UI Coach guide
Data Visualization for Dashboards
Chart and table choices for product dashboards, with an emphasis on questions rather than decoration.
By UI Coach Editorial ·
A dashboard chart exists to answer a question. If you cannot state the question, you are decorating. The default should be the simplest representation that does not hide the comparison the user needs.
Match the question to the form
- How big is this number? Use a labeled metric, not a gauge.
- How is it changing? Use a line or sparkline with a clear time axis.
- How do parts compare? Use a bar chart, not a pie with twelve slices.
- What is the exact value among many? Use a table.
Axes and units are content
Unlabeled axes are incomplete sentences. Truncated y-axes can exaggerate change. Dual axes invite misreading. If two series need different units, they may need two charts.
Empty, loading, and sparse data
A chart with one data point should not pretend to be a trend. Explain sparse data. Provide a table fallback when the visualization would mislead.
How to judge the work
A dashboard visualization is honest when each widget has a written question, when axes and units are labeled, when a single data point cannot impersonate a trend, and when a table exists for anyone who needs the exact number. If the prettiest chart cannot survive those checks, it is decoration.
Practice this on UI Coach
In dashboard challenges, write the question above each widget. Remove any visualization that does not answer it. That habit shows up in product work more than a custom illustration style.
Sources
- Nielsen Norman Group on data visualization