Home/Docs/KPI Definitions

Docs

KPI Definitions

Canonical KPI definitions used by VOCAL scoring and reporting layers, including formulas and governance notes.

Why Canonical KPI Definitions Matter

This page is the governance reference for metric definitions and implementation alignment. It complements the public KPI glossary. Use Call Analytics KPIs for practical interpretation and this page for canonical contract logic.

KPI Governance Principles

  • - Canonical definitions are versioned and governed centrally.
  • - Signal-to-metric mapping must remain auditable.
  • - Rubric logic should align with documented interpretation guidance.
  • - Dashboard and API outputs should reflect the same KPI contract.
  • - Export schemas should remain consistent with canonical metric labels.

Metric Grouping

Conversation structure metrics
Customer signal metrics
Operational outcome metrics
Quality and governance metrics

Formula and Derivation References

  • - KPI formulas and interpretation guidance: /call-analytics-kpis.
  • - Signal and field registry relationships: analysis registry and pipeline docs.
  • - QA/compliance scoring structure: rubric-driven score components and governance rules.
  • - Export contract alignment: metric fields exposed through export/API delivery paths.

Interpretation Cautions

  • - A single threshold rarely fits all call types or business motions.
  • - Metric interpretation should include transcript and context evidence.
  • - Trend direction is often more actionable than isolated point values.
  • - Metric drift after rubric/config changes should be treated as a governance event.

Alignment with Dashboards, API, and Exports

KPI definitions should be semantically consistent across dashboard views, API payloads, and export files. If a KPI changes, governance workflows should track the version, effective date, and reporting impact before operational rollout.

Change Control and Versioning

Metric, rubric, and mapping updates should follow explicit change control with validation against historical outputs and downstream consumers. This reduces interpretation drift across teams.

Related Pages