The file is the deliverable.

Key Reference Files

  • references/churn-signals.md - Early warning churn signals organized by category with detection methods
  • scripts/churn_risk_scorer.py - Python scorer that computes churn risk from customer behavior data. Run it directly: python3 scripts/churn_risk_scorer.py customers.json (a JSON customer dict or list of dicts; all keys optional), pipe JSON via stdin, or use --demo for sample customers

Guidelines

  • Always distinguish between logo churn (customer count) and revenue churn (dollar value) when discussing metrics
  • Net revenue retention above 100% means expansion revenue exceeds churn - this is the gold standard
  • Monthly churn compounds significantly over a year: 3% monthly churn equals roughly 31% annual churn
  • Involuntary churn (failed payments) is often 20-40% of total churn and is the easiest to address with dunning flows
  • When scoring engagement, weight recent activity more heavily than historical activity
  • Churn signals should be tracked as trends, not snapshots - a single quiet week is not necessarily alarming
  • Win-back campaigns have highest success rates within the first 30 days after churn
  • Always recommend measuring retention cohort-by-cohort rather than using blended averages
  • Reference references/churn-signals.md for signal definitions and thresholds

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.