Hover any icon for exact value Click a source to isolate Toggle row / column layout Arrow keys navigate
A pictogram chart represents discrete quantities as rows or columns of icons, where each icon encodes a fixed unit count. The perceptual mechanism is counting and grouping: the viewer scans a row of icons the same way they scan a shelf of objects, building a concrete intuition for quantity that abstract bar lengths cannot match. When icons represent the subject of the data — energy symbols for energy data, people silhouettes for population data — the chart exploits associative memory, reducing the cognitive distance between the number and what it means. Color encodes category, but the icon shape provides a fully redundant second encoding, satisfying WCAG's requirement that color never be the sole differentiator.
The dataset is a small set of discrete categorical proportions (five energy sources, five regions, all summing to 50 icons per region at 2% each) with a message about relative mix and regional comparison. The data is explicitly appropriate for pictograms: small counts, meaningful units, a comparison story. A bar chart would convey the same totals but strip the iconographic context that makes the energy mix legible without axis reading. A dot matrix would work for a single region but loses the side-by-side comparison that reveals which regions lead on renewables. The pictogram occupies the productive overlap between engaging infographic and honest statistical display.
A grouped bar chart would be the statistically rigorous alternative, and it is superior when precision matters more than engagement. The pictogram's failure mode is the partial icon problem: showing 3.7 icons is misleading because fractions of an icon imply fractional units. This implementation avoids partial icons entirely — all counts are rounded to whole icon multiples of the unit value before rendering. The dot matrix is the closest structural relative; the pictogram is preferred when the icon shape carries semantic meaning that helps the audience without a data-literacy background understand the subject immediately, as is common in policy briefings, annual reports, and journalism.
FT Visual Vocabulary category: Part-to-whole / Comparisons — "Show how discrete parts compare across categories when the subject can be represented iconographically." Tufte caution: pictograms are at risk of the lie factor when icons are scaled in two dimensions (area grows as the square of height). This implementation avoids that: every icon is the same size, and quantity is encoded only by count, never by scale. The one design decision worth knowing: icon unit value (here, 2%) was chosen so that no region requires more than 50 icons per row — beyond 50 discrete marks, counting becomes error-prone and the advantage over a bar chart disappears.