Side-by-side charts only support comparison across panels if they all share the same category order
Aliases: shared ordering · small-multiple sorting
What it is
For juxtaposed charts (small multiples, faceted panels, grouped bar sets) to support "same category, across charts," every chart must share one category order. Letting each panel sort by its own values makes each chart prettier in isolation but relocates every category on every chart — cross-chart comparison decays into re-locating the category panel by panel, which forfeits the entire point of juxtaposition.
Why it happens
Cross-chart comparison runs on positional correspondence: the reader locks a category's position in the first chart, then checks the same position in the rest — no re-searching anywhere. With a shared order the correspondence holds, and the cost of n charts approximates the cost of one; with per-chart sorts the correspondence breaks, every chart demands a fresh "find the category" search, and cost grows linearly with chart count. The shared basis should be the reference quantity of the primary comparison task — typically an aggregated total or the first chart's values — so the positional order is valid everywhere.
Where it stops holding
A shared order sacrifices each single chart's value-aligned elegance — an inherent price of juxtaposition: the primary job of side-by-side charts is comparison, not solo optimisation. The exception is charts that display rank change explicitly (bump charts, slope charts) — rank movement is the content, and order must vary per chart; that is the "rank is the message" special case, and the variation must itself be directly readable. With no dominant reference quantity, a unified alphabetical order is also legitimate — what matters is "shared," not "by what."
Applying it
- Fix the shared sort basis before laying out (total, first-period value, or name); every panel uses it, and no panel re-sorts alone.
- State the basis in the group title or caption ("regions ordered by total sales").
- Verification: have readers compare one category across three charts; vertical gaze movement between charts (no lateral re-search) means the shared order works — per-chart hunting means it does not.
Related
- Same group: U5.04.1 Category-axis order is a designer's choice, not a property of the data · U5.04.2 Value sorting makes rank and extremes readable without search · U5.04.3 Categories with intrinsic order must not be re-sorted by value · U5.04.5 State the sorting basis in the axis title or caption
- Nearby: U5.06.2 Past the density ceiling, aggregate, sample, or facet · U5.01.3 Legend order should match the graphic order
- Search terms:
small multiples·shared axis order·cross-chart comparison
Cards in the same group
- U5.04.1Nominal categories have no natural order, so whatever order a chart shows was somebody's choice
- U5.04.2Sorting a bar chart by value hands readers the ranking and the extremes for free
- U5.04.3Age bands or rating tiers should keep their natural order even if sorting by value looked tidier
- U5.04.5An unstated sort order leaves readers guessing, and they often guess alphabetical when it wasn't