Module 8 · Data Analytics Foundations
Business intelligence
How BI turns one-off analysis into dashboards people use every week, and the pipeline behind them.
About 25 minutes
The problem
Every Monday, Kolanut's sales manager asks for the same numbers: last week's revenue, revenue by region, overdue customers. Every Monday an analyst spends two hours rebuilding the same spreadsheet. The numbers are useful; the process is wasteful, and each rebuild risks a new mistake.
Business intelligence (BI) fixes this: build the analysis once, connect it to the data, and let it refresh.
The concept
Analytics vs BI. The words overlap, but a useful distinction:
- Analytics answers a question, often a new one: why did North West fall?
- BI monitors the business with the same questions, repeatedly: how is each region doing this week?
Good analysis often becomes BI. Once you know North West matters, you put it on the dashboard.
The BI pipeline
Source systems → Extract, transform, load (ETL) → Data warehouse / model → Reports & dashboards
(orders, finance, (clean, combine, calculate) (one trusted version) (Power BI, Tableau,
HR, CRM) Looker Studio)- Sources: the systems where work happens.
- ETL: copying data out, cleaning it and shaping it. In Power BI this is Power Query.
- Model / warehouse: clean tables with relationships and agreed calculations, a single source of truth.
- Dashboards: the views people look at, refreshed on a schedule.
Dashboards vs reports. A dashboard is a one-screen summary of KPIs for monitoring. A report goes deeper, with several pages and details to explore. Most BI tools produce both.
Common BI tools: Microsoft Power BI, Tableau, Google Looker Studio, Qlik. Power BI is widely used in Nigerian companies because many already pay for Microsoft 365; it's covered in its own course here.
Example
A sensible first dashboard for Kolanut's sales manager:
| Area | Visual | Why |
|---|---|---|
| Top row | 4 KPI cards: revenue this month, vs same month last year, active customers, average order line | Answers "are we OK?" in two seconds |
| Middle | Line chart of monthly revenue, this year vs last | Shows the trend and season |
| Middle | Bar chart of revenue by region, sorted, with growth % | Shows where to look |
| Bottom | Table of customers whose orders dropped most vs last quarter | Tells reps who to call |
| Side | Filters (slicers) for region, channel, product category | Lets each manager see their own area |
Four KPIs, three visuals, one action list. The discipline is leaving things out.
Walkthrough
When a law firm like Ashgrove Chambers builds BI for its partners, it follows the same steps:
- Agree the KPIs with the partners: open matters, hearings adjourned, invoices overdue, collection rate.
- Define each one precisely. Collection rate = paid invoices ÷ all invoices issued × 100, counted by number of invoices.
- Connect the sources: the matter management system and the billing system.
- Check the numbers against a manual calculation before anyone relies on the dashboard.
- Schedule the refresh, for example every morning at 7.
Step 4 is the one people skip. A dashboard that is wrong once loses trust for months.
Practice
Practice
Check a KPI before it goes on the dashboard. Using invoices.csv, what is Ashgrove Chambers' collection rate by number of invoices (invoices with status Paid ÷ all invoices × 100), to one decimal place?
Practice
How many of Ashgrove Chambers' invoices are Overdue?
Check your understanding
Answer every question to check.