Glossary
What is the data-ink ratio?
Data-ink ratio: The data-ink ratio is the share of a chart's ink that presents data, a measure introduced by Edward Tufte in 1983. The idea is to erase ink that carries no information.
The data-ink ratio is the share of a chart's ink that shows data. Edward Tufte, an American statistician and writer on information design, introduced the idea in The Visual Display of Quantitative Information (1983). Data-ink is the ink you could not erase without losing information. Everything else, such as heavy frames, decorative fills and marks that repeat other marks, is ink that does not help the reader.
Data-ink ratio = data-ink ÷ total ink used in the graphic
Nobody measures it with a ruler. It works as an editing question: if I erase this mark, does the reader lose anything? If not, the mark is a candidate for removal. Tufte's advice is to remove ink that carries no data and ink that repeats other ink, as long as the chart stays easy to read.
Where it shows up on a slide
The slide below compares revenue by region. The values are written on the columns, so the chart needs no value axis. There is one baseline, and the title states the finding. Nothing repeats anything else.
The same chart before editing might look like this:
Low data-ink ratio
Revenue by region: 3D columns with shadows and gradient fills, a heavy frame, a value axis, dark gridlines and a value label on every column. The same values are shown twice.
High data-ink ratio
Revenue by region: flat columns, the values written on them and one baseline. Each number appears once.
How to apply it
- Start with the message. Write the title as a full sentence, then ask what the chart needs to prove it.
- Run the erase test. Remove one element at a time and keep it removed if nothing is lost. The chartjunk entry has the full list of candidates.
- Label directly. Put names and values on the data instead of in a legend or on an axis. Do not keep both an axis and labels for the same numbers.
- Lighten what stays. Draw grid and axis lines in pale gray, so the data marks stand out.
- Put back what the reader needs, such as a zero baseline, a unit or a source.
Typical mistakes
- Treating the ratio as a score to maximize and stripping out labels, units and sources until the chart is hard to read.
- Removing the zero baseline from a column chart, so the columns float and their lengths mislead.
- Deleting annotations. A note that explains a spike carries information, so keep it.
- Confusing minimal with clear. A chart with nothing but bars and no title tells the reader nothing about the point.
- Applying it as a law. A 2010 study by Bateman and colleagues (CHI 2010) found that embellished charts were interpreted about as accurately as plain ones. Chartjunk covers how to judge each case.
The ratio is one test among several when you are choosing the right chart, and the space you free up is best left as white space.
FAQ
Frequently asked questions
What is the data-ink ratio?
The data-ink ratio is the share of a chart's total ink that shows data. It is data-ink divided by all the ink used in the graphic. In practice it works as a rule of thumb for erasing what is not needed, not as a number you measure.
Who invented the data-ink ratio?
Edward Tufte introduced it in his 1983 book The Visual Display of Quantitative Information. He used it to argue that ink carrying no information, and ink that repeats other ink, should be removed from charts.
Should you always maximize the data-ink ratio?
No. Remove what carries no information, but keep what makes a chart readable, such as labels, units and a baseline. A 2010 study by Bateman and colleagues (CHI 2010) also found that embellished charts were interpreted about as accurately as plain ones.
How do you improve the data-ink ratio of a chart?
Apply the erase test: remove one element at a time and keep it removed if nothing is lost. Then write values and names directly on the data and lighten what remains. The chartjunk entry lists the usual candidates.
Keep reading
Charts and numbers
- Free toolSlide contrast checker
Check the WCAG 2.2 contrast ratio of two slide colors. See AA and AAA results for text and charts, a live slide preview and the nearest colors that pass.
- GlossaryDriver tree
A driver tree breaks a metric such as revenue or profit into the measurable drivers behind it, linked by simple math. See a worked example you can copy.
- GlossaryWaterfall chart
A waterfall chart shows how a start value becomes an end value through a series of increases and decreases. Definition, how to read it and typical mistakes.
- GlossaryGantt chart
A Gantt chart shows a project's tasks as bars on a time axis, with milestones as markers. Definition, how to read it and how to simplify it for a slide.
- GlossaryHarvey balls
Harvey balls are round pictograms that show how far a criterion is met, from empty to full in quarter steps. Meaning, symbols and how to use them on slides.
- GlossaryTwo-by-two matrix
A two-by-two matrix sorts items into four quadrants using two criteria. Definition, well-known examples such as the Eisenhower matrix, and how to build one.
Where to go next
- ConvertTurn a spreadsheet into slides
Tables keep editable cells, and you change the numbers later in the editor or by request.
- ProductRevise an existing deck with AI
You review every change one by one and can restore an earlier version.
Less theory, more finished slides
Describe the deck you need. The AI plans, writes and designs it, and hands you an editable PowerPoint.
Build your first deck freeFree plan, no card needed