Module 1 · Data Analytics Foundations
What is data analytics?
What analysts actually do, the four kinds of analytics, and the steps every analysis follows.
About 20 minutes
The problem
Kolanut Distribution sells drinks, snacks and household goods to shops across Nigeria. Every order is written into a system: who bought, what, how many, at what price, on which day. After eighteen months that system holds more than four thousand order lines.
The managing director doesn't want four thousand rows. She wants answers:
- Are we selling more than last year?
- Which regions are growing, and which are slipping?
- Should we keep giving wholesalers 10% discounts?
Turning the rows into those answers is data analytics.
The concept
Data analytics is the work of collecting, cleaning and examining data to answer questions and support decisions.
The word that matters is decisions. A table of numbers is not analysis. Analysis ends when someone can act: restock earlier, call a customer, drop a product, hire in one region instead of another.
Analytics questions come in four kinds, each harder than the last:
| Kind | Question it answers | Kolanut example |
|---|---|---|
| Descriptive | What happened? | Revenue in March 2025 was ₦46.3 million. |
| Diagnostic | Why did it happen? | North West revenue fell because shops there ordered less often, not because we lost them. |
| Predictive | What is likely to happen? | December 2026 sales will again be far above an average month (in 2025 they were about 45% higher). |
| Prescriptive | What should we do? | Send more stock to Lagos warehouses in November. |
Most day-to-day analyst work is descriptive and diagnostic. They are the foundation: you can't predict what you can't describe.
Example
Here is Kolanut's revenue for the first six months of 2025, rounded to millions of naira:
| Month | Revenue (₦ million) |
|---|---|
| January | 37.1 |
| February | 36.1 |
| March | 46.3 |
| April | 44.6 |
| May | 40.3 |
| June | 39.8 |
A descriptive reading: revenue ranged from ₦36.1m to ₦46.3m, and March was the best month.
A diagnostic question it raises: why were March and April stronger? (In this data, drink sales rise in the hot, dry months, and Easter shopping falls in that period.)
Walkthrough
Every analysis, big or small, follows roughly the same steps:
- Ask. Agree the question with the person who will use the answer. "How are sales?" is vague. "Did first-half revenue grow compared with last year, and in which regions?" is answerable.
- Collect. Find the data that can answer it: which system, which tables, which dates.
- Clean. Fix what would mislead you: duplicates, inconsistent spellings, missing values.
- Analyse. Summarise, compare, look for patterns and exceptions.
- Share. Present the finding so the audience understands it in a minute: a clear chart, a short summary.
- Act. Someone makes a decision, and you check later whether it worked.
The rest of this course takes each step in turn. By the end you'll run the whole cycle yourself on real-looking company data.
Practice
Practice
Using the table in the Example, which month in the first half of 2025 had the lowest revenue?
Practice
How many million naira separate the best month from the worst month in that table? Give the answer in millions, for example 3.5.
Check your understanding
Answer every question to check.