Image by Author | Ideogram

Pareto charts are simple yet powerful visualization artifacts based on statistical concepts. They are frequently used in data analysis, decision-making processes, and quality control. Interested in understanding what they are, what they are used for, and how to visually interpret them? This article has you covered.

What are Pareto Charts and What are They Used For?

Pareto charts are the combination of a bar chart and a line graph, and they are primarily used to describe the relative importance of different categories or groups in your data. They are named after the well-known Pareto Principle or “80/20 rule”, according to which a small number of causes (about 20%) usually account for the majority of effects (about 80%). In the original scenario that led to the formulation of this principle, Vilfredo Pareto observed that 80% of the land in Italy was owned by 20% of the country’s population. In terms of sales, the Pareto principle would state that about 80% of sales come from approximately 20% of customers, and so on.

Aligning with this principle that focuses on the “vital few” rather than the “trivial many”, Pareto charts shows absolute frequencies of data categories in bars ranked in descending order, as well as total cumulative frequencies in the data displayed by a line. This combined view provides an invaluable approach for identifying and prioritizing opportunities, problems, or priorities in several domains, such as:

• Quality control: Pareto charts help pinpoint defects contributing to most complaints in products.
• Customer service: they are used to analyze customer complaints and shift efforts to the most common “pain points”.
• Inventory management: they help identify items that incur the majority of costs.

Let’s see a first visual example of a Pareto chart. The following chart displays a dataset of customer complaints categorized by type.

Pareto chart of customer complaints categorized by type

Here’s how to interpret this chart:

• Just like in plain bar charts, the vertical bars represent frequencies of data by categories. In this example, bars show the frequency of each complaint type, with late delivery being the most common cause for customer complaints, followed by damaged items.
• The cumulative percentage line shows the cumulative percentage of complaints as we move from left to right across the previously listed categories.
• The 80% threshold represented by a horizontal dashed line is used to easily identify the causes 80% of complaints come from. In this scenario, by looking at this intersection between the cumulative line and this threshold, we can state that roughly 80% of the complaints come from the two most common categories: late delivery and damaged items.

The key insight is, therefore, that by addressing just two of the existing issues -late delivery and damaged items- we can resolve nearly 80% of customer complaints. This insight is used to leverage resource prioritization or make more informed decisions upon data, with the ultimate goal of improving processes or solving key problem areas or pain points.

Another example Pareto chart used for quality control processes shows the frequencies and cumulative counts of defect types in manufactured components.

Pareto chart of defect types for quality control in a manufacturing plant

In this case, the frequency distribution is slightly more even, such that a moderately larger range of possible causes (defect types) account for 80% of the effects (the faults). Scratches, cracks, color issues, misalignments, and rough edges, would all be the major issues to further investigate in order to address the 80% of faults in manufactured components.

Best Practices and Limitations of Pareto Charts

To build a useful Pareto chart, it is important to keep in mind these key guidelines:

• Ordering categories in descending order of frequencies is vital to easily identify the most significant contributor causes.
• Calculate and use cumulative percentages -rather than cumulate frequencies- in the line to clearly illustrate the contribution of categories (issues) from most to least frequent ones.
• Stick to the ’80’ in the 80/20 rule, but use the 80% threshold to identify categories contributing to 80% of the effects to guide decisions. It is normal if sometimes the contributing causes are a bit more than 20% of existing types, as seen in the second example above. Your compass will be the 80% threshold in the last instance.

Regarding the limitations of Pareto charts, while they are very useful for identifying major contributors to effects, they do not take into account the context, failing to capture why certain issues occur. They also oversimplify the data, which might be an issue when dealing with complex datasets where some relationships or interactions matter between factors.

Wrapping Up

This article provided an introductory guide to understanding and leveraging Pareto charts as an easy-to-interpret yet useful data visualization tool in processes and scenarios like quality control, customer management, and other contexts where problem-solving and decision-making assets are needed to identify causes for common problems. Despite having existed for a while, they are still a must-tool for data analysts to assist them in where to focus their attention on the most critical areas.