How Health Scores Work

Last Updated: August 21, 2026 Estimated Reading Time: 7 minutes

Overview

Every Configuration Item (CI) in cmdbx receives a health score from 0-100. This score helps you prioritize CMDB improvements and track data quality over time.

Health Score Ranges

| Range | Status | Color | Action Needed | |-------|--------|-------|---------------| | 90-100 | Excellent | Green | No action needed | | 75-89 | Good | Olive green | Minor improvements recommended | | 50-74 | Fair | Yellow | Should be addressed soon | | 25-49 | Poor | Orange | Requires attention | | 0-24 | Critical | Red | Immediate action needed |

How cmdbx Calculates Health Scores

cmdbx calculates each health score from 6 weighted components:

Every stored score records its rule-pack version. Production uses health-score v3, while the code retains health-score v2 as a safe default for deployments where the server-owned v3 rollout switch is not enabled. Tenant requests cannot select a rule pack.

Health-score v3 uses deterministic scoring evidence. AI relationship findings remain advisory and do not change component scores or authoritative issue counts.

1. Data Completeness (25% weight)

Measures how many required fields are filled for the CI's class.

Example: A server with no IP address or FQDN, no environment, or no owner will score lower.

Required fields vary by CI class:

  • All CIs: Name, class, and ServiceNow sys_id
  • Servers: Owner, operational status, environment, and an IP address or FQDN
  • Applications and services: Version, environment, and owner
  • Databases: Version, owner, and operational status
  • Network CIs: Owner, operational status, and an IP address or FQDN

How to improve:

  • Fill in missing required fields in ServiceNow
  • Use bulk update tools for common patterns
  • Set defaults for new CIs

2. Relationship Health (20% weight)

Evaluates the quality and validity of CI relationships.

Example: An application with no "Runs on" relationship to a server is likely incomplete.

Common relationship issues:

  • Orphaned CIs: No relationships at all
  • Applications without a parent dependency
  • Servers without a downstream dependency
  • Too few relationships for the CI's class

How to improve:

  • Connect isolated CIs to their verified dependencies
  • Review whether relationship direction matches the real dependency
  • Use CSDM Workbench to model services correctly
  • Use the CSDM Validator for deeper relationship-pattern checks

This component measures topology coverage. It does not currently detect duplicate relationships or validate every relationship type.

3. Data Freshness (20% weight)

Checks when the CI record was last updated.

Freshness is based on the last ServiceNow update. Thresholds vary by CI class, and the score degrades progressively after the class's stale threshold rather than jumping between four fixed buckets.

Example: A server not updated in 6 months may have outdated information.

How to improve:

  • Verify that automated discovery and integrations are still updating the CI
  • Review whether the CI still exists and should remain active
  • Correct discovery scope, credentials, or reconciliation when source updates have stopped

Any edit can change sys_updated_on, so this component measures record recency and does not prove that Discovery recently verified the CI.

4. Ownership Clarity (15% weight)

Checks whether the CI has a resolvable owner and support contact. CIs with both values plus an assignment group or manager receive a small bonus within this component.

How to improve:

  • Assign an accountable owner
  • Set the support contact or group
  • Populate the assignment group or managed-by field

5. CSDM Compliance (10% weight)

Checks a limited set of CSDM-related class, lifecycle, environment, naming, and class-specific data hygiene rules. Detailed relationship-model validation remains in the CSDM Validator, so a score of 100 does not certify full CSDM compliance.

For services identified through ServiceNow's cmdb_ci_service hierarchy, health rule pack v2 checks the CSDM lifecycle pair:

  • Both lifecycle stage and lifecycle status present: complete
  • Exactly one present: incomplete-pair penalty
  • Both missing: missing-pair penalty

The score does not reject a complete pair based on its wording. ServiceNow may validly configure stages such as Design or End of life for a service class. Other CIs continue to use the base operational and install status checks in this component.

Example: A service with lifecycle stage "Operational" but no lifecycle status is incomplete and should have the missing status filled in ServiceNow.

How to improve:

  • Use the CSDM Workbench to model services correctly
  • Review CSDM relationship rules documentation
  • Restructure non-compliant service hierarchies
  • Run CSDM validation reports

6. Operational Status (10% weight)

Considers the CI's current operational and installation state. This remains separate from CSDM lifecycle completeness: a service can have a complete lifecycle pair while still being unavailable.

The component starts at 100. A missing operational status subtracts 40 points, an unknown operational status subtracts 25 points, a non-operational status subtracts 50 points, and a retired operational status subtracts 70 points. Installation-state penalties range from 20 points for installing or pending installation to 60 points for retired, stolen, or absent CIs. Operational and installation penalties can accumulate, and the result cannot fall below zero.

This component does not measure uptime, incidents, change performance, or the separate Operational Stability signal.

Example: A CI marked as "In Maintenance" for 12 months may actually be retired.

How to improve:

  • Update operational status regularly
  • Retire decommissioned CIs instead of leaving them inactive
  • Mark production CIs with appropriate criticality
  • Clean up "Unknown" status CIs

Criticality and the overall rollup

After the six weighted components are combined, the individual result is rounded. Under production health-score v3, business criticality does not change an individual CI health score. Criticality instead controls how much influence the CI has on the tenant-level rollup and remediation ordering. AI relationship analysis is advisory and never changes the score.

