Pareto Analysis

Our ISO9001, ISO14001 and OHSAS18001 management systems require us to maintain records of checks, audits and inspections. Analysing this data can help us set priorities for process and system improvement and to assist us in defining performance targets.

Pareto analysis can help us identify the most significant issues from the general morass of data and requires no real mathematical or statistical knowledge. It has also been called the 80/20 rule and underlines the premise that 80% of our problems tend to have 20% of the possible causes.

Vilfredo Pareto was an Italian economist and sociologist who identified that in his country approximately 80% of the wealth was owned by 20% of the population. The Pareto principle can be applied to all kinds of data for instance we might identify that:

  • Approximately 80% of or sales come from 20% of our customers
  • Approximately 80% of customer complaints come from 20% of our distributors.
  • Approximately 80% of audit findings are against 20% of our processes.

Of course the premise is a general approximation and not a rule that can be consistently proven mathematically.

To apply the Pareto analysis to set of data follow these simple stages:

  1. Review the data and decide how best to analyze it. Remember that we are looking at items that recur. We can analyse by recurring issues or recurring causes.
  2. Once decided list the issues or causes
  3. Use tally charts to identify the number of times each cause or issue occurs.
  4. Plot the frequency of occurrence against the issue starting with the highest frequency first and progressing to the lowest. See example below:

The above example shows types of employee injuries over 4 years for a scrap recycling plant. (The data is ficticous but does illustrate typical results for a Pareto analysis.) As a general rule with this type of data, we can see that by tackling the top 3 or 4 issues we can significantly reduce the number of overall accidents. (A client uses a similar analysis as part of an improvement programmed to support their OHSAS18001 integrated management system)

In this example we have analyzed the number of occurrences for each type of injury. Our next step should be to analyze why so many lacerations occur by identifying the root causes. We may need additional data to do this effectively for instance to know how many of the injured persons were wearing protective gloves at the time. This may necessitate introducing a process for collecting more detailed data; for example: amending our accident reporting form to collect data about the types and extent of PPE used.

Collecting additional data, if it is not readily available, will take time. However, in order to effect an immediate reduction in accidents we can use alternative techniques (like brainstorming) to identify potential root causes and use this data to improve our overall process.

If we have further analysed our data to identify the root causes of each issues (or potential root causes), we could carry out a Pareto analysis of causes as in the example below:

The three key system standards: ISO9001 (Quality), ISO14001 (Environmental), OHSAS18001 (Health and Safety) all contain specific requirements for analysing data.

In the example above, the results of such analysis give a clear indication of where we need to concentrate our process improvements in order to reduce our audit non-conformances.

The three key system standards: ISO9001 (Quality), ISO14001 (Environmental), OHSAS18001 (Health and Safety) all contain specific requirements for analysing data and for continuous improvement.

Pareto analysis is a simple and powerful tool for helping quality practioners to analyse data in order to identify the vital few issues and separate them from the trivial many. Used as input to management review and performance review activities it can be invaluable in focusing management’s attention on the key issues in the continuous drive for process and business results improvement.

Leave a Reply