Start here: what data analysts do and your roadmap from zero to job-ready
A data analyst turns raw numbers into answers that help people decide. A supermarket asks "Which branch is losing money and why?" A county asks "Which wards have the lowest clinic attendance?" A SACCO asks "Which members are likely to default?" The analyst finds, cleans and studies the data, then explains the answer clearly.
What a data analyst does every day
- Talks to managers to understand the question
- Finds and collects data (spreadsheets, databases, surveys, systems like M-Pesa statements or POS exports)
- Cleans it: fixes errors, duplicates and missing values
- Analyses it: totals, trends, comparisons, statistics
- Visualises it: charts and dashboards
- Presents findings and recommendations
The six-step analysis process
| Step | Question to ask | Example (a chain of shops) |
|---|---|---|
| 1. Ask | What decision needs to be made? | "Should we open longer hours on Saturdays?" |
| 2. Prepare | What data do we need, and where is it? | Hourly sales from the POS for 6 months |
| 3. Process | Is the data clean and complete? | Remove test sales, fix missing branch names |
| 4. Analyse | What does the data say? | Saturday 6–9pm sales are 18% of weekly revenue |
| 5. Share | How do we explain it? | A one-page report with two charts |
| 6. Act | What do we do now? | Pilot longer hours in two branches, measure results |
Most beginners rush to step 4. The best analysts spend real time on steps 1 and 3.
The tools you'll learn (in this order)
| Tool | What it's for | Where to learn |
|---|---|---|
| Excel / Google Sheets | Quick analysis, PivotTables, charts; used in almost every office | This subject + Excel |
| SQL | Getting data out of databases | This subject + SQL |
| Statistics | Knowing what the numbers really mean | Lessons 3–7 of this subject |
| Power BI (or Looker Studio) | Dashboards for managers | Lesson 13 |
| Python | Bigger data, automation, advanced analysis | This subject + Python |
Your roadmap (about 4–6 months part-time)
- Month 1: Excel basics, PivotTables, charts; the statistics lessons here.
- Month 2: SQL: SELECT, WHERE, GROUP BY, JOINs.
- Month 3: Data cleaning, visualisation and Power BI.
- Month 4: Python basics for data (lists, dictionaries, CSV files; then pandas).
- Months 5–6: Two or three portfolio projects with real Kenyan data, a polished CV and LinkedIn, applications or freelance gigs.
See the full data analyst career roadmap and the data analysis portfolio project.
Your first analysis (run it)
Here are a week of sales from a small shop. Click Run, then try changing the numbers.
sales = {"Mon": 4200, "Tue": 3900, "Wed": 4100, "Thu": 4600, "Fri": 6800, "Sat": 9100, "Sun": 5200}
total = sum(sales.values())
best_day = max(sales, key=sales.get)
average = total / len(sales)
print("Total for the week: KSh", total)
print("Average per day: KSh", round(average))
print("Best day:", best_day, "with KSh", sales[best_day])
print("Weekend share:", round(100 * (sales["Sat"] + sales["Sun"]) / total), "%")That's analysis: a question (how did we do?), data, calculations and an insight (weekends matter a lot).
Think about it: A manager says: "Give me a report on sales." What should you do before opening Excel?Show answer
Ask questions first: What decision is the report for? Which period, branches and products? Compared to what (last month, last year, target)? Who will read it, and how much detail do they want? A clear question saves hours and makes the report useful.
Summary
- Data analysts turn data into answers and recommendations.
- Follow the process: ask, prepare, process, analyse, share, act.
- Learn Excel, SQL, statistics, Power BI and Python, in that order.
- Build real projects; they matter more than certificates alone.
Check yourself
What is the first step of the analysis process? (one word)
Show answer
ask
Which tool gets data out of databases? (abbreviation)
Show answer
sql
In the example, which day had the highest sales? (three letters)
Show answer
sat