/health-score-monitor

name: health-score-monitor
description: Design and maintain customer health scoring systems with automated alerts and trending analysis
license: MIT
metadata:
  author: ClawFu
  version: 1.0.0
  mcp-server: "@clawfu/mcp-skills"

Health Score Monitor

> Build systematic customer health monitoring with composite scores, trend tracking, and automated alerting for proactive customer success.

When to Use This Skill

• Designing health score frameworks
• Setting up monitoring dashboards
• Creating alert thresholds
• Analyzing health trends across portfolio
• Optimizing existing health models

Methodology Foundation

Based on **Gainsight Health Score Design** and **Totango Customer Success metrics**, focusing on:

• Multi-dimensional scoring
• Leading vs lagging indicators
• Score normalization
• Trend analysis
• Alert prioritization

What Claude Does vs What You Decide

| Claude Does | You Decide | |-------------|------------| | Designs scoring framework | Dimension weights | | Calculates composite scores | Alert thresholds | | Identifies trending patterns | Intervention triggers | | Suggests monitoring cadence | Resource allocation | | Recommends improvements | Business rule exceptions |

What This Skill Does

1. **Framework design** - Multi-factor health model 2. **Score calculation** - Weighted composite scores 3. **Trend analysis** - Direction and velocity 4. **Alert rules** - When to notify teams 5. **Portfolio view** - Aggregate health visibility

How to Use

Design a health score monitor for my customer portfolio:

Business Context:
- Product type: [SaaS/Platform/Service]
- Contract model: [Annual/Monthly/Multi-year]
- Key value metric: [What shows customer success?]
- CSM:Account ratio: [1:X]

Available Data Points:
- Product: [List usage metrics available]
- Support: [List support metrics available]
- Financial: [List financial signals]
- Relationship: [List engagement data]

Current Challenges:
- [What's not working with current approach?]

Instructions

Step 1: Define Health Dimensions

**Standard 4-Pillar Model:**

| Dimension | Weight | What It Answers | |-----------|--------|-----------------| | **Product** | 30-40% | Are they using it? | | **Support** | 15-25% | Are they happy? | | **Financial** | 20-25% | Are they paying/growing? | | **Relationship** | 20-25% | Are we connected? |

Adjust weights based on your business:

• High-touch: Increase Relationship
• Usage-based pricing: Increase Product
• Support-intensive: Increase Support

Step 2: Select Metrics per Dimension

**Product Health Metrics:** | Metric | Type | Scoring | |--------|------|---------| | DAU/MAU | Leading | % of benchmark | | Feature adoption | Leading | % features used | | Time in product | Leading | Minutes vs avg | | Key feature usage | Leading | Yes/No or frequency | | Usage trend | Leading | Up/Flat/Down |

**Support Health Metrics:** | Metric | Type | Scoring | |--------|------|---------| | CSAT score | Lagging | 1-5 scale | | Ticket volume | Leading | vs baseline | | Escalations | Leading | Count (negative) | | Response sentiment | Leading | Positive/Neutral/Negative | | Time to resolution | Lagging | vs SLA |

**Financial Health Metrics:** | Metric | Type | Scoring | |--------|------|---------| | Payment status | Lagging | Current/Late | | Expansion | Leading | Pipeline/Discussion | | Contract type | Lagging | Multi-year bonus | | Renewal date | Context | Days remaining | | ARR trend | Lagging | Growth/Flat/Decline |

**Relationship Health Metrics:** | Metric | Type | Scoring | |--------|------|---------| | Champion engaged | Leading | Active/Passive/None | | Exec sponsor | Leading | Yes/No | | NPS/CSAT | Lagging | Score | | QBR attendance | Leading | Regular/Sporadic/None | | Last touch | Leading | Days since contact |

Step 3: Create Scoring Rules

**Example Scoring Matrix:**

Product Score (0-100):
+30: DAU/MAU > 70%
+20: DAU/MAU 50-70%
+10: DAU/MAU 30-50%
+0:  DAU/MAU < 30%

+25: Feature adoption > 60%
+15: Feature adoption 40-60%
+5:  Feature adoption < 40%

+20: Usage trending up
+10: Usage stable
-10: Usage trending down

+15: Key feature used weekly
+5:  Key feature used monthly
-10: Key feature abandoned

Max: 100 (or normalize)

Step 4: Set Alert Thresholds

| Condition | Alert Level | Action | |-----------|-------------|--------| | Score drops >15 pts | 🔴 Critical | Immediate CSM review | | Score below 40 | 🟠 High | Intervention required | | Score 40-60 | 🟡 Medium | Proactive outreach | | Renewal <60 days + <60 score | 🔴 Critical | Save plan | | Usage drop >30% WoW | 🟡 Medium | Check-in call |

Step 5: Define Monitoring Cadence

| Activity | Frequency | Owner | |----------|-----------|-------| | Score refresh | Daily/Weekly | Automated | | Alert review | Daily | CSM | | Portfolio review | Weekly | CS Manager | | Model calibration | Quarterly | CS Ops |

Examples

Example 1: SaaS Health Score Design

**Input:**

Design health score for our B2B SaaS platform:

Business Context:
- Product: Project management software
- Contract: Annual, avg $50K
- Value metric: Projects completed, team adoption
- CSM ratio: 1:40

Available Data:
- Product: Logins, projects created, tasks completed, integrations used
- Support: Zendesk tickets, CSAT, NPS
- Financial: ARR, payment status, expansion pipeline
- Relationship: Meeting attendance, email engagement

Challenges:
- Too many "yellow" accounts, can't prioritize
- Alerts are noisy, CSMs ignoring them

**Output:**

## Health Score Framework: Project Management SaaS

### Design Principles

Given 1:40 ratio and "alert fatigue" issue:
1. **Prioritize leading indicators** - Catch issues early
2. **Reduce noise** - Tighter alert thresholds
3. **Enable triage** - Clear Red/Yellow/Green actions

---

### Scoring Model

#### Dimension Weights
| Dimension | Weight | Rationale |
|-----------|--------|-----------|
| Product | 40% | Usage = value realization |
| Support | 15% | Lower to