Skip to main content

Customer Health Model - Project Details / Task List

Tag: health-model Total Hours: 30h Structure: Single Milestone (<=50h)


Milestone: Customer Health Model

Description: A technical and strategic implementation project that creates a multi-dimensional scoring system to measure customer health by combining product usage signals, CSM sentiment, support interactions, and engagement metrics into a unified health score that enables proactive churn prevention.


Task List: (Customer Health Model) 1. Discovery & Design

Contains: Parts 1-2

TaskEstDescription
1. Analyze Historical Churn Data2.5hReview past churn events to identify patterns and signals that preceded customer departures. End state: Documented list of leading indicators correlated with churn in your specific business.

• Pull list of all churned customers from past 12-24 months
• Categorize churn reasons (product fit, support issues, champion left, budget, competitor)
• Identify common behaviors before churn (usage decline, support ticket spikes, missed QBRs)
• Interview 2-3 CSMs on what they noticed before accounts churned
• Document 5-10 potential health score signals based on churn analysis
2. Define Health Score Dimensions and Metrics2.5hIdentify the specific metrics that will comprise the health score across multiple dimensions. End state: Finalized list of 5-10 metrics organized by dimension with clear definitions.

• Define Product dimension metrics (login frequency, feature adoption, license utilization, active user trend)
• Define Engagement dimension metrics (QBR attendance, email response rate, executive sponsor engagement)
• Define Support dimension metrics (ticket volume, escalation frequency, time to resolution)
• Define Financial dimension metrics (payment timeliness, expansion history, renewal timing)
• Include CSM sentiment as a qualitative override dimension
• Document data source for each metric (CRM, product analytics, support system)
3. Validate Metrics with Stakeholders1.5hReview proposed metrics with CS leadership and cross-functional partners to ensure buy-in. End state: Approved metric list with stakeholder alignment on what signals matter.

• Schedule review meeting with VP Customer Success and CS managers
• Present proposed metrics with rationale from churn analysis
• Gather feedback on missing signals or irrelevant metrics
• Validate data availability with RevOps/Data team
• Finalize approved metric list with documented rationale
4. Establish Metric Weights2hAssign relative importance weights to each metric based on correlation with retention outcomes. End state: Weighted scoring model with percentages assigned to each metric.

• Analyze historical correlation between each metric and renewal/churn outcomes
• Assign weights totaling 100% across all metrics (e.g., product usage 40%, engagement 25%, support 20%, financial 15%)
• Consider segment-specific weights (enterprise vs SMB may weight differently)
• Document rationale for each weight assignment
• Create scoring formula combining weighted metrics
5. Define Scoring Thresholds and Categories2hEstablish the score ranges that define Red (at-risk), Yellow (needs attention), and Green (healthy) accounts. End state: Clear threshold definitions with category descriptions.

• Set numeric thresholds for each category (e.g., 0-40 Red, 41-70 Yellow, 71-100 Green)
• Define what each category means operationally (Red = immediate intervention required)
• Document edge cases and override rules (e.g., CSM can override if major event occurred)
• Consider adding "Black" category for accounts with insufficient data
• Test thresholds against historical churned accounts to validate accuracy
6. Design Segment-Specific Scoring Rules2hCustomize scoring criteria for different customer segments to ensure fair comparison. End state: Segment-specific scoring rules that account for usage patterns by account size.

• Define customer segments (Enterprise, Mid-Market, SMB, or by product tier)
• Adjust usage expectations by segment (50% login rate may be healthy for Enterprise, concerning for SMB)
• Set different metric weights per segment if needed
• Document lookup windows for each metric (e.g., trailing 30 days vs 90 days)
• Create segment mapping rules for automated categorization

Task List: (Customer Health Model) 2. Implementation & Enablement

Contains: Parts 3-5

TaskEstDescription
7. Configure Data Integrations3hSet up data flows from source systems into the customer success platform. End state: All required data sources connected and syncing reliably.

