Lessons from IBCS – Eliminating Chart Clutter
Interested in exciting thought leadership and revenue insights?
Sign up for the Fluence newsletter.
The core job of a chart is to convey information to the viewer as clearly, concisely, and quickly as possible. One thing that can detract from this is chart clutter – distracting or unnecessary additional elements which can mask the overall message.
“Simplicity is the ultimate sophistication” Leonardo da Vinci
Edward Tufte coined the term ‘chartjunk’ in his book “The Visual Display of Quantitative Information” as far back as 1983. While there is some debate around the finer details of what specific elements are considered chartjunk, most field experts agree that chartjunk / clutter detracts from rather than adds to a chart’s usefulness. Avoiding clutter forms the ‘Simplify’ component of IBCS’s ‘SUCCESS’ formula. This recommends avoiding unnecessary elements such as pictures and backgrounds, borders and cluttered layouts.
To take a familiar and simple example from Excel, the first chart below is from an old version of office using the default settings with 3D turned on, and the second chart is a mildy formatted version from the current version. Different software will give differing levels of chart clutter by default, but the example below starting from a deliberately bad base helps show some of the elements.
The second chart is much clearer and easier to read, but what has actually changed? The column colours are better, and the legend has been moved to the bottom, but aside from that it’s a case of taking things away – less is often more.
The chart clutter which has been removed or toned down includes:
- Remove the 3D effect – it never (ever!) helps…
- Grey backfill on the chart area
- Borders on the columns
- Border on the legend
- Heavy gridlines have been subdued
- Vertical axis bar
- Axis now displayed in thousands
Another consideration is the use of colour. While a colourful chart can be eye-catching, we will instinctively look for meaning in the colour and if there isn’t any the colour is a distraction. Take a look at the examples below.
The top chart has a legend which is merely duplicating the categories already shown on the x-axis, and the colours are meaningless. The chart below simplifies the layout to provide a clearer view of the data.
XLCubed’s Dynamic Charts apply sensible and clutter-free defaults to help set you on the right path to effective reporting with minimal effort. We have also added a ‘Message’ styling option to quickly align the charts to the IBCS standards.
Value Axes, Grid Lines and data labels
IBCS suggests that the value axis and Grid lines in charts are often superfluous and can be replaced by data labels. Initially a chart without a value axis sounds a little odd, but in truth we see that type of chart regularly and when you take a look at the examples below, which are both XLCubed Dynamic Charts, you can see it works very well in many cases.
With the first chart, the gridlines help the eye allocate an approximate value to each column, but there is quite a lot of ‘eye-juggling’ to and from the axis to do so. Simply removing the axis and indeed the gridlines and adding data labels means the values are instantly available. It’s important with this approach to avoid unnecessarily large numbers by rounding to thousands or millions as appropriate, which should then be included in the report title e.g. Sales in mGBP.
In XLCubed, to move from the first chart to the second, you simply apply the ‘Message’ format as shown below, which aligns the formatting to the IBCS standards, or you can of course make the individual changes manually.
Another consideration around chart clutter is ‘over-labelling’. Depending on the chart type, labelling every data point may cause the overall view of the chart to become obscured in a sea of numbers or captions. This can often be the case with small multiple, or trellis charts as shown below.
In this example, across the 16 charts there is a lot of data label text, and it makes the comparison of the different charts more difficult. In this case it’s actually better to remove some of the labels to create a clearer overall view. XLCubed has several options for this, but one which often works well is ‘High, Low, First, Last” as shown below.
Alternately, for a final version of the report for formal communication, XLCubed Message gives complete control over the display and positioning of data labels and captions.
We hope this gives some food for thought in your reporting and helps you de-clutter – much more to come in the following weeks.