Design elements: colour, labels and accessibility

Why this matters

Choosing the right chart type is half the job. The other half is the design around the data: the title that says what the chart shows, the labels that make it readable, the colours that separate or rank, and the removal of everything that distracts. Good design is not decoration. It decides whether a reader understands the chart at all, and whether everyone can, including people with colour vision deficiency or who use assistive technology.

The exam's visuals objective includes design elements. This lesson covers titles, labels, legends and annotations; colour palettes; accessible design; clutter and emphasis; and consistency through style guides.

The lesson

Titles, axis labels, legends and annotations

The words on a chart do as much work as the shapes.

A title should say what the chart shows, and often what it means. A descriptive title such as "Monthly orders, 2025" identifies the chart. An insight title such as "Orders held steady through 2025, with no seasonal peak" tells the reader what to see, which is usually more useful in a report or presentation.

Axis labels name each axis and its unit: "Revenue (£ thousands)", "Response time (seconds)". An unlabelled axis leaves the reader guessing whether the numbers are counts, percentages or currency. Tick labels should be readable without rotation where possible, and numbers should be formatted for reading, such as 1.2M rather than 1200000.

A legend explains colours or symbols. Where possible, label series directly, placing each series' name at the end of its line or on its bar, which saves the reader looking back and forth between the chart and a legend.

Annotations add context at the point it matters: a note marking when a price change took effect, a label on an outlier, a line showing the target. One or two well-placed annotations often explain a chart better than a paragraph beside it.

Charts should also state their source and, where relevant, the date of the data, usually in small text at the bottom.

Colour palettes: sequential, diverging and categorical

Colour carries meaning, and the right palette depends on what the data is.

A sequential palette runs from light to dark in one hue, or through a controlled range of hues, and suits values that run from low to high, such as sales by region on a choropleth. Darker means more.

A diverging palette runs from one colour through a neutral midpoint to another colour, and suits values with a meaningful centre, such as change from last year (decrease to increase) or deviation from a target. The midpoint, often zero or the target, should be the neutral colour.

A categorical (qualitative) palette uses distinct hues with no implied order, and suits unordered categories, such as product lines. It works for a limited number of categories, about six to eight, before colours become hard to tell apart.

Using the wrong kind misleads: a rainbow palette on sequential data suggests boundaries that do not exist, and a sequential palette on categories suggests an order that does not exist. Colour should also be used sparingly: when every element is coloured, none stands out.

Colour-blind safe design and contrast

Around one in twelve men and a much smaller share of women have some form of colour vision deficiency, most often difficulty distinguishing red from green. A chart that relies on red and green alone to separate "bad" from "good" fails them.

Accessible design follows a few rules:

• Use colour-blind safe palettes, which avoid problem pairings such as red against green; blue and orange is a common safe pair.
• Do not rely on colour alone: add labels, patterns, shapes or position so meaning survives without colour, and in greyscale printing.
• Provide sufficient contrast between text and background, and between adjacent colours. Accessibility guidelines such as the Web Content Accessibility Guidelines (WCAG) set minimum contrast ratios for text.
• Use readable font sizes, and avoid thin, light text.
• For digital reports, provide alternative text describing each chart's message, and a table of the underlying data where possible, so screen reader users can access the same information.

These practices improve charts for everyone, not only for readers with a particular need.

Clutter, data-ink and emphasis

Every mark on a chart should earn its place. The idea of the data-ink ratio, popularised by Edward Tufte, is that the ink used to show data should be a large share of the total, and ink that shows nothing should be removed.

Common clutter includes heavy gridlines, borders and backgrounds; 3-D effects that distort values; shadows and gradients; redundant labels that repeat the axis; and unnecessary decimal places. Removing them makes the data easier to see.

Clearing clutter makes room for emphasis: drawing the eye to what matters. Techniques include a single strong colour for the key series against grey for the rest, bold or larger text for the main number, a reference line for a target, and ordering bars so the most important comes first. Pre-attentive attributes, such as colour, size and position, are noticed before conscious attention, so using them for the one thing the reader should see first is the most effective way to make a point.

The goal is not minimalism for its own sake, but a chart where the message is the most visible thing on it.

Corporate style guides and consistency

Many organisations have a style guide for reports and dashboards: approved colours, fonts, logo placement, chart conventions and templates. Following it makes reports recognisable and professional, and it saves decisions.

Consistency also has an analytical purpose. When the same measure always has the same colour, the same region always sits in the same position, and the same scale is used across related charts, readers can compare across charts and pages without re-learning each one. Inconsistency creates false signals: if revenue is blue on one page and orange on the next, or one chart's axis starts at zero and its neighbour's does not, readers may read differences that are only design.

A practical approach is to build from templates and shared definitions, so consistency is the default rather than something each author must remember.