Skip to content
Topic Hub · Business Intelligence

BI tells you what happened. You still need to decide what to do.

Business intelligence (BI) is the discipline of turning raw data into reports and dashboards that describe past performance. In 2026, BI is unbundling: semantic layers (dbt, Cube), headless BI engines (GoodData, Lightdash), and operating intelligence platforms (Fairview) are replacing monolithic BI suites. BI is the description layer; operating intelligence is the decision layer above it.

§ 01 · Definition

What is business intelligence?

Business intelligence is the technology, processes, and disciplines for collecting, integrating, analyzing, and presenting business data. Traditional BI is built around dashboards and reports. Modern BI is unbundling into a stack: warehouse (Snowflake, BigQuery) → semantic layer (dbt, Cube) → presentation (Looker, Tableau, Mode) → decision layer (operating intelligence).

§ 02 · Context

Why business intelligence matters in 2026

  • 01

    BI software is a $30B+ market growing 8% annually, but dashboard adoption inside companies remains under 25% of intended users (Gartner).

  • 02

    The semantic layer is becoming the new center of gravity — defining a metric once and consuming it everywhere prevents the "every dashboard shows a different number" problem.

  • 03

    Headless BI lets product teams embed metrics into operational tools, killing the dashboard-context-switch problem.

  • 04

    For operators, traditional BI tools require an analyst translator. Operating intelligence is "BI for non-analysts."

  • 05

    The fastest-growing BI segment is embedded + operator-facing — not the monolithic-suite category.

§ 03 · Metrics

Core metrics & concepts

Every metric below has a definition page in the Fairview glossary — formulas, benchmarks, and worked examples.

Business Intelligence (BI)

Business intelligence turns raw data into reports and dashboards. It tells you what happened — operating intel

Business Intelligence vs Operating Intelligence

BI tells you what happened — dashboards, queries, charts. OI tells you what to do next — risks, opportunities,

Operating Intelligence

A category of software that connects CRM, finance, and marketing data into a single operating view and surface

Data Warehouse

A centralized storage system that collects, structures, and stores data from multiple business systems (CRM, E

Data Lake

Data lake = centralised raw-data repository on cheap object storage (S3, GCS), schema-on-read. Dominant 2010–2

Data Lakehouse

Data lakehouse = lake-class storage (Parquet on object storage) + warehouse-class properties (ACID, schema, in

Data Mart

Data mart = subject-area subset of analytical data (sales/finance/marketing) modelled for one team's reporting

Semantic Layer

A translation layer that sits between a data warehouse and reporting tools, defining business metrics (revenue

Metric Store

Metric store = centralised metric definitions exposed via API to any consumer. Largely synonymous with headles

Metric Layer

Metric layer = architectural layer for centralised metric definitions. Synonymous with metric store, semantic

Headless BI

Headless BI = decoupled metric semantic layer that any consumer (dashboards, AI tools, reverse-ETL) can query

Embedded Analytics

Analytics capabilities built directly into a software product's interface, so users access dashboards, reports

KPI Dashboard

A visual display that shows an organization's key performance indicators in real time, combining metrics, tren

Operating Dashboard

A single-screen view that aggregates revenue, margin, pipeline, and forecast data from multiple business syste

Self-Serve Analytics

A data access model where non-technical users (operators, managers, executives) can explore, query, and visual

Data Product

Data product = dataset treated as a managed product (owner, consumers, SLAs, versioning, lifecycle). Discovera

Data Catalog

Data catalog = searchable inventory of data assets with metadata, ownership, documentation, classification, li

Data Lineage

Data lineage = documented dependency graph of analytical data. Levels: table-level (most common), column-level

Data Governance

Data governance = policies + tooling for responsible data management (quality, security, privacy, access, rete

Data Normalization

The process of cleaning and standardizing data from multiple sources so it can be compared and analyzed togeth

Single Source of Truth

SSoT = every critical metric defined and calculated in one canonical place all teams reference. When marketing

Metric Tree

Metric tree = hierarchical decomposition of a top-level metric (revenue, NRR, CM) into driving sub-metrics, do

Connected Data

Data from multiple business systems — CRM, finance, e-commerce, and marketing — unified into a single normaliz

ELT (Extract, Load, Transform)

ELT = Extract → Load → Transform (in warehouse). Modern default for cloud analytical workloads. Won because wa

ETL (Extract, Transform, Load)

ETL = Extract → Transform (in staging) → Load. Dominant 1990s–early 2010s when warehouse compute was expensive

Reverse ETL

Reverse ETL = warehouse → operational systems sync. Closes the loop after ELT pulls data inbound. Use cases: C

CDC (Change Data Capture)

CDC = read database transaction logs (Postgres WAL, MySQL binlog) to capture inserts/updates/deletes increment

Star Schema

Star schema = central fact table + denormalised dimensions in star layout. Dominant analytical pattern (Kimbal

Snowflake Schema

Snowflake schema = normalised dimensional model with dimension sub-tables. Storage savings vs query complexity

Dimensional Modeling

Dimensional modeling = analytical-database design around facts (events + measures) and dimensions (context). K

§ 05 · Pillar guides

The definitive guides

Long-form references on the core jobs — written for operators, not analysts. Updated 2026.

§ 11 · FAQ

Frequently asked

What is business intelligence in simple terms?

Software and processes that turn raw data into dashboards and reports describing past performance. Examples: Looker, Tableau, Power BI, Metabase.

How is BI different from operating intelligence?

BI describes what happened. Operating intelligence prescribes what to do. BI is a window onto data; operating intelligence is a steering wheel for decisions.

What is the semantic layer?

A translation layer that defines business metrics (revenue, margin, churn) once and exposes them consistently to every downstream tool. Prevents the "every dashboard shows a different number" problem.

Do I need a data warehouse for BI?

For modern BI: yes. Snowflake, BigQuery, or Redshift acts as the central data store. Modern BI tools (and operating intelligence platforms) connect to the warehouse rather than to source systems directly.

What is headless BI?

BI architecture where the metric layer and the query engine are decoupled from the visualization layer. Lets product teams embed metrics into operational tools without rebuilding the underlying logic.

Stop reading about business intelligence. Start running on it.

Connect your stack. See business intelligence in your data within 24 hours. No credit card required.

Editorial standards

Sources & references

Fairview maintains a public bibliography for every topic hub. Each citation below was verified at publication. We update sources every 12 months as new benchmark studies are released. See our editorial standards.

  1. 1 Magic Quadrant for Analytics and Business Intelligence — Gartner, 2025. View source .
  2. 2 The State of Analytics Engineering — dbt Labs, 2025. View source .
  3. 3 Headless BI: The Future of Embedded Analytics — GoodData Research, 2024. View source .

Fairview cites primary sources only — government data, academic research, industry benchmarks from named publishers, and official vendor documentation. See our editorial standards.