visual design
Information visualisation
What is information visualisation?
Information visualisation is the practice of turning data into visual form - charts, graphs, maps, diagrams - so that patterns, comparisons and trends become visible at a glance. It works by mapping numbers to visual properties the eye reads instantly (position, length, colour, size), letting people see in a moment what a table of figures would take minutes to work out.
Also known as: information visualization, information visualisation, data visualisation
Prefer to watch?
Watch the recap 0:27
The demo
Twelve months of sales - the same numbers, shown two ways. Find the trend and the odd month out. First in a table, then as a chart. Time yourself: one of these is a lot faster.
What this demo shows (text version)
The same twelve months of sales data is shown two ways. As a table of numbers it's precise, but the overall trend and the one unusual month are buried - you'd have to read every value and compare them in your head. As a chart, the same data maps each month to a position and height, so the rising trend and the dip stand out instantly, without any calculation.
That's information visualisation: turning data into visual form so the eye can perceive patterns, comparisons and outliers directly. The craft is choosing the right chart for the question and keeping it honest - sensible axes, clear labels, no distortion or chart-junk - because the same power that reveals a real pattern can just as easily manufacture a false one.
A good visualisation turns reading-and-calculating into seeing. The eye is brilliant at spotting patterns in position and length, so plotting data lets people grasp a trend, an outlier or a comparison instantly that a table hides in rows of numbers. The craft is choosing the right form for the question (trend over time → line; comparison → bar; part-to-whole → carefully; relationship → scatter) and being honest: truncated axes, misleading scales and chart-junk distort the truth as easily as they reveal it. Match the chart to the insight, keep it clean and clear, label it well, and never let the visual exaggerate or hide what the data actually says.
In the table, the pattern was buried - you'd have to read every number and do the maths in your head. As a chart, the same data shows its trend (and that one outlier) at a glance. That's information visualisation: mapping numbers to things the eye reads instantly. Just pick the right chart for the question, and keep it honest - the same power that reveals a pattern can fake one.
Visualisation works because human perception decodes some visual properties almost instantly and pre-attentively - position along a common scale and length are read most accurately, followed by angle, area and colour. Mapping data to those properties lets people perceive trends, outliers, clusters and comparisons directly, offloading work from slow conscious calculation to fast visual perception. A table stores values precisely; a chart reveals relationships.
Choosing the right form is the core skill, driven by the question: trends over time call for line charts; comparisons across categories for bar charts; relationships between two variables for scatter plots; part-to-whole for stacked bars or (cautiously) pie charts; geographic data for maps. The wrong chart can hide the very insight it should reveal, and over-complex visuals overwhelm rather than clarify.
Honesty and clarity are non-negotiable. Truncated or inconsistent axes, distorted aspect ratios, dual axes and 3D effects can manufacture or hide patterns; "chart-junk" and decoration obscure the data. Good practice (in the spirit of Tufte) maximises the data-ink, starts axes sensibly, labels clearly, uses colour purposefully and accessibly, and reduces clutter - so the visualisation reveals what the data genuinely says, quickly and truthfully, rather than dressing it up or misleading.