Data visualisation style guide

This guide covers how I make charts and maps. It sets out the rules I follow, the colours and type I use, and the reasons behind each choice. You can apply it as it stands to your own work, or take it as a starting point and adapt it into a guide of your own.

Start with the principles Open the colour picker


How this guide is organised

The guide puts what to do first and why second.

Section Use it to
Principles Check if a visual is ready to publish.
Planning Write a brief, size up your audience and pick a format: a number, a chart, a map or a table.
Colour and Typography Pick colours and type.
Big numbers Show one key number at a glance.
Charts and Maps Lay out a chart or map, choose a type and keep within its limits.
Dashboards Combine numbers, charts and maps on one screen that people check regularly.
Make your own Adapt this guide to another brand or organisation.
Background Read the research behind the rules: perception, cognitive load, colour science and typography. Optional.

Four kinds of notes recur throughout:

A firm rule. Break it only when you can explain why.

A common mistake, and what to do instead.

What to change, and where to start, if you are building a guide for your own brand.

A short reason, with a link to the background reading.

The short version

If you read nothing else:

  1. One visual, one message. Write the message as the title: “Incidents fell by a third after the ceasefire”, not “Incidents per month”.
  2. Put what, where and when in the subtitle: units, categories, period and place.
  3. Pick the simplest form that shows the message. Use a big number for one fact, a horizontal bar chart for comparing categories, and a line chart for a trend. Use a map only when location is the point.
  4. Label directly. Put labels on the data, not in a legend, wherever they fit.
  5. Use colour on purpose. Use one colour, plus grey for everything that isn’t the story. Add more colours only for real categories.
  6. Never use colour alone. Add labels, position, shape or line style.
  7. Remove what doesn’t help. No borders, 3D, shadows or heavy gridlines.
  8. Check it at the size people will see it, on screen and in print.

Pre-publish checklist

  • The title states the takeaway, and the subtitle gives units, period and scope.
  • The chart or map type matches the message (chooser).
  • Axes and scales are honest: bars start at zero, and any cut axis is clearly marked.
  • Data is labelled directly, with a legend only where direct labels don’t fit.
  • Colours come from the palette, and the visual still reads in greyscale.
  • Text is the house typeface, at least 9 pt in print.
  • The source is credited, and notes explain any data limitations.
  • Someone who wasn’t involved can state the message in one sentence.