-
Six Sigma eCourses
T H E F U T U R E O F S I X S I G M A W E B - B A S E D T R A I
N I N G H A S A R R I V E D
Six Sigma Metrics
Section 1: Types of Project Metrics Section 2: Measuring Metrics
Section 3: Defects vs. Defectives Section 4: Yield
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 2 of 40
Module Objectives
In this module we will provide key concepts and definitions for
use in the formal project definition process. The concepts covered
here will include: Types of project metrics:
o Business Metrics o Primary Metrics o Consequential Metrics o
Financial Metrics
Calculating Baseline, Entitlement and Benchmark The difference
between Defects and Defectives The role of DPMO and sigma level The
difference between classical yield and Rolled Throughput Yield
These terms and definitions are the foundation for building your
ability to provide Six Sigma practitioners with properly defined
projects.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 3 of 40
DMAIC Roadmap
The definition of Six Sigma metrics for a given project is of
invaluable importance. Defined in the Define phase, these metrics
track the progress of the project throughout the MAIC phases. In
addition to management, Green Belts and Black Belts require these
skills as well.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 4 of 40
Section 1: Types of Project Metrics Introduction
The single biggest contributor to the success of Six Sigma is
its strong focus on the use of metrics. One of the differentiators
of Six Sigma is the way in which project progress is defined and
tracked. Continuous improvement efforts of the past deployed
organizational projects without a structured, quantifiable manner
of measuring progress.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 5 of 40
Metrics Overview
Project metrics are commonly documented in a Project Charter or
a Project Definition Worksheet. These metrics are: Business Metrics
Primary Metrics Consequential Metrics Financial Metrics
The single most important role of a Champion and a Process Owner
is the definition of projects. Six Sigma practitioners must sanity
check the definition of projects assigned to us by the
Champions.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 6 of 40
Business Metrics
The business metric is a high-level existing management
performance indicator that Champions care a lot about. This
category of metrics measures the high level health of a business, a
functional area, or a department. Examples include: profitability
percentages, inventory levels, customer satisfaction indices, scrap
or yield rates, throughput measures, capacity calculations, or
virtually anything that measures a single or multiple business
processes. Business metrics are useful for business management.
Business metrics are influenced by multiple processes and/or many
outputs.
Business metrics are often mistakenly given to Green Belts or
Black Belts as Primary Process Metrics
that need to be improved in a project. These high level metrics
are usually outside the direct control of a Green Belt a Black Belt
and/or their team members. For this reason, we like to avoid
assigning business metrics as the primary metrics for a
project.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 7 of 40
Primary Metrics
Primary Project Metrics are the opposite of Business Metrics.
These are what the Six Sigma practitioner cares about and can
influence. Primary Metrics are used to directly measure the success
of the project. Primary Metrics are direct output characteristics
of the process. We also refer to them as process metrics. The
primary metric of a project almost always falls into one of three
categories: Defects Cycle time Consumption
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 8 of 40
Consequential Metrics
A consequential metric is a business or process metric that
measures anything that could go wrong as a result of improving the
primary metric. It measures any potential negative consequences.
This metric is sometimes referred to in industry as a secondary
metric. Theres usually only one primary metric for a project, but
there can be multiple Consequential Metrics.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 9 of 40
Categories of Primary and Consequential Metrics
Defect-based metrics are commonly expressed as: Percent
defective PPM (parts per million defective) DPU (defect per unit)
DPMO (defects per million opportunities)
We can take a positive spin on things by measuring yields
instead of defects. A percent defective would become a classical
yield A DPU would become a Rolled Throughput Yield A DPMO would
become a sigma level
All of these metrics are defect-based measures of a process.
With these types of project metrics, our focus is on eliminating
errors, defects, non-conformances, or mistakes in order to increase
the capacity for good units or reduce rework or scrap rates.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 10 of 40
Categories of Primary and Consequential Metrics (Continued)
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 11 of 40
Categories of Primary and Consequential Metrics - Continued
Cycle time metrics usually measure the time to process a single
unit, time per unit or units per time. Cycle time projects can be
tricky in the Define and Measure phases because we dont necessarily
standardize when the clock starts or stops. The intent with these
types of projects is to eliminate unnecessary or redundant steps in
a process or to decrease the time it takes to complete a particular
step.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 12 of 40
Categories of Primary and Consequential Metrics - Continued
Consumption metrics measure the ability of a process to consume
a particular resource: A raw material Water Power
The goal becomes more efficient consumption of this resource.
