How to create a

Bar Chart

The bar chart is probably the most all-rounded chart type out there. Learn how to get the most out of it.

What you learn


1.

When to use a bar chart

When the labels are long, this often is the best option

2.

New features in Datawrapper

Example: You can place long labels over the bars

3.

Extra features

Added features for even better bar charts

Intro: When to use a bar chart

Bar charts are among the most often used chart types in Datawrapper. Strictly speaking we need to distinguish bar charts (horizontal) from column charts, although many people talk of "bar charts" for both. 

The main difference is that horizontal bar charts provide a way to display longer labels. With a column chart the only space you have is the width of the X-axis. If your labels are very short, this often is the best option. But, what if the labels are questions from a survey and quite long?

Let us visualize a dataset in different ways to show how to get to a great looking bar chart with the new features of Datawrapper released in 1.10.

The Data

The Story: More and more people search for jobs via a variety of online platforms. This is the key finding from data published by PEW Research Center, published on November 2015. ("Searching for Work in the Digital Era").

Below is an example of how this would look like in a column chart. Datawrapper actually handles the labels quite well. But there are issues with breaks. It's too much information cramped into too little space.

So, this is why we use a bar chart (below is one of our "old" bar charts, the version we provided before the new release 1.10). Problem here is that the line breaks of the labels are not good, the text on the left looks mixed up. Before the update of our charts the only way to fix this would have been to somehow shorten the labels. Not a good compromise, was we would loose relevant information. 


The new bar chart

Below is the same chart, using the new bar chart module. One change is that the line breaks are cleaner.

Still, the chart is quite "crowded". If you look into "Refine" you will find new options, which will help us to make this chart better and easier to understand.

This is just one feature that is now available: 

Via "Refine" > Labeling we can choose the option "Display labels in separate line". This will be a feature that will come handy always when the labels are long, such as questions from a survey. 

But there are even more options, which help us to make the chart cleaner and even easier to understand. 

We can "swap bar and value labels", which would put the percentage to the front and the labels into the bars. 



Below is an example how the chart will look after the swap: 


Using additional options in "Appearance" we can add a dotted line between the rows, make the bars a bit thicker and add a light grey bar in the background. 



Below is the final chart, with all of the features from above enabled:

Separate rows with dotted line (not absolutely needed here, but it's nice)

Thicker bars (same as above, not strictly needed in this example. Use this feature if you have only few bars to make them visually more present. If you have many bars leave the bars as they are)

Adding color: As a final stroke we have colored the use of online resources for the job hunt from more traditional ways to get a new job, to show the large change at first sight. 


Other Interesting Tutorials

How to create a Stacked Column Chart

Learn how to create a stacked column chart in minutes with Datawrapper.

View Guide

All-new line charts

Datawrapper has completely revamped line charts: More responsive, cleaner and with annotations.

View Guide