4. Communicating vs Charting


In this lesson, we start our second case study by considering a stacked bar chart. This will show us how an overly complex chart can sometimes try to answer several questions, and fail to answer any of them

To explore more Kubicle data literacy subjects, please refer to our full library.


  1. Lesson Goal (00:13)

    The goal of this lesson is to learn about the potential issues with overly complicated charts.

  2. Understanding Stacked Bar Charts (00:34)

    The second case study in this course deals with a stacked bar chart. This is a common chart used to condense data. In our case, we consider a stacked bar chart showing sales over several quarters for different product categories. The chart uses location for the data, size to represent sales, and color to represent product categories. With three traits in use, the chart can become difficult to read and interpret.

  3. Getting Information from a Bar Chart (01:39)

    People want to obtain answers to various questions from a visualization. On our chart, we want to know what are the sales trends, which category has the highest sales, and what quarter has the highest and lowest sales for each category.

    Although the bar chart contains all the information we need to answer these questions, it’s not easy to find a lot of the answers. We often have to compare segments that are very similarly sized or differently colored. Some small segments are difficult to identify at all. Effectively, the bar chart is trying to display too much information and not displaying any of it clearly. When creating visuals, it’s better to consider what question or questions it should answer, and ensure it does that effectively. If this means creating multiple visualizations instead of a single one, then that’s what you should do.


In the previous lesson, we completed our first case study in this course. We're now ready to move on to the second case.

In this lesson, we'll learn about the potential issues with overly complicated charts.

We're now going to move on to our next case study.

In this case study, we're going to look at a stacked bar chart, identify some issues with it and see how we can tweak this visualization in a way that makes it easier to interpret all while following the visual hierarchy.

The stacked bar chart is a very common way to condense data. Here, we can see a stacked bar chart that shows sales values over several quarters for various different product categories.

It's clear straight away that this chart is trying to convey a lot of information.

If you think about this in terms of the visual pieces, it's an X, Y plot where X is time and Y is the sum of sales.

So the plot as a whole uses location, then we're using size to represent sales and color to represent the category.

And that creates a challenge. While it's easy to read the total sales value, having three traits of the hierarchy makes it harder to analyze the data.

With the stacked bar chart, people are trying to answer important questions like, what is the trend? Which category had the highest sales? Which quarter had the highest and lowest sales for each category? Let's consider these questions individually, starting with the trend.

The size of the bars gives us an indication of the overall sales trends, but this format doesn't make it easy to determine the trends for individual categories as it's tricky to compare their size from one quarter to another.

In most quarters, it's relatively easy to identify the largest category. Usually, binders and orange or storage in yellow has the largest sales within a bar but sometimes like in Q419, there's not a lot in it, and it's difficult to be certain.

It's pretty much impossible to identify which category has the highest sales overall.

Finally, it's very difficult to identify which quarter had the highest or lowest sales for each category.

For most categories, the segments are simply too small to compare quarters easily.

This visualization makes it very difficult to answer this question easily.

In effect, this bar chart is trying to answer all of these questions and in doing so it makes it difficult to answer any of them well, because we have to scan the chart in detail to try and find answers.

When creating charts, it's tempting to add as much information as we can just because it's possible to do so. However, this is a bad idea.

It's better to step back and ask yourself what question should this chart answer and make sure it answers that question effectively.

This may mean creating a series of visualizations instead of a single one, but if this is the best way to convey the relevant information, then that's what you should do.

So how do we begin to simplify this chart and make it easier to see what's going on with the data? Remember, the first step is to try and eliminate the trait that's hardest in the hierarchy. Here that's color.

Let's stop the lesson here and see how to do this in the next lesson.