Method of identifying the vital few causes ( typically 20% ) which cause 80% of the problem
History
In the 19th century, Vilfredo Pareto, an Italian studied the distribution of wealth and found that this distribution was not equal across the populace. He found that the majority of the wealth was concentrated in relatively few hands.
Applications
Quality
Production
Stock Control
Sickness
Absenteeism
Accident Occurrence
Resource Collection
How to Make the Pareto Chart :
Identify problem and likely causes
Collect Information about Causes
Prepare Pareto Chart Causes on the X- axis the problem expressed as no. of occurrence, cost or frequency on Y –axis
Identify the Vital few causes
Apply Improvement Techniques to deal with causes in order of Importance
Example
One washing machine manufacturer had a quality crisis. A major consumer magazine had ranked their products last in an extensive reliability test and advised consumers not to buy them. The manufacturer’s response was to increase the warranty period for its goods and offer consumer cash compensation for faults. Clearly however they could not sustain this level of support if faults continued.
The Company had kept good records of faults. There were 22 categories reported. However, a Pareto Analysis showed only 4 were responsible for 83% of the occurrence. Faults in the other categories had only happened occasionally and were reckoned to be rare.
Fault in washing machine
Fault in washing machine
When to use it
Can be used in a wide variety of situation where there are a number of variables contributing to a problem and you need to know which are the most important. Particularly useful at the start of an improvement program
Won’t be needed if more sophisticated system are in place, such as Statistical Process Control.
Versatility
First step in making improvement
Has strong impact
Universal applicability
Helps in selecting concrete goal
Every one concerned cooperates
Can speak language of money
Steps - Using Microsoft Excel
· Enter titles ‘ Category’, ‘Frequency’, ‘% Frequency’, ‘% Cum
· Frequency’, in cells A1, B1, C1 and D1 respectively
· Enter categories in column A
· Enter corresponding frequency in the column B
· Using mouse block ( select ) column A and B
· On the Main tool bar click Data/Sort, a pop Up menu shall appear
· Sort on column ‘B’, and Click on descending order
· Take the sum of frequency using syntax ( =sum(A1:A10) if there are 10 no. of frequencies are there.
·
· In column ‘C’ get the % frequency by dividing the sum of the frequency column to corresponding frequency.In first cell using syntax ( = c2/$B$11)
· In the column D get the cumulative frequency.
· First % cum frequency is equal to % frequency of first cell.( value of C2 = Value of D2 )
· In column ‘C’ get the % frequency by dividing the sum of the frequency column to corresponding frequency. In first cell using syntax ( = c2/$B$11)
· In the column D get the cumulative frequency.
· First % cum frequency is equal to % frequency of first cell.( value of C2 = Value of D2 )
· In column ‘C’ get the % frequency by dividing the sum of the frequency column to corresponding frequency.In first cell using syntax ( = c2/$B$11)
· In the column D get the cumulative frequency.
· First % cum frequency is equal to % frequency of first cell.( value of C2 = Value of D2 )
· In column ‘C’ get the % frequency by dividing the sum of the frequency column to corresponding frequency.In first cell using syntax ( = c2/$B$11)
· In the column D get the cumulative frequency.
· First % cum frequency is equal to % frequency of first cell.( value of C2 = Value of D2 )
Interpreting the results
The focus generally translate into the tallest bar on the Pareto chart, but the team should choose this bar only after review of the impact of the other problems.
If the distribution is flat, then try to look at the data in a different way. If the team used frequency measure, try cost, location, time or any other measure important to the customer. In the end a flat distribution may also indicate that the process is in a state of chaos, in which any problem has equal chance of occurring.
