When and How to Use Column Charts and Bar Charts (Stacked and Unstacked)

This is the second post in a multi-part series on data visualization. For a deeper dive into visualizations, download our whitepaper 6 Essential Chart Types for any Data Visualization Need.

If pie charts are the placebos of the data visualization world, column charts and bar charts might just be Advil. When it comes to simple, effective data visualizations, they’re tried and true.

We’ve all been exposed to column and bar charts. There’s a graph, with an x and a y axis, just like on your old TI-84 calculator. On that graph there are at least two bars which, at a glance, immediately show you the relative positioning of data — it really is that simple.

These chart types are the bread and butter of rankings or comparisons, whether you’re looking to share election polling data, or need to visualize sales numbers over time.

Sales by Month Column Chart - Stytch Example

Column and Bar Charts: Similarities and Differences

Though they’re similar in style, column and bar charts actually have different strengths and should be used in different ways. Understanding their nuances is key to creating effective and impactful data visualizations.

So how exactly do they differ? Column charts present the data vertically while bar charts do so horizontally. That might not seem like much, but it’s important because readers react to these two different presentation types in different ways.

When to Use a Column Chart

Column charts are great when the number of values you’re comparing is small (we’d recommend 12 at the most, but use your discretion based on the message you are trying to convey and how much space you have to display it). This might be a comparison of different categories (i.e. which devices your clients are using to order your product online) or a comparison of value changes for one category over time (i.e. share of employees using their tablet to buy your product online over the last twelve months).

If you want to show a part-to-whole relationship, there are also stacked column charts—where the column is split into several sections. These provide an opportunity to visualize a comparison of multiple groupings over time. To use the same example, if you want to see monthly sales of each of your product groups, a stacked column chart could be a great way to go.

Sales of Product Groups Over Time Stacked Column Chart - Stytch Example

When Not to Use a Column Chart

In order to maintain a clean, easily-understandable visual, you don’t want to use a column chart when there are a large number of data points. If your chart starts looking like a downtown skyline, you’ve probably gone too far and need to dial it back.

This issue is amplified when it comes to the use of stacked column charts. The colors add a degree of visual complexity, and can quickly get messy.

When to Use a Bar Chart

As a general rule, a bar chart should be used when the number of data points you want to compare is larger than seven. Our eyes are used to moving down a page, so the bar chart offers a more natural visualization experience when you have to share a larger data set. Imagine you want to show which websites are referring the most traffic to your website. A bar chart would allow you to present and compare the top 10 to 15 traffic sources at a glance. I often think that a bar chart is a great way to show the reader “who’s winning”.

Bar charts are also great for more practical reasons, like when you have really long category names and need space for long labels on your bars. If you wanted to visualize the performance of individual members of your sales team, writing their full names (like, say, Arnold Schwarzenegger) on a column chart would likely feel cramped and ruin the visual. Not so with the pragmatic bar chart.

Sales by Top 10 Lines of Business Bar Chart - Stytch Example

When Not to Use a Bar Chart

Small data sets (anything below four, really) will easily fit on a column chart, so there’s no reason to use a bar chart in that scenario (again, unless your labels are really long). And huge data sets don’t work well either. Everything has its limits, and with bar charts that’s around 15 data points. Once you hit that number, start considering other chart types, or think about how to simplify the data set you’re sharing.

You also don’t want to use bar charts for comparing a single category over time. Our eyes read time left to right, so we expect time to be on the x-axis. That doesn’t work with a bar chart.

4 Tips for Using Column and Bar Charts

  1. Make sure your numeric axis (this will vary based on whether it’s a bar or a column chart) starts at 0. People will assume it is, and starting at a different number will paint an entirely different picture of your chart.
  2. If you’re trying to show data over time, make sure it’s portrayed on the horizontal, or x, axis. People don’t read time as bottom to top; it’s always left to right.
  3. With stacked column charts, use contrasting colors to ensure readability.
  4. The simpler the better. We often instinctively try to include everything in our visualizations so as to leave nothing important out, but we end up diluting the power of the chart. Sorry to say it, but a stacked column chart with 12 columns and 12 different colors on each is a pointless visualization. Tone it down for simplicity.

Remember that these are guidelines, “rules of thumb”, starting points to help you quickly pick what will likely be the best visualization for your data. There are exceptions to the rules and the most important thing to remember is to look at your charts and ask yourself, “Does this picture tell me a story about my data that I can quickly and easily understand?”

Want to learn more about data visualizations? Download our whitepaper 6 Essential Chart Types for Any Data Visualization Need – How to Choose and Apply the Right Chart Type.

Download Whitepaper

 

By on November 25, 2016

Leave a Reply

Your email address will not be published. Required fields are marked *

Demo

Copyright © 2016 Stytch Inc. All rights reserved unless otherwise stated.