Stacked bar charts are often maligned, but there are many places where a stacked bar chart is the best application for a data display. Our examples of stacked bar charts will help you figure out which are the right choice for your needs. When used with other graphs, stacked bar charts can be very powerful tools.
Best Examples of Stacked Bar Charts
In the event that you need to present a lot of data in a very comprehensive and easy-to-consume way, there’s usually no better choice than a chart of some sort. However, there are many different kinds of charts to choose from, and picking the one best for your data needs can be a difficult process.
One of the most controversial charts out there is the stacked bar chart. For those unfamiliar, a stacked bar chart is like a general bar chart, but instead of each bar correlating to a specific set of data, one bar will represent many sets of data, each one represented by a different color. So, instead of one bar on your chart being entirely blue, your single bar will have different “stacks” of blue, green, and orange.
These charts are controversial since they can produce confusion in readers who are used to seeing a single bar represent a single type of data, rather than multiples. In fact, there are many people who argue that stacked bar charts shouldn’t be used at all
Should I Use a Stacked Bar Chart?
However, we disagree with the above linked post. There are many reasons why stacked bar charts are the best choice in certain applications. While stacked bar charts shouldn’t be used constantly, there are many reasons to use them, and it’s very possible that you can make use of them well.
When Should I Not Use a Stacked Bar Chart?
If your goal is simply to show the differences between data sets in a single category, you should not use a stacked bar chart. You should use a regular bar chart for this application. For example:
This is a simple bar chart. In this particular example, there is a single category being displayed (“excuses for being late to class”) and then there are five different data sets. In this case, a stacked bar chart would be confusing, since then you’d just have all the data sets stacked on top of each other in a single large bar. This would confuse the reader and detract from the point that you are making. With a regular bar chart in this application, it is clear that the most popular excuse for being late to class is forgetting to set the alarm, while the least-used one was believing that it was still dark. Putting this information in a stacked bar chart would be counterproductive.
Stacked bar charts should be used when there are multiple categories that need to be compared, though they aren’t necessarily the right choice all of the time. There is also the option to use a grouped or a multi-series bar chart.
This is an example of a multi-series bar chart:
Clearly, this is a more complicated bar chart with more data. Not only are there multiple products in this example, but also multiple categories. In this example, five different sales strategies are being compared when used for four different products. Here, you can easily say that while Strategy 4 was very effective for products C and D, it was not nearly as successful for product B when compared to Strategy 2.
This sort of bar chart allows you to compare the impact of each strategy on each particular product. A multi-series bar chart allows for this kind of breakdown.
If what you wanted to know was which was the most effective sales strategy for each product, you could go back to the simple bar chart. Again, a stacked bar chart is not necessary to display this since you are looking at the averages for each strategy, which is a single data set. So, in this case you could display the products on the X axis and then display the revenue from the strategies on the Y axis.
In these instances, you should not use a stacked bar chart.
When Should I Use a Stacked Bar Chart?
Here is an example of the data in the multi series chart being displayed in a stacked bar chart. Again, as said earlier, you could have all the bar charts be the same color and it would tell you that Strategy 2 was overall the most effective sales strategy across all the products.
However, what the stacked bar chart will tell you at a glance is what percentage of that success was attained in each product. While the normal, simple bar chart will tell you that Strategy 5 was overall the least effective, what it will not tell you is that a large reason why it is the least effective overall strategy was due to how poorly it performed for Product D. If you look at the other three products, you’ll see that Strategy 5 actually didn’t do all that badly, and for some of the products (namely Product A), Strategy 5 was actually one of the more effective options for that product.
If you just looked at the simple bar chart for this, you’d simply believe that Strategy 5 was a total dud. The stacked bar chart lets you know that this may be the overall case, but it is not necessarily the case when it comes to certain products.
Basically, if you were showing this to a CEO who wanted to know which sales strategies worked best across five products in his company, the simple bar chart would make him tell you to throw out Strategy 5. But the stacked bar chart would tell him that while Strategy 5 should not be used for Product D, it is actually the best choice for Product A and a strong contender for Product C.
Couldn’t I Use a Line Chart?
There are many similarities between a stacked bar chart and a line chart, but it largely depends on the kind of effect you want to get across to the person consuming your information. For example, you could certainly use a line chart to graph the trajectory of the success of the sales strategies above relative to the products they are used on, but this won’t have as strong as a visual impact.
