Pareto Principle

Quick reference

• 00:04 Hi, this is Ray Sheen.
• 00:06 And I want to take a few minutes to discuss a principle we will use again and
• 00:10 again, it's called the Pareto principle.
• 00:14 One thing that we continually find we must do is prioritize our work.
• 00:18 We don't have unlimited time and money, so
• 00:21 we need an easy method to decide what to work on and what to ignore.
• 00:25 The Pareto principle does this for us.
• 00:28 I mentioned that we will use this again and again on our Lean Six Sigma projects.
• 00:31 The first time it is used is by the stakeholders to help them decide which
• 00:36 process the team should work on next.
• 00:38 One of the nice points about Pareto is that it works with any type of
• 00:41 measurement.
• 00:42 It could be cost, customer impact such as sales, business process
• 00:46 impact such as defects, or even the number of people needed to work on the project.
• 00:51 The stakeholders can decide what metric is most important based upon the business
• 00:55 strategy and pain points.
• 00:57 Then all of the potential Lean Six Sigma projects are gathered up, and a quick
• 01:01 count is made of how each project measures up based upon the stakeholder's metric.
• 01:06 Taking the value for each process, we can create a Pareto chart and
• 01:10 be able to visually see which process should be the next one to
• 01:13 work on in our now prioritized list.
• 01:17 Let me show you how the Pareto process works.
• 01:19 In a Pareto analysis we first list all the potential projects, or
• 01:23 list whatever else it is that we are trying to prioritize.
• 01:27 You will find in a later phase we prioritize root causes with Pareto.
• 01:31 Now based upon the selected metric, we create a score or account for
• 01:35 each of the processes.
• 01:37 Let's say our metric is errors, and we have ten processes.
• 01:41 We could look at the last three months and
• 01:43 count the number of times an error occurred in each process.
• 01:47 Now we plot those on a vertical bar chart.
• 01:51 The process with the highest number of errors is in the farthest left position.
• 01:55 Then the next highest number, and the next, and the next.
• 01:59 Until we get to the process with the fewest number of errors,
• 02:02 which is in the farthest right column.
• 02:04 Now take a minute and look at this graph.
• 02:07 There are two vertical scales.
• 02:09 The left scale is the actual count of whatever you are measuring,
• 02:13 in our case it's errors.
• 02:14 The blue bars are using this scale, and you can see the process A has 76 errors.
• 02:20 The next process only has 28 errors.
• 02:22 We keep going down until we get to the last two processes,
• 02:26 each with only one error in the past three months.
• 02:29 The right-hand vertical scale has units of percentage,
• 02:33 the red line is using this scale.
• 02:35 The scale shows the cumulative percentage of all errors across all processes
• 02:40 that are included to the left of the position on the line.
• 02:44 So the first process, process A, represents 62% of all errors.
• 02:50 And processes A and B combined represent 81% of all errors.
• 02:54 This line will always end at 100%, because by the time we have included all of
• 02:59 the processes we have accounted for all of the errors.
• 03:03 Priority is given to the leftmost column.
• 03:05 This is because if I can reduce or eliminate that column,
• 03:08 it will have the biggest impact on the overall number of errors.
• 03:12 So in our example, if we can expect the Lean Six Sigma project to reduce errors to
• 03:17 just 25% of what they had been initially, doing project A would eliminate 57 errors,
• 03:22 whereas doing project B would only eliminate 21.
• 03:26 If the amount of work was the same,
• 03:28 we would get a much larger business impact working on project A.
• 03:33 Now you may say to yourself but why do we need a chart,
• 03:36 why not just use the numbers in a table?
• 03:39 Well the reason is that many people understand a picture of data faster than
• 03:42 they can understand numbers in a data table.
• 03:45 The chart improves communication, and it isn't hard to make.
• 03:50 Now you may be wondering why is it called a Pareto chart?
• 03:53 It was developed by Vilfredo Pareto, an Italian engineer,
• 03:57 economist, and behavioral scientist.
• 03:59 Pareto was studying the Italian economy and found a consistent
• 04:02 pattern across the country, 20% of the population controlled 80% of the wealth.
• 04:07 The principle became known as the 80/20 rule, and people found that that principle
• 04:12 applied to much more than just wealth in a country.
• 04:15 In almost any kind of ranking there were a few major items with a large value,
• 04:20 and lots of items with a small value.
• 04:23 Within the process improvement community, the 80/20 rule is often applied to root
• 04:29 causes where we find 80% of the defects come from just 20% of the defect causes.
• 04:34 So whether we are prioritizing processes, root causes,
• 04:38 or impact of solutions, we will find the Pareto principle to be a quick and
• 04:42 simple analysis that graphically displays what is most important to work upon.
• 04:49 The Pareto principle is one of the easiest and yet
• 04:51 most powerful tools that we'll use with Lean Six Sigma projects.
• 04:54 And that just shows that advanced statistics are not always required to have
• 04:59 a successful Lean Six Sigma project.

Lesson notes are only available for subscribers.