What is the data-ink ratio and how do you apply it when designing a chart?
Coined by Edward Tufte, the data-ink ratio is the proportion of ink (or pixels) in a chart that encodes actual data, divided by the total ink used. Maximizing it means removing every element — gridlines, borders, tick marks, legends, decorative shading — that does not carry information the viewer cannot infer from the remaining ink.
How to think about it
The formula in plain terms
data-ink ratio = data ink / total ink used in the graphicA ratio of 1.0 means every pixel encodes data. In practice, some non-data ink is necessary for legibility (axis labels, a title, a legend when groups cannot be directly labeled), but the goal is to cut anything that does not earn its space.
Common sources of chart clutter
| Element | Action |
|---|---|
| Heavy outer border / box | Remove or lighten to a subtle gray |
| Dense gridlines | Replace with sparse, light gridlines or remove and annotate key values |
| Tick marks pointing inward | Remove or reduce to stubs |
| Redundant legend (colors labeled on bars directly) | Remove legend; label bars directly |
| Background fill / gradients | Remove |
| 3D effects, shadows | Remove |
| Excessive decimal places in labels | Round to the significant digit |
Applying the principle step by step
- Start with the default chart output from your tool.
- Ask for each element: “If I removed this, would the viewer lose information?” If no, remove it.
- Lighten what you keep: gray gridlines instead of black, thinner borders.
- Use direct labels over a legend whenever space permits — it removes one decoding step.
- Move the title from a generic label (“Bar Chart of Sales”) to a declarative sentence (“APAC leads in Q2 revenue growth”).
The erasure test
Tufte suggests the erasure test: erase any non-data element and ask whether the chart lost information or became harder to read. If neither, the element was clutter. Apply iteratively.
Tufte’s principle applies equally to dashboards: every tile, color, icon, and border should justify its presence. A dashboard that scores well on data-ink ratio typically has fewer elements but each one does more work.