Tenant rollup weights:

  • Critical: 1.2
  • Important: 1.1
  • Standard: 1.0
  • Low: 0.9

Example: Critical production CIs should have higher data quality than dev/test systems.

How to improve:

  • Set appropriate criticality levels for all CIs
  • Ensure critical CIs have complete data (100% fields filled)
  • Prioritize improvements for business-critical systems
  • Review criticality during change management

How to View Health Scores

Dashboard View

The Health Dashboard shows:

  • Overall CMDB health score (criticality-weighted average of scored CIs)
  • Health distribution (count of CIs in each range)
  • Health trend (12-week historical chart)
  • Top issues (most common problems across CIs)

CI Health Workspace View

When you browse CIs:

  • Each CI displays its health score as a colored badge
  • Click a CI to see detailed health breakdown
  • Filter by health range using the status cards
  • Sort by health score to find worst offenders

Individual CI View

Click any CI to see:

  • Overall health score (large number)
  • Breakdown by component (6 components with percentages)
  • Specific issues found (for example, missing fields or stale data)
  • AI-generated remediation recommendations
  • Health history over time (trend chart)

How to Improve Health Scores

Quick Wins (30 minutes)

Focus on these high-impact, low-effort improvements:

  1. Fix isolated CIs: Add verified dependency relationships where they are genuinely missing
  2. Fill missing fields: Update IP addresses, environments, criticality
  3. Update stale records: Review CIs not updated in 90+ days
  4. Set operational status: Mark CIs as Operational, Retired, or Maintenance

Medium Effort (1-2 days)

Tackle these for sustained improvement:

  1. CSDM restructuring: Use CSDM Workbench to model services correctly
  2. Relationship cleanup: Remove invalid or duplicate relationships
  3. Bulk updates: Use AI Remediation to generate improvement plans
  4. Criticality review: Ensure all production CIs have appropriate criticality

Long-Term (Ongoing)

Implement these processes for continuous improvement:

  1. Automated discovery: Set up ServiceNow Discovery to keep data fresh
  2. Process integration: Update CMDB during change/incident workflows
  3. Quarterly audits: Review and improve health scores regularly
  4. Onboarding process: Ensure new CIs are created with complete data

Use the AI Remediation feature in the CI Health Workspace to generate bulk improvement plans. It analyzes all filtered CIs and provides prioritized recommendations.

Frequently Asked Questions

Q: Why did my health score drop?

A: Health scores can drop if:

  • Data becomes stale (not updated in 90+ days)
  • Relationships are removed or invalidated
  • Required fields are cleared
  • Sync brought in new CIs with missing data
  • CSDM compliance issues are detected

Check the health trend chart to see when the drop occurred, then review recent changes.

Q: How often are health scores recalculated?

A: cmdbx recalculates health scores at these times:

  • Automatically during the daily sync (default)
  • Manually via "Recalculate Health" button in CI Health Workspace
  • After changes in CSDM Workbench
  • On-demand via CI Health Workspace

Q: What is a "good" overall CMDB health score?

A: Target scores by maturity:

| Score | Assessment | Recommendation | |-------|------------|----------------| | 90-100 | Excellent | Maintain current controls and monitor for regression | | 75-89 | Good | Focus on the lowest component and recurring failures | | 50-74 | Fair | Prioritize the largest evidence-backed gaps | | 25-49 | Poor | Establish a governed improvement plan | | 0-24 | Critical | Address systemic data quality and ownership failures |

Q: Can I customize health scoring weights?

A: Not currently. Health scoring uses industry best practices and ServiceNow CSDM standards. If you have specific needs, contact [email protected] to discuss custom scoring.

Q: Why do some CIs have N/A scores?

A: CIs may show N/A when they are newly synced and have not yet been scored, when a health calculation is still running, or when the latest calculation failed before a score was stored.

Wait for the next sync, or start the recalculation manually.

Q: How does health scoring compare to ServiceNow's built-in scoring?

A: The two scores use different models and should not be expected to match:

| Feature | cmdbx | ServiceNow CMDB Health | |---------|-------|------------------------| | Overall score model | Six weighted CI components | Completeness, Correctness, and Compliance KPIs | | Completeness | CMDBx class requirements and approved customer mappings | Mandatory and recommended ServiceNow fields | | Staleness | A separate class-sensitive Freshness component | A Correctness metric using configurable staleness rules | | Relationships | Class-sensitive topology coverage | Orphan, duplicate, stale, suggested, containment, and hosting relationship tests | | Compliance | Limited CSDM hygiene in this score, with deeper checks in the CSDM Validator | Template and scripted compliance audits | | Ownership and operational state | Explicit weighted components | Can be represented through configured fields, rules, and audits rather than default top-level KPIs |

CMDBx complements ServiceNow CMDB Health by providing a consistent cross-environment prioritization model. It does not reproduce ServiceNow's native KPI calculation.

Related Articles

Need Help?

Contact [email protected] or use the AI Assistant to ask questions about your health scores.