What is Business Intelligence (BI)?
Business intelligence is the set of practices and tools that turn an organisation’s raw operational data into reporting that people can act on. It covers the movement of data out of source systems, its consolidation into a consistent structure, and its presentation as dashboards, reports and defined metrics.
BI is usually descriptive rather than predictive. It answers what happened and what is happening now, in a form that is consistent enough to be compared week to week. Forecasting and modelling sit alongside it but are a different discipline.
Key Takeaways
- BI answers what happened and what is happening; it is descriptive by design.
- Its hardest problem is definitional, not technical: agreeing what a metric means across departments.
- A dashboard nobody has agreed on produces argument rather than decisions.
- BI is the layer most analytics and data science work depends on for clean, shared inputs.
Understanding Business intelligence
The technical shape of BI is a pipeline. Data is extracted from the systems that generate it, reshaped into a consistent model, and stored somewhere designed for reading rather than transacting. Reporting is then built on that store, so that a query does not slow the systems running the business and so that everyone is reading the same numbers.
The harder half is semantic. Two departments will often mean different things by an apparently obvious word such as active customer, order or region. BI forces those definitions to be written down and applied consistently, which is why BI projects surface organisational disagreements that predate them. A metric layer that everyone accepts is the real deliverable; the charts are the visible part.
Real-World Example
A retailer’s finance team reports monthly revenue from the accounting system while the trading team reports it from the ecommerce platform. The two never match, because one recognises revenue at despatch and the other at order. A BI project that only builds dashboards reproduces the disagreement in colour. One that first settles the definition, then builds to it, removes a recurring monthly argument.
Importance in Business or Economics
Without a shared reporting layer, decisions are made from whichever extract the loudest person brought to the meeting. BI makes the numbers boring, which is its purpose: an organisation that agrees on what happened can spend its attention on what to do instead.
Types or Variations
- Reporting and dashboards: Scheduled and live views of defined metrics, the most common form of BI.
- Self-service BI: Business users build their own views against a governed model, trading consistency for speed.
- Embedded BI: Reporting delivered inside another application rather than a separate tool.
- Operational BI: Near real-time reporting used to run a process rather than review it after the fact.
Related Terms
Quick Reference
- Orientation: Descriptive: what happened and what is happening
- Core deliverable: An agreed metric layer, not the charts
- Typical stack: Source systems, pipeline, warehouse, reporting tool
- Hardest problem: Consistent definitions across departments
Frequently Asked Questions
What is the difference between business intelligence and analytics?
BI is generally descriptive and standardised: agreed metrics, reported consistently. Analytics is usually investigative and answers a specific question, often with statistical methods. In practice the boundary is soft and many teams do both under one name.
Does business intelligence require a data warehouse?
Not strictly, and small organisations often report directly from source systems. A separate store becomes necessary when reporting queries would slow the operational systems, or when data from several systems has to be combined consistently.
Why do dashboards go unused?
Usually because they answer a question nobody is accountable for, or because their numbers are not trusted. A dashboard earns use when someone has to defend a decision against it.