Data Analysis

BI, KPI, and Dashboard Terms

Learn how metrics, KPIs, dimensions, measures, star schemas, semantic models, dashboards, filters, and refreshes support decisions.

12 matching terms

BI concepts

Business intelligence (BI)

Meaning

Processes and systems that turn organizational data into information for decisions.

When to use it

Use it to deliver governed reports, analysis, metrics, and dashboards.

Practical example

BI combines sales, finance, and operations data into one decision view.

Metrics

Key performance indicator (KPI)

Meaning

A metric selected to show progress toward a specific strategic or operational objective.

When to use it

Define the objective, formula, owner, target, frequency, and response action.

Practical example

KPI: on-time delivery rate; target >= 96%; owner: logistics.

Metrics

Metric

Meaning

A quantitatively defined measure used to monitor or compare performance.

When to use it

Maintain one documented formula and grain across reports.

Practical example

conversion_rate = completed_orders / checkout_sessions

Data model

Dimension

Meaning

A descriptive attribute used to filter, group, or label measures.

When to use it

Use dimensions such as date, product, customer, region, or channel for analysis.

Practical example

Analyze revenue by month, region, and product category.

Metrics

Measure

Meaning

A numeric calculation evaluated in an analytical context, usually through aggregation.

When to use it

Centralize reusable business calculations in the semantic model.

Practical example

Net Revenue = SUM(Sales[Revenue]) - SUM(Sales[Refund])

Data model

Fact table

Meaning

A table that stores events or observations together with dimension keys and measures.

When to use it

Declare its grain before adding measures or relationships.

Practical example

FactSales grain: one row per order line.

Data model

Dimension table

Meaning

A table that describes business entities used to filter and group facts.

When to use it

Use stable keys and user-friendly attributes for dates, products, customers, and locations.

Practical example

DimProduct(product_key, product_name, category, brand)

Data model

Star schema

Meaning

A dimensional model with a central fact table connected to surrounding dimension tables.

When to use it

Use it to improve analytical clarity, filtering behavior, and model usability.

Practical example

FactSales connects to DimDate, DimProduct, DimCustomer, and DimStore.

Data model

Semantic model

Meaning

A governed analytical layer that defines relationships, measures, hierarchies, and business meaning.

When to use it

Use it to keep definitions consistent across reports and users.

Practical example

Publish certified Revenue, Margin, and Active Customer measures once.

Delivery

Dashboard

Meaning

A focused visual view of important metrics and status for monitoring and decisions.

When to use it

Design it around decisions, exceptions, trends, and clear ownership.

Practical example

Executive dashboard: revenue, margin, forecast gap, and top exceptions.

Interaction

Filter context

Meaning

The active set of filters that determines which data contributes to a calculation.

When to use it

Use it to explain why the same measure changes by slicer, row, or visual.

Practical example

Revenue is recalculated for Region = Seoul and Year = 2026.

Delivery

Data refresh

Meaning

The process that updates an analytical model or report with newer source data.

When to use it

Define frequency, latency expectations, failure alerts, and data completeness checks.

Practical example

Refresh hourly; alert if source data is more than 90 minutes old.