What is Business Intelligence

BI classifications, ROI, business use cases, Tableau licensing tips.

What Business Intelligence is

Wikipedia’s definition, in short:

• BI is the strategies and technologies companies use for data analysis and management of business information.
• BI tools handle large amounts of structured (and sometimes unstructured) data to help organizations find and develop new strategic business opportunities.
• The aim of BI is to make large volumes of data easy to interpret.
• Finding new opportunities and acting on the insights gives a business a competitive advantage, long-term stability and better strategic decisions.

Strip away the formal wording and BI is about turning raw data into answers that help people make business decisions.

Business Intelligence, the short version

It’s a joke, but it is also the most honest definition. Every BI project is an investment, and it has to pay for itself by cutting costs or growing revenue.

What colleagues thought about BI in 2012

A 2012 mind map put Business Intelligence in the center and linked it to everything the field touched:

Branch	What it covered
Analysis	Data discovery, reporting, visualization, data mining, forecasting
Expertise	Foresight, hindsight, insight
Data governance	Availability, security, usability
Data quality	Unambiguous, true, accurate
Raw data	Information → knowledge → wisdom
Business users	Data checking, support, teaching, involvement in development
Business value	KPI monitoring, cost reduction, fixing bottlenecks, a single version of the truth
Self-Service BI	Agile BI, multiple data sources, transparency, an intuitive interface, rich analysis capabilities, flexibility

More than a decade later the list still holds. Tools have changed, but governance, data quality, business value and self-service are still what BI teams work on.

What BI looks like in practice

A typical BI architecture has three layers, connected by data pipelines:

Layer	What’s in it
Source layer	Files and SFTP, IoT devices, APIs, OLTP databases
ETL / ELT	Pipelines that extract data from sources, transform it and load it into storage
Storage layer	Data warehouse, big data solution, data lake
Access layer	Tableau Desktop / Tableau Server, Power BI, R, Python

What BI gives the business:

• Access to data for non-technical users
• Visualization and data manipulation tools
• Business terminology instead of table names (revenue, cost, product, etc.)
• Shared reports
• Report automation
• Simplicity
• A base for statistics and data science

BI tools sit in the access layer, but they are only as good as the layers underneath. Clean sources, reliable pipelines and a well-modeled warehouse come first.

BI tools classification

BI tools visualize and analyze business metrics through reports and dashboards. There are many of them, and they can be grouped by three criteria:

Criterion	Type	What it means
Cost	Open source	Free to use and modify
	Commercial	Paid, usually with vendor support and extra features
Usage model	Enterprise	Built for large companies. Needs a BI development team to install, configure and maintain
	Self-service	Easy to use. Non-technical users build reports and dashboards on their own
Deployment model	Client/server, self-hosted	Run by the company’s own IT department
	Client/server, managed	Infrastructure run by the vendor
	Cloud (SaaS)	Fully cloud-based, flexible and easy to start with

Any tool can be placed on all three axes. For example, Metabase is open source, self-service, and can be self-hosted or used as a cloud service.

ROI (Return on Investment)

When a company implements BI, it thinks about ROI. BI pays off through:

• Automation: the same work in fewer hours
• Business advice: data-backed recommendations
• Proactive actions: spotting problems and opportunities early
• Better decisions: higher quality of business solutions

If you can’t say which of these a dashboard delivers, it probably doesn’t deliver any of them.

The tool is not important

What matters is the value the tool delivers. Start from business use cases, not from the tool:

• As a sales manager, I want…
• As a category manager, I want…
• As a marketing manager, I want…
• As a CEO, I want…
• As a marketing or product analyst, I want…

Writing requirements as user stories keeps the focus on who will use the report and what decision it supports. Picking the tool comes last.

We already did BI

• In Module 1 we did BI by building a dashboard in a spreadsheet from row-level CSV data.
• In Module 2 we did BI by querying data with SQL.

Both times we answered business questions, which means we delivered business value. In this module we do the same with proper BI tools.

The idea is simple: make access to data easier and open it up to everyone in the company (data democratization).

Assignment

Try Tableau with the Superstore dataset: 02-BI-admin-with-Tableau.