Data Analysis Is Not “Making Charts”: A Beginner's End-to-End Workflow

If someone tells me:

“I'm learning data analysis.”

My next question is usually:

“What have you analyzed?"

Not:

“Which course are you taking?”

Not:

“Which certificate are you getting?”

Not even:

“Which tool are you learning?”

What did you analyze?

Because data analysis becomes much easier to understand when you stop treating it as a collection of tools and start treating it as a workflow.

Here's a realistic beginner workflow.

1. Start with an ugly dataset.

Don't start with a perfectly formatted Kaggle dataset.

Those are useful, but real-world data is often messy.

Imagine:

Date       Product       City       Sales
01/09/26 Laptop        Awka       450000
02/09/26 laptop Awka 450000
03/09/26 Phone         Enugu
04/09/26 Phone         Enugu      180000
05/09/26 LAPTOP Awka 450000

Immediately, you have problems.

Laptop, laptop, and LAPTOP may represent the same thing.

One sales value is missing.

Dates may not be in the format you eventually want.

This is where analysis actually begins.

1. Clean before you visualize.

A common beginner mistake is jumping straight into Power BI and creating colorful charts.

Don't.

First ask:

Are the columns correct?
Are there duplicates?
Are values missing?
Are categories consistent?
Are dates valid?
Are numerical fields actually numerical?

Cleaning is not the glamorous part.

It is one of the most important parts.

1. Ask a business question

Don't open Power BI and ask:

“What dashboard can I make?”

Start with:

“What do I need to know?”

For example:

Q1: Which product generates the most revenue?

Q2: Which city has the highest sales?

Q3: What month performs best?

Q4: Is revenue increasing?

Q5: Which products are declining?

Now your analysis has direction.

1. Use SQL when the dataset grows.

Suppose the data is in a database.

You might eventually write something like

SELECT
city,
SUM(sales) AS total_sales
FROM transactions
GROUP BY city
ORDER BY total_sales DESC;

Now you're no longer manually scrolling through thousands of rows.

You're asking the database a question.

That is where SQL becomes extremely useful.

1. Visualize the answer.

Once you've analyzed the data, you might use Power BI or another visualization tool.

For example:

Revenue by City

Awka       █████████████
Enugu █████████
Aba        ███████
Owerri █████

But remember:

The chart isn't the analysis.

The chart is how you communicate the analysis.

The important part is the conclusion behind it.

1. Explain what the numbers mean.

Imagine the final result says:

“Awka generated 42% of total revenue, but Aba had the fastest month-on-month growth.”

That is useful.

A manager can act on that.

A client can ask another question.

A business owner can investigate why.

This is why data analysis is more than Excel formulas or Power BI dashboards.

It is a process:

RAW DATA
↓
CLEAN
↓
ASK
↓
ANALYSIS
↓
VISUALISE
↓
EXPLAIN
↓
DECIDE

And that workflow is becoming increasingly relevant to Nigeria's digital-skills ecosystem. NITDA's current 3MTT program lists Data Analysis & Visualization as one of its 12 technical skill areas.

So where should a beginner start?

If you're looking for data analysis training in Awka, I'd recommend thinking about your learning path in layers.

Layer 1 — Spreadsheet fundamentals

Learn:

Excel
Data cleaning
Formulas
Pivot tables
Basic charts
Layer 2—Data querying

Learn:

SQL
Filtering
Aggregation
Joins
Grouping
Layer 3—Visualization

Learn:

Power BI
Dashboard design
Data storytelling
Layer 4—Programming

Then consider:

Python
Pandas
NumPy
Matplotlib

You don't necessarily need all four layers immediately.

The important thing is to build projects while learning.

For someone in Awka who prefers structured physical learning, TEKHUB provides a technology-training environment where practical digital skills can be developed alongside project work.

You can explore the academy here:

https://tekhub.ng/

And learn more about TEKHUB's wider technology programs here:

https://tekhub.ng/

One project I'd give every beginner

Download a dataset containing sales transactions.

Then answer five questions:

1. What is the total revenue?

2. Which product sells the most?

3. Which location performs best?

4. Which month performs best?

5. What would you recommend based on the data?

Don't copy someone else's dashboard.

Build yours.

Make mistakes.

Then explain your conclusions.

If you can do that, you're no longer just “learning Excel.”

You're beginning to think like an analyst.

And that's the real skill.