Designing a Health Score Model

A health score is a composite metric that quantifies the strength, risk, and opportunity within a customer relationship. It replaces the account manager's gut feeling with a structured assessment that is consistent, comparable, and trackable over time. The gut feeling is not wrong — experienced relationship managers have excellent intuition. But intuition cannot be audited, compared across accounts, or tracked for trends. A health score can.

The model has three dimensions, each scored independently and then combined. Engagement health: how actively is the customer interacting with you? Meeting frequency, response times, executive access, and stakeholder breadth all contribute. Customers who are engaged are healthy. Customers who go silent are at risk. Value health: is the customer receiving measurable value from the relationship? Are they using what they bought? Are the outcomes they expected materializing? Value delivery is the foundation of retention. Strategic health: is the customer's business trajectory aligned with your capabilities? A customer growing into markets where you have deep expertise is a strategic fit. A customer pivoting away from your domain is a strategic risk.

1. Define Engagement Signals Identify the specific, measurable behaviors that indicate engagement: meeting frequency (increasing/stable/declining), email response time, executive participation in meetings, number of stakeholders involved, inbound requests. Each signal gets a weight based on its predictive value for relationship health.
2. Define Value Signals Identify the metrics that indicate value delivery: product usage rates, reported outcomes vs. promised outcomes, support ticket volume and sentiment, expansion conversations, reference willingness. A customer who is receiving value behaves differently from one who is not — the value signals capture that behavioral difference.
3. Define Strategic Signals Identify the external factors that indicate strategic alignment: customer growth trajectory, market expansion direction, technology adoption patterns, leadership stability. Strategic signals are forward-looking — they predict whether the relationship will grow, plateau, or contract.
4. Weight and Combine Assign weights to each dimension based on what best predicts retention and expansion in your specific context. A common starting point: engagement 40%, value 40%, strategic 20%. Combine into a single composite score (0-100). The score alone is useful. The dimensional breakdown is more useful — a customer with high engagement but low value delivery is a different risk profile than one with low engagement but high value delivery.