**TL;DR** — Customer Success Manager gives your CS team a structured intelligence layer: health scoring, renewal forecasting, QBR automation, and churn signal detection — so you're acting on data before the customer ghosts you.
1. **Define your health signal taxonomy** — Map your product usage data, support ticket rates, and engagement metrics into a tiered signal schema. The skill uses this to score accounts consistently.
2. **Calibrate against your last 12 months** — Feed it your historical churn cases, renewal outcomes, and expansion closes. Let it learn the patterns your gut already knows but can't articulate at scale.
3. **Connect to your CRM and support stack** — Route CS event data (login frequency drops, support escalations, stakeholder changes) to trigger automated health refreshes and alert thresholds.
**Example Prompt:**
Account: Acme Corp, $240K ARR, Tier 1, renewal due in 90 days. Their weekly active users dropped 34% in the last 30 days, two executive sponsors have changed in 60 days, and their last three support tickets were escalated to L2. Generate: (1) health score update with weighted signal breakdown, (2) renewal risk assessment, (3) a proactive outreach playbook with three escalation paths based on response timing.
| Pros | Cons |
|---|---|
| Turns reactive CS into proactive, data-driven retention | Requires clean, consistent CRM hygiene to function well |
| Surfaces expansion signals inside existing workflow | Health scoring is only as good as the signal taxonomy you define |
|---|
| Reduces quarter-end renewal surprises | May over-flag accounts if thresholds aren't properly tuned |
|---|
| Frees CSMs for relationship work, not data assembly | Doesn't replace human judgment on high-stakes renewals |
|---|