Exercise Overview

Time Required: 45 minutes

Format: Individual or small group

Deliverable: A customer health scoring model with calculation formula and action thresholds

Objective

Develop a comprehensive customer health scoring system that can predict churn risk and identify expansion opportunities. You'll define the factors to include, assign appropriate weights, and create actionable thresholds for your Customer Success team.

Instructions

Step 1: Select Your Health Score Factors (15 minutes)

Review the potential factors below and select 5-7 that are most relevant to your SaaS business. For each factor you select, define how you'll measure it:

Potential Factors:

• Product Usage: Login frequency, feature usage depth/breadth, active users
• Engagement: Email opens, webinar attendance, community participation
• Support Activity: Number of tickets, severity, resolution time
• Customer Feedback: NPS score, CSAT scores, survey responses
• Account Changes: Payment status, contract changes, champion turnover
• Business Outcomes: ROI achieved, goals met, success milestones
• Relationship Strength: QBR completion, CSM touch points, executive engagement
• Adoption Velocity: Speed of feature adoption, team growth, integration usage

For each factor you select, document:

1. Factor Name:
2. Why it matters: How does this predict churn or expansion?
3. Data Source: Where will you get this data?
4. Measurement: How exactly will you quantify it?

Step 2: Assign Weights and Scoring Thresholds (20 minutes)

Create your health scoring model:

Weighting Exercise

Assign weights to your selected factors (must total 100%):

Factor	Weight (%)	Rationale for Weight
Example: Product Usage	40%	Most predictive of churn in our data
Factor 2
Factor 3
Factor 4
Factor 5
TOTAL	100%

Scoring Scale

For each factor, define the Green/Yellow/Red thresholds:

Factor 1: [Name]

• Green (3 points): [Define threshold, e.g., "80%+ users active weekly"]
• Yellow (2 points): [Define threshold, e.g., "50-79% users active weekly"]
• Red (1 point): [Define threshold, e.g., "<50% users active weekly"]

Factor 2: [Name]

• Green (3 points):
• Yellow (2 points):
• Red (1 point):

(Repeat for each factor)

Overall Score Calculation

Formula:

Total Score = (Factor1 Score × Factor1 Weight) + (Factor2 Score × Factor2 Weight) + ... 

Health Status Ranges:

• 🟢 Green (Healthy): Score 2.5 - 3.0 → Low churn risk, expansion ready
• 🟡 Yellow (At Risk): Score 2.0 - 2.4 → Some concern, needs attention
• 🔴 Red (Critical): Score < 2.0 → High churn risk, immediate action required

Step 3: Test with Example Customers (10 minutes)

Apply your health score model to three hypothetical customers:

Customer A - "Star Account":

• [Describe their characteristics across your factors]
• Calculate health score:
• Status:
• Recommended action:

Customer B - "Middle Ground":

• [Describe their characteristics]
• Calculate health score:
• Status:
• Recommended action:

Customer C - "At Risk":

• [Describe their characteristics]
• Calculate health score:
• Status:
• Recommended action:

Deliverable Template

Create a document that includes:

1. Health Score Model Overview
   • Selected factors and rationale
   • Weights assigned
   • Scoring thresholds (Green/Yellow/Red for each factor)
2. Calculation Formula
   • How scores are computed
   • Overall health status ranges
3. Tested Examples
   • 3 sample customers scored with your model
   • Actions recommended for each
4. Implementation Plan (brief)
   • Data sources needed
   • Tools/systems required
   • Team training needs

Discussion Questions

1. Predictive Validity: How confident are you that your model would accurately predict churn? What historical data would you analyze to validate it?
2. Weights: Which factor did you weight most heavily and why? Would you adjust weights for different customer segments (e.g., SMB vs Enterprise)?
3. Actionability: For each health status (Green/Yellow/Red), what specific actions should the CSM take?
4. Automation: Which factors could be automatically calculated vs require manual input?
5. Iteration: How often would you recalculate scores? How often would you revisit and refine the model itself?

Bonus Challenge

CSM Playbook Integration:

For each health status, create a quick playbook:

• Green (2.5-3.0):
  • CSM Action: [e.g., "Identify upsell opportunities, request referral"]
  • Cadence: [e.g., "Monthly check-in, quarterly QBR"]
• Yellow (2.0-2.4):
  • CSM Action: [e.g., "Schedule deep-dive call, investigate issues"]
  • Cadence: [e.g., "Bi-weekly check-ins until green"]
• Red (<2.0):
  • CSM Action: [e.g., "Escalate to manager, execute save plan"]
  • Cadence: [e.g., "Weekly calls, executive sponsor engaged"]