Stop 05 · Data Analysis for Business AnalystsTopics · 4/10

Data Visualization Basics

In short

The same numbers can confuse a room or convince it in five seconds. It mostly comes down to picking the right chart and removing everything that doesn't help the point.

Choosing the right chart type and presenting data clearly — a BA's analysis is only as useful as a stakeholder's ability to quickly understand what it shows.

The purple blob in a kitchen chooses between a frying pan, a pressure cooker and a tall pot.

Think of it like this

The right utensil for the dish

You fry in a kadhai, make dal in a pressure cooker and boil milk in a tall pot. Use the wrong one and it still sort of works, but it's messy. Charts are utensils: each one is built for a specific kind of message.

Watch it happen

Pick a chart by what you want to show

What to show?

Change over time

Compare items

Part of whole

Relationship

Spread

Line chart

Bar chart

Stacked bar

Scatter plot

Histogram

1/5Decide the message first. The chart type follows from it.

The big ideas

What you need to know

Title says the finding

'Pune refunds doubled in March' beats 'Refunds by city'. Readers should get the point from the title alone.

Remove clutter

Drop 3D effects, heavy gridlines and extra colours. Every ink mark should help the message.

Colour with purpose

Use grey for context and one strong colour for the thing you want people to notice.

Honest axes

Bar charts should start at zero. A cut-off axis can make a tiny difference look huge.

Save yourself the pain

Common mistakes

Oops

Using a pie chart with nine slices.

Do this instead

Use a sorted bar chart. People compare bar lengths much more easily than slice angles.

Oops

Making every bar a different colour.

Do this instead

Keep bars one colour and highlight only the one that matters to your point.

Your mission

Fix a bad chart

~25 min

Quick check

You want to show how monthly active users changed over the last 12 months. Which chart fits best?

Go deeper

1Build 3 different chart types from the same dataset and note which communicates best.
2Learn why a cluttered chart with too many data series often obscures the actual insight.
3Practice writing a one-line takeaway to accompany each chart you build.