• Connect CRM (Salesforce/HubSpot) to CS platform via native integration
• Set up product analytics integration (Segment, Amplitude, Mixpanel, or direct API)
• Configure support system integration (Zendesk, Intercom, Freshdesk)
• Establish billing/financial data sync from payment system
• Verify data freshness and sync frequency for each source
• Document any data transformation or field mapping required
8. Build Health Score in CS Platform3.5hConfigure the health score calculation in Gainsight, ChurnZero, Vitally, or similar platform. End state: Automated health score calculating correctly for all accounts.

• Create health score measures for each defined metric
• Configure metric calculations with correct lookback windows
• Apply weights to each measure per approved model
• Set up segment-based scoring variations if applicable
• Configure score refresh frequency (daily recommended)
• Set up CSM sentiment input field for manual override
9. Create Health Score Dashboard3hBuild reporting dashboards for health score visibility across the CS organization. End state: Dashboard showing account health distribution with drill-down capability.

• Create executive summary view (health distribution pie chart, trend over time)
• Build CSM-level view filtered by book of business
• Include health score trend charts for individual accounts
• Add drill-down to see which metrics are driving low scores
• Configure alerts for accounts crossing threshold boundaries
• Set up scheduled dashboard delivery to CS leadership
10. Test Score Accuracy Against Historical Data2hValidate the health score model by checking if it correctly identifies known churned accounts. End state: Validated model with documented accuracy metrics.

• Run health scores retrospectively against accounts that churned 6+ months ago
• Calculate false positive rate (Green accounts that churned)
• Calculate false negative rate (Red accounts that renewed/expanded)
• Adjust thresholds or weights if accuracy is below 70%
• Document model accuracy baseline for future comparison
• Identify any segments where model performs poorly
11. Conduct Pilot with CSM Team2hRun pilot period with subset of CSMs to validate score usefulness and gather feedback. End state: Pilot complete with documented feedback and refinements.

• Select 2-3 CSMs for pilot program (mix of tenure and segment coverage)
• Review their top 10 accounts by health score (high and low)
• Validate scores match CSM intuition and known account situations
• Gather feedback on missing signals or incorrect assessments
• Refine model based on pilot learnings
• Document changes made and rationale
12. Develop Intervention Playbooks2hCreate standardized response playbooks for each health score category. End state: Documented playbooks defining actions for Red, Yellow, and Green accounts.

• Define Red account intervention steps (executive escalation, QBR scheduling, product review)
• Define Yellow account monitoring cadence and proactive touchpoints
• Define Green account expansion opportunity identification
• Create escalation path for accounts declining into Red status
• Document SLA for CSM response to score changes
• Build playbook templates in CS platform if supported
13. Train Customer Success Team1.5hEnable the CS team on health score interpretation and usage. End state: Team trained and confident using health scores in daily workflow.

• Schedule training session (45-60 minutes) with full CS team
• Explain each metric, weight, and threshold rationale
• Demonstrate dashboard navigation and account drill-down
• Review intervention playbooks and expected actions
• Practice with sample accounts (what would you do for this Red account?)
• Create quick-reference guide for ongoing use
14. Hand Off to Client2.5hTransfer ownership and establish ongoing governance process. End state: Client self-sufficient with admin access and quarterly review process defined.

• Transfer admin access to CS Operations or RevOps owner
• Deliver configuration documentation (metrics, weights, thresholds, data sources)
• Establish quarterly health score review cadence to assess accuracy
• Define process for requesting metric or threshold changes
• Schedule 30-day check-in to review adoption and address questions
• Close out project with success metrics documentation

Summary

  • Total Task Lists: 2 (consolidated from 5 Parts)
  • Total Tasks: 14 (one per Step)
  • Total Hours: 30h
  • Generated from: playbook_customer-health-model.md
  • Generated on: 2025-12-31