What is a Column Chart?

A column chart is a method of displaying data with categories represented by a rectangle—sometimes called vertical bar charts. They allow easy comparisons among a number of items and trends analysis. In general, statistics and figures are difficult to understand when presented in tables or written format. Column charts make things easy, immediate, and understandable.

Column Chart Example

A column chart is distinct from a bar chart. While a bar chart plots the variable horizontally and the fixed dimension vertically, the column chart does the opposite. However, most people don’t make the distinction, and refer to column and bar charts interchangeably.

The original inventor of column and bar charts is thought to be William Playfair. In the late 1700’s, he published an atlas about the commercial and political statistics at the time. The first bar chart was said to be about the exports and imports of Scotland from Christmas 1780 to 1781.

Column Chart Demo
Visualizations/Charts with Spotfire
Check out this demo to see how easy Spotfire makes it to start visualizing all aspects of your data.

When to Use Column Charts

As one of the most common charts, there are a range of scenarios where they are excellent at conveying information in a logical, clear manner. Column charts are best used when:

  • The data has a small number of discrete categories. Each of those categories has a single value.
  • The chart needs to compare the values for each category.
  • The goal is to make the information as easy to understand as possible.

In business settings, column charts are regularly used in reporting. Whether the data is sales information, KPI outcomes, or a customer analysis, the column chart is ideal. They are far more flexible in their use than a pie chart, for instance, and present more data and categories.

To Compare Data Values of Related Categories

Column charts, at their simplest, compare a range of categories in a single measure. This allows readers to judge how well each single category is performing against the counterparts. Sales performance is an example of this.

To See Changes in Dependent Variables Over Time

Seeing how changes happen over a timeframe is another excellent use of column charts. Often, line charts can be replaced with a column chart. An organization may want to see their sources of internet traffic over a year, for instance.

To Compare Contributions of Different Category Members

The stacked bar charts show how different groups contribute to the overall total of something. Total sales in various branches achieved by individual salespeople is an example.

Comparing Negative and Positive Values

Many charts do not display negative data well. However, column charts are ideal for showing negative vs positive data and comparing performance against a benchmark.

When Not to use Column Charts

Avoid column charts and find an alternative chart when there are a large number of categories, when you need cumulative values, the values are small, or if the values are rates. Consider using line graphs if there are many categories to compare.

Good Practices for Column Charts

In order to create a chart that is easy to understand and is not misleading, there are a few basic practices to follow.

Unless there is a natural order to the data, such as age groups, sort the categories so the biggest category is on the far left. This is in line with the Pareto Principle.

The Y axis should start at zero. In a few cases, it may be possible or acceptable to “crunch” the scale at the bottom, but this should be used sparingly. For instance, if all the data starts at 1000 and finishes at 1500, then you could crunch the scale so it starts at 1000 as long as it’s not misleading.

Unless you are choosing specific colors for a reason or doing groups of data, use one color. It can be distracting and confusing otherwise.

If the actual figures are important, it could be beneficial to remove the axis labeling and write the corresponding values on the end of the column. However, if the important thing to display is the trends, then always use the axis.

Think about the gaps between the columns. The gaps need to be distinct, in order to allow a clear and easy distinction. However, the column width should always be wider than the gap, or it can make the graph harder to interpret.

Variation to Column Charts

There are a great many number of variations to column charts. Not all of them are suitable for use in every scenario.

Clustered Column Chart

Clustered column charts group items together. For instance, if there was an analysis of sales in a department store, there could be groups for each branch and the total sales for each division within the branch. So, there could be the electronics and furniture department sales for each branch across America.

Stacked Column Charts

Stacked column charts are useful when the total amount adds up to 100 percent. A chart has bars that, stacked together, show a percentage of 100. These aggregated figures can easily show percentages of total volumes across categories or timeframes.

For example, for the same department store chain, a stacked bar chart may show the percentage of sales each salesperson has achieved within the branch. There are multiple stores, and each one has a stacked bar chart that shows total sales for that branch and the percent of sales achieved by each salesperson within that branch.

These stacked charts are ideal when there are multiple values for one category.

Bar Charts

Bar charts are column charts tipped on their side. The quantities, rather than height, are communicated by the width of the bar. They offer no particular benefit or challenge in comparison to a column chart, but some data may be easier to understand when presented in a bar chart. Bar charts can have text in the bars as a label.

Column Chart Software
Try TIBCO Spotfire - Free Trial
With TIBCO Spotfire, the most complete analytics solution in the market, easily discover new insights from your data.

Benefits of Column Charts

Summarize a Large Data Set Visually

If a data set is grouped appropriately, column charts are easy to understand at a glance. Categories can be summarized so they are visually almost instantaneous to understand and interpret.

Good Heuristics

Because column charts are easy to understand, they allow a quick visual check for basic accuracy and reasonableness. You can see at a glance whether the data matches and makes sense.

Easily Understood

Due to their widespread use, column charts are easily understood. They are easier to interpret than a table and allow fast understanding of data.

Disadvantages of Column Charts

Can Require Additional Explanation

When someone looks at the chart, they may find there is simply not enough information to draw valid conclusions. This is often the case when there is too much data presented at once, or it is too complex for a column chart.

Solution: Display only one set of data at a time. Label both axes clearly, have a descriptive title, and if there are groups of categories, have a key to allow easy interpretation. If the data is still not presented clearly, consider using another chart type, such as a scatter plot or bubble chart.

Can be Easily Manipulated

If a chart maker wanted to intentionally distort data, it is very easy to do so. Simply by changing the scale or even the bar width, data can be manipulated to suit any agenda. For instance, if a company wanted to show sales volumes change over a period of time, they could have the scale starting at a figure other than zero, and then have a long distance between two figures, showing huge differences between branches, although in reality the differences could be small.

Solution: No collapsed scales, and always start the axis at zero. Readers need to be aware of the potential for manipulation and call out data presentations.

Failure to Reveal Effects or Patterns

If the data is presented poorly, it will not be meaningful. If categories are not equal, or the data not presented in an orderly way, the chart will not show causes, effects, and patterns.

Solution: Have meaningful, equal data categories. Unless there is a reason for doing so, the biggest category should be on the left with a downward trend to the right.