# What is chartjunk? > **Chartjunk:** Chartjunk is decoration in a chart that adds nothing to the reader's understanding, such as 3D effects, heavy grids, shadows or background pictures. Edward Tufte coined the term in 1983. - Source: https://ai-presentations.com/glossary/chartjunk/ - Updated: 21 October 2026 - Language: English - German version: https://ai-presentations.com/de/glossar/chartjunk/ Also called: chart junk **Chartjunk** is visual decoration in a chart that does not help the reader understand the data. The statistician Edward Tufte coined the term in *The Visual Display of Quantitative Information* (1983). Typical examples are 3D effects on flat data, heavy patterns and gridlines, drop shadows, background pictures and ornaments. Tufte's chapter on the subject describes three kinds: - **Vibrations:** patterns and hatching that seem to shimmer. - **Grids:** heavy gridlines that compete with the data. - **Ducks:** graphics in which the design takes over from the data. The word borrows from the architects Robert Venturi, Denise Scott Brown and Steven Izenour, who called a building shaped like the thing it sells a duck (*Learning from Las Vegas*, 1972). ## Where it shows up on a slide Decoration often starts as a wish for emphasis. The slide below needs the reader to look at one site. Instead of an arrow, a burst shape or a 3D column, it keeps one column at full strength and dims the others. The emphasis sits in the data and costs nothing in readability. *One column at full strength does the work an icon or a burst would do. Example with invented numbers.* 3. Berlin has 64 weekly active users, Leipzig 17; Leipzig is the one to watch ## The erase test For every element, ask what job it does for the reader. | Element | Verdict | Why | | ------------------------------- | ------------------- | ------------------------------------------ | | 3D columns or pie slices | Erase | Perspective makes values harder to compare | | Shadows and gradient fills | Erase | They add ink and no information | | Dark or dense gridlines | Lighten or erase | They compete with the data | | Picture behind the plot | Erase | It lowers contrast and adds noise | | Icons or clip art over the data | Erase or move aside | They hide the marks or compete with them | | Values written on the data | Keep | They carry data | | Zero baseline of a column chart | Keep | Columns need a reference to compare | | Note on a spike | Keep | It explains the data | ## What research says The rule is not absolute. A 2010 study by Scott Bateman and colleagues, presented at the CHI conference, compared plain charts with charts that had pictorial embellishments. People interpreted both kinds about equally well, and they remembered the embellished charts better two to three weeks later. The authors were cautious about recommending the style for every chart. In practice, decoration is not automatically harmful, but it has to earn its place. In a business deck the chart's job is to let people compare numbers and see the point, so plain is the safe default. If you want a memorable image, put it beside the chart and not on the data. ## Typical mistakes - **Decorating by default** because the chart tool offers shadows and gradients. - **Using 3D** for two-dimensional data. - **Emphasizing with effects** such as arrows, bursts and glows instead of contrast, for example one strong color on a muted base. - **Over-correcting.** Removing labels, units or the baseline in the name of cleanliness makes a chart harder to read. The [data-ink ratio](https://ai-presentations.com/glossary/data-ink-ratio/) explains where the idea stops being useful. - **Judging by taste.** Ask what each element does, not whether it looks busy. For the choice of chart type itself, see the guide to [choosing the right chart](https://ai-presentations.com/guides/choosing-the-right-chart/). [Visual hierarchy](https://ai-presentations.com/glossary/visual-hierarchy/) shows how to direct the eye without decoration. ## Frequently asked questions ### What is chartjunk? Chartjunk is decoration in a chart that does not help the reader understand the data, such as 3D effects, heavy grids, shadows and background pictures. The term was coined by Edward Tufte in 1983. ### Is all decoration in charts chartjunk? No. An element that does a job, such as one strong color that guides the eye to the key column, is not junk. A 2010 study found that embellished charts were interpreted about as accurately as plain ones and remembered better after two to three weeks, but its authors were cautious about recommending the style for every chart. ### What is the difference between chartjunk and the data-ink ratio? Both ideas come from Edward Tufte. Chartjunk names the decoration to avoid, and the data-ink ratio measures how much of a chart's ink shows data. Removing chartjunk raises the ratio. ### How do you remove chartjunk from a chart? Remove one element at a time, such as 3D effects, shadows, background pictures and heavy gridlines, and keep it removed if the reader loses nothing. Then lighten the grid, write values directly on the data and use one strong color to highlight what matters. ## Related pages - [Data-ink ratio](https://ai-presentations.com/glossary/data-ink-ratio/): The data-ink ratio is the share of a chart's ink that shows data, a concept from Edward Tufte. What it means, how to apply it and where it stops being useful. - [Choosing the right chart](https://ai-presentations.com/guides/choosing-the-right-chart/): Which chart fits which message: bars to compare, lines for trends, a waterfall for bridges. Includes when a pie works, when a table beats a chart, common traps. - [Visual hierarchy](https://ai-presentations.com/glossary/visual-hierarchy/): Visual hierarchy is the arrangement of elements so the eye reads them in order of importance. Definition, the five signals and how to apply them on slides. - [White space](https://ai-presentations.com/glossary/white-space/): White space is the empty area around and between elements on a slide. Definition, why it improves clarity and how to use it, with examples and mistakes. - [Presentation rules](https://ai-presentations.com/guides/presentation-rules/): What the 10-20-30, 5-5-5 and 7-7-7 rules say, where each one helps, where it misleads and which are worth following in a work deck. --- ai-presentations.com is published by Slaide (https://www.slaide.de), a sole proprietorship run by Nathalie Scholl in Heilbronn, Germany. [How this site is written, checked and corrected](https://ai-presentations.com/editorial-policy/).