Data visualisation consulting

Right data, wrong chart

Many dashboards contain accurate data presented in ways that slow people down or mislead them. Pie charts with too many slices, 3D effects that distort size, dual axes that suggest false correlations and colours that change meaning from page to page are common.

We review each chart against the question it should answer. Comparing values usually needs a bar chart. Showing change over time usually needs a line. Highlighting one figure needs colour on that figure only. These choices sound basic, but they change how quickly people understand the data.

Start from the decision

Dashboards often grow by adding every metric someone asks for. The result is a page full of numbers where nobody knows what to look at first. Users export to Excel to find the answer.

We start by asking who uses each report and which decision it supports. The main question goes at the top, with the most important comparison clearly shown. Supporting detail sits lower or on separate pages.

We also design for accessibility. Colour palettes are checked for colour-blind users, and important information never relies on colour alone.

• One main question per dashboard
• Chart types chosen for the comparison
• Colour used to highlight, not decorate
• Colour-blind safe palettes