Pareto Charts show which factors contribute the most to a given problem. The power of Pareto Charts lies in the Pareto Principle, which states that a small number of factors will typically control a given outcome the most. Identifying and focusing on these critical factors will maximize results, oftentimes with far less effort than would be required if all factors were treated equally.
Here is a typical Pareto Chart (see excel templates for the files that generated these images):
Pareto categories -
Left-hand vertical axis -
Right-hand vertical axis -
Creating a Pareto Chart
The following steps can be used to create a Pareto Chart.
1. Define the Problem Scope
Clearly state the problem or goal that the Pareto chart will address. Goal statements like, “Improve product quality in the Final Assembly area,” are too general and will not provide the focus needed for good data collection. Instead, focus on a specific metric like, “Reduce the final audit defect rate from 17,000 PPM to 9,000 PPM in Final Assembly.”
2. First Look for Existing Data
Oftentimes there is sufficient data already available to create a top-level Pareto Chart for the problem being addressed. In the case of our assembly defect rate example, the production floor will likely have historical data showing defect types and associated quantities for whatever time period is needed. Unless there is a specific timeframe of interest, retrieve historical data for at least one month, and preferably three months. Too short a timeframe (less than one month) may not show the complete picture of all causes, and too long a timeframe (more than three months) may dilute the current set of problems with past problems that have already been solved.
3. Plan for Data Collection if Needed
If sufficient data does not already exist, then a data collection plan must be implemented. Structuring the data collection effort with appropriate check-sheets or forms will go a long way toward a successful outcome.
4. Assemble Data Into Pareto Categories
A typical Pareto Chart will have ten or fewer problem categories, and data should be grouped to achieve something in this range. If there are too many categories then the chart will be difficult to read. Too few problem categories (typically less than five) are an indication that problem categories are not specific enough to create an “actionable” Pareto Chart.
5. Arrange the Pareto Categories in Descending Order, and Enter Into the Spreadsheet
Pareto data is usually organized by frequency of occurrence (number of defects over a given timeframe for a given defect type). Sort the groups in descending order, and you are ready to enter your data into the Excel spreadsheet.
Pareto Charts are simple but powerful tools for problem solving – give them a try and they will likely find a permanent place in your problem solving arsenal.