Line charts can also be somewhat confusing since they are a combination of thin lines in different colors that cross over each other. This becomes even more evident the more categories you have to work with. In the above example, there are only four different products and five different strategies. Imagine what it would be like with hundreds of products and hundreds of strategies! Your line chart would look more like an etch-a-sketch than anything resembling actual information.
Largely, the points of a line chart and a stacked bar chart are similar in that both are meant to get a good representation of the big picture. Line charts are also most often used to compare the difference between data points over time. They don’t have to be used in this way, but this is definitely their most common use, so if your data points don’t necessarily have anything to do with the amount of time spent gathering the data (as our above example does not), then line charts can be misleading.
But Aren’t Multi Series Bar Charts Better?
Not necessarily. Again, it depends on the point of your data and what questions you are trying to answer with that data. Let’s look at another example:
This is a multi-series bar chart. You can see that the sales have been compared across all regions and divided up by quarters. This way you can easily see how each region performed in sales across all quarters.
Just like the previous examples, this chart makes it easy to compare quarters. You can see that the first quarter was very slow for both the North and the East, but off the charts in the South and the West. You can also see that the second quarter was strong in the North and the South, weaker in the East, and very weak in the West.
Again, multi-series bar charts make this sort of analysis easy and possible. If your question is “How did each region perform in each quarter?” this makes it easy to answer and easier than a stacked bar chart, since comparing the slices of a stacked bar chart is more visually difficult as compared to comparing completely separate bars.
Take the same data compared in a stacked bar chart:
Now, what the stacked bar chart does for you is let you easily analyze how each region did in sales compared to each other and also let you see which quarters were responsible for what component of that region’s sales. So, if you want to know which region produced the highest sales overall, the stacked bar chart easily lets you know that it’s the South region.
If you look at the multi-series bar chart, you would be misled by the fact that the East had a high spike in sales during the third quarter. Many people would look at the multi series chart and assume this since due to the spike in the East. However, the stacked bar chart tells you that while the South’s sales were a little more stable, they were also more overall.
Basically, it largely depends on the way you want to display your information. Stacked bar charts are not the best all the time, but they are certainly an effective bar chart to have in your arsenal when it comes to displaying information effectively.
How Do I Know What Kind of Chart To Use?
As we’ve mentioned, stacked bar graphs are really great for showing totals, but they may not always be the best choice for your data display. One of the really fun things about data is how many ways that you can display it, and it can be astonishing how the data can tell you different things depending on how it is arranged. Fortunately, there are many different kinds of charts and graphs you can use to display your data, so you can arrange it in a manner that is most effective for your purposes.
Bar graphs (stacked and non) are great to use when you have more than 10 items to compare or if you have long data labels. They are also very useful when displaying negative numbers, as it’s very clear if a value drops below the 0 point. (This can be less clear with line graphs depending on their use.)
As mentioned earlier, line graphs are best to use when you are tracking the change of something over time. Line graphs are almost always used with a time metric attached to them; using them without one can be misleading and confusing. When you are using a line chart you should only use solid lines (dotted lines cause confusion), and you really shouldn’t plot more than four different lines on one graph. More than this can throw a reader off and make your line graph hard to follow.
Pie charts are unbeatable when it comes to displaying the relation of pieces to a whole. The sum of all pieces of a pie chart has to equal 100 percent, as 100 percent is the value of the entire pie. If you have data that requires values outside of this, then you shouldn’t choose a pie chart. Pie charts should also be avoided for use with negative values, unless the entire pie chart is being used to show negative values.
Scatter plot charts look like line graphs, but without the lines. They are typically used to showcase the distribution of the data. They are best used when you have lots of different data points and you want to highlight where the data points are most similar. Scatter plot charts should always have the Y axis starting at 0 for maximum clarity.
Choose the Chart To Fit Your Needs
Stacked bar charts are powerful, but they aren’t necessarily the right choice all the time. Make sure to think about what you want your data to actually say and what kinds of charts will best showcase your data to make your point. A good way to frame this is to think about the kind of questions you need your data to answer. Are you looking for simple comparisons? Or comparisons of the whole? Or how many parts make up a whole?
Then use the chart that best helps answer your particular questions.