Module 6 · Data Analytics Foundations
Data analysis
The handful of calculations behind most business analysis - totals, averages, shares and growth - and the traps in each.
About 30 minutes
The problem
Kolanut's data is clean. Now the managing director asks: "Did we do better in the first half of 2026 than the first half of 2025, and where did the money come from?"
Answering needs only a few calculations, used carefully. Most business analysis is built from the same small toolkit.
The concept
| Calculation | What it tells you | Formula |
|---|---|---|
| Total | How much, overall | Add the values |
| Count | How many | Count the rows (or distinct items) |
| Average (mean) | A typical value | Total ÷ count |
| Median | The middle value when sorted | Half the values are above, half below |
| Share of total | How much one part contributes | Part ÷ total × 100 |
| Growth rate | How much something changed | (New − old) ÷ old × 100 |
Mean or median? The mean is pulled towards extreme values; the median isn't. When a few values are much larger than the rest, as with salaries or order sizes, the median is often the fairer "typical" figure. Report both when they differ a lot.
Compare like with like. Compare January with January, the first half of one year with the first half of the next. Comparing December with January mostly shows the Christmas season, not performance.
Revenue at Kolanut is calculated per order line as:
revenue = quantity × unit_price × (1 − discount_pct ÷ 100)
Example
Revenue by region, first half (January–June) of each year, in millions of naira:
| Region | H1 2025 | H1 2026 |
|---|---|---|
| Lagos | 118.2 | 152.8 |
| South West | 28.9 | 51.7 |
| North Central | 25.3 | 27.2 |
| South South | 19.7 | 23.1 |
| South East | 20.9 | 19.4 |
| North West | 31.1 | 16.6 |
| Total | 244.2 | 290.7 |
Growth. (290.7 − 244.2) ÷ 244.2 × 100 = 19.0%. The business grew.
Share. Lagos brought in 152.8 ÷ 290.7 × 100 = 52.6% of H1 2026 revenue: more than half the business comes from one region.
The exception. North West went the other way, from 31.1 to 16.6. A total that grew 19% hides a region that nearly halved. That's why analysts always break totals down.
Walkthrough
Mean versus median on real data. Across all 4,266 of Kolanut's order lines:
- The mean order line is worth ₦194,689.
- The median order line is worth about ₦166,680.
The mean is higher because wholesalers place a smaller number of very large orders (up to ₦713,400 on a single line), which pull the average up. If a manager asks "what does a typical order line look like?", the median is the more honest answer.
Averages by group often tell the real story. Average quantity per order line:
| Channel | Average packs per line |
|---|---|
| Wholesale | 19.0 |
| Supermarket | 11.1 |
| Kiosk | 3.6 |
One overall average (13.8 packs) would describe none of these customers well.
Practice
Practice
Using the table in the Example, what share of H1 2026 revenue came from the South West, to one decimal place?
Practice
By what percentage did North West revenue fall from H1 2025 to H1 2026? Give the size of the fall as a positive number, to one decimal place.
Practice
In orders.csv, what is the average quantity per order line, to one decimal place?
Check your understanding
Answer every question to check.