Choosing the Right Chart: A No-Nonsense Guide

The wrong chart doesn’t just look bad — it actively misleads. Choosing the right visualisation is one of the most underrated skills in data communication. Here’s a practical guide to matching your data to the chart it deserves.

The Big Four and When to Use Them

Most data stories can be told with four chart types. Everything else is a special case.

• Bar chart — comparing values across categories. Simple, honest, works almost everywhere.
• Line chart — showing change over time. If there’s a time axis, start here.
• Scatter plot — revealing relationships between two variables. Great for spotting correlations (and correlation ≠ causation, don’t forget).
• Table — yes, a table. When people need to look up exact numbers, a chart is the wrong tool.

When to Bin the Pie Chart

Pie charts have one job: showing part-to-whole relationships with a small number of clearly distinct slices (ideally 2–4). The moment you have five or more slices, you’ve created a visual puzzle where the reader has to work hard to extract information that a simple bar chart would deliver instantly.

The rule of thumb: if you need a legend to decode your chart, reconsider the chart type.

The Question to Ask First

Before picking a chart, ask: what comparison am I making? Comparison over time? Line. Comparison between groups? Bar. Composition of a whole? Maybe a stacked bar. Relationship between variables? Scatter. Answer the comparison question first and the chart choice usually follows naturally.