top of page

# 5 GREAT INFOGRAPHIC AND DATA VISUALIZATION DESIGN TIPS

## WHAT IS THE 100 % BASIS?

If numbers sum up to 100 %, illustrate them as part of a whole 100 %. However, if numbers don’t sum up to 100 %, you can still show them in correlation to each other, but never as part of a whole 100 %.

### COMBINING DIFFERENT NUMBERS WITHIN ONE DATA VISUALIZATION

Sometimes it makes sense to combine several different numbers within one data visualization. The more different number sets go into one infographic, the more complex it is, of course. Let me show you an example:

This big data visualization is part of an infographic series I created for the smartphone repair place clickrepair. The outlines of the circles show how many smartphone users - depending on the brand - protect their smartphone with a protective case (86.1 % of Apple users do so). The orange circle fillings show how many smartphone users protect their smartphone with a screen protector (50.8 % of Apple users do so). Both case and screen protector are protective measures for a smartphone. This is one topic that gets fed by two different number sets.

It’s important to have a visual distinction between different number sets so that you can compare numbers within one and the same data set in an instant. At the same time, you want people to see the correlation between different datasets too. Another example is this one here:

This is one of many infographics I designed for Minor Projektkontor für Bildung und Forschung, funded by the Office for the equal treatment of EU Workers and the German Federal Government Commissioner for Migration, Refugees and Integration. It’s about notice periods employers are obliged to. There are two numbers that correlate to each other but don’t run linear: 20 years of employment (see bottom right) are illustrated by 20 dark green trapezoids. When somebody has been employed for 20 years, their boss has to give a seven months’ period of notice (seven months are illustrated by seven lighter points). However, ten years of employment don’t mean three and a half months of period of notice, but four months of period of notice. So the numbers are not linear, as you see. Nevertheless, you can very quickly compare different notice periods (the light dots) with each other, and at the same time you instantly see the rise of the corresponding years of employment (the dark green trapezoids.)

#### DESIGN FOLLOWS FUNCTION

The data visualization must serve the purpose of use. An infographic that’s used for a social media campaign must be designed in a snappy, quick-to-absorb way, with only little text and a not too little font size. I have a test account on Facebook which is invisible to the public, where I test how my infographic looks like in a Facebook timeline. Testing in a real scenario is always recommended.

If an infographic is printed in a huge poster size format for a fair, to give you another scenario, the infographic can of course be more complex as people will spend more time reading the infographic compared to scrolling down on Facebook.

##### COLOUR SELECTION: WORK WITH SHADES

A good data visualization looks great despite only a few colours. Colour differences can be achieved by working with shades. To give you an example: I have a green tone, then I add ten, 20 or 30 % of white or I add ten, 20 or 30 % of black. This gives me a wide selection of different green shades that all look very well together. This is how you show different colours within an infographic. In my opinion, there should be a maximum of five to eight colours in a data visualization. Very outstanding colours can be used for certain highlights or header bars to allow for an easy navigation for the reader.

###### MAKING INVISIBLE DIFFERENCES VISIBLE

Sometimes, you have to make differences visible even though the differences are only tiny. An example can be seen here:

This is again one of the infographics of the clickrepair study I mentioned earlier. This one is about the repurchase probability by device. The differences lie between 90 and 97 %. This data visualization doesn’t really show the differences very well.

The data visualization above is the design that I delivered to the client in the end. The focus lies on the area between 87 and 97 %. However, you must not simply shorten the bars as this would visually falsify the data.

It‘s important to include the icon marked in red above, which shows a break of the 100 % line. This makes people understand that this graph is zoomed into the upper percentage area to make the differences more visible.