Metadata

In depth

Think of metadata as the label and instructions that travel with your data. It removes guesswork by answering who created the data, what it measures, where it came from, when it was last updated, and how to use it.

Common types you’ll see in analytics:

• Descriptive metadata: names, titles, summaries, tags, keywords.
• Structural metadata: how parts relate, like table joins, hierarchies, and dimensions.
• Administrative and technical metadata: owners, permissions, retention rules, data types, formats, and refresh schedules.
• Lineage and quality metadata: data source, transformation steps, version history, freshness, and quality checks.
• Semantic metadata: business definitions and rules, like the approved definition of “Active Customer.”

Strong metadata helps both humans and software parse and apply business meaning without repeated clarifications or custom code.

Pro tip

Treat metadata like product requirements. Set mandatory fields for every metric and dataset: clear name, one‑sentence description, owner, unit or currency, aggregation rule, time grain, and tags. Make these part of your workflow so nothing ships without context.

Why Metadata matters

• Faster self‑serve: people can search, filter, and trust results without pinging a data teammate.
• Consistent answers: shared definitions reduce conflicting numbers in meetings.
• Governance and security: owners, access levels, and retention notes prevent misuse.
• Auditability: lineage shows where numbers come from and which steps transformed them.
• AI‑readiness: clear structure and definitions let assistants and NLP features answer questions accurately.

Metadata - In practice

• Supply Chain: an “On-Time Delivery” metric with definition, source tables, filters, owner, and last updated timestamp helps sales and shipping reach the same count.
• Finance: revenue metrics with currency, accrual vs cash basis, and period alignment stop reconciliation churn.
• Product: usage metrics with event names, user scopes, and exclusion rules prevent double counting.

Metadata and PowerMetrics

PowerMetrics uses metadata throughout the Metric Catalog to make metrics easy to discover, trust, and share. When you create or govern metrics, you can attach and manage details such as:

• Name and description
• Owner and certification status
• Tags and categories for search
• Data source and connection details
• Aggregation method and calculation logic
• Dimensions and hierarchies (like region, plan, channel)
• Unit, currency, and formatting
• Time grain, time zone, and default date range
• Refresh schedule and last updated timestamp
• Access controls for users, groups, and roles
• Lineage context so teams understand where numbers come from

This metadata powers search, filters, goals, notifications, published views, and the PowerMetrics Knowledge Graph, so business users can assemble dashboards quickly and use consistent, trusted numbers.