For every project, a primary metric needs to be defined. The
primary metric directly measures the process producing a particular
product or service.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 13 of 40
Common Primary and Consequential Metric Pairs
Primary metric categories help us define potential consequential
metrics. Therefore, primary and consequential metrics can be
defined in pairs. We dont want to decrease defect rate at the
expense of increasing cycle time. Therefore a possible
consequential metrics for a defect based project is cycle time
per unit. You could reduce the cycle time by adding additional
labor. But the intent of the project is to reduce cycle
time without adding labor. Therefore labor would be tracked as a
consequential metric.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 14 of 40
Common Primary and Consequential Metric Pairs - Continued
If consumption is the primary metric, theres a potential that
you may negatively affect your quality levels if you use less raw
material. The strength of your product will suffer if you use less
titanium in it.
At a hospital, bed sheets may not be clean enough if you use
less detergent.
A defect-based quality metric can be a consequential metric for
the consumption primary metric of raw material usage.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 15 of 40
Finance Metrics
Financial metrics convert process improvements into tangible or
intangible dollars. Financial metrics are commonly translated from
business and process metrics and are usually defined in conjunction
with a finance rep. Common financial metrics include: Cost per unit
Dollars per hour Dollars per overtime hour Scrap costs Revenue
dollars Dollars to dispense with safety and hazard material Dollars
spent on outsourcing a particular good or service
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 16 of 40
Recognizing Metrics (Activity)
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 17 of 40
Section 2: Measuring Metrics Introduction to Measuring
Metrics
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 18 of 40
Baseline
Baseline is the average, long-term performance of an output
characteristic (Y) when all input variables (Xs) are running in an
unconstrained fashion. This means that we let all input variables
vary across their entire range of expected values.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 19 of 40
Entitlement
Entitlement is the opposite of the Baseline. Entitlement is the
best case, short-term performance level of an output characteristic
(Y), when all input variables in the process are running centered
and in control or in a constrained at particular optimum level.
This optimum performance usually occurs when Xs are constrained at
particular levels. This is a very important measure of the Y
because it helps us determine what were capable of and allows us to
set realistic goals.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 20 of 40
Benchmark
Benchmark is the performance level of the best possible in the
industry. The benchmark is the best that anyone has ever done, not
just your company. What has the competition achieved? What do the
industry trade journals say is possible? What are other business
units accomplishing with this process or with a similar
process?
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 21 of 40
Section 3: Defects vs. Defectives Introduction to Defects and
Yield
The most common category of primary metric is the defect rate,
because sigma level is about defect reduction. Defect based metrics
can be negatively stated: Percent bad Percent rejected Percent out
of spec
Or positively stated: Percent good Percent yield Percent in spec
or sigma level
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 22 of 40
Defects vs. Defectives - Introduction
In order to select the right quality metric for a Six Sigma
project, you need to know the difference between defects and
defectives. Drive your organization toward using defect based
measures instead of defectives.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 23 of 40
Defects vs. Defectives Introduction (Activity)
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 24 of 40
Defects vs. Defectives Definitions
Defect: a feature on a product or service that fails to meet one
of possibly several internal or external acceptance criteria. A
unit may have a single defect or even multiple defects but will not
necessarily be defective. Defective: when an entire unit fails to
meet some external customer-imposed acceptance criteria.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 25 of 40
Defects vs. Defectives Definitions (Continued)
In this example, there are 8 defects. Where there was a crack,
whether or not it leads to a loss of functionality, is a defect. In
this example, there are 3 or 4 defectives, depending on the
acceptance criteria.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 26 of 40
Defects vs. Defectives - Calculations
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 27 of 40
Defects vs. Defectives - Analysis
The proportion of defects is higher than the proportion
defective. A single unit could have more than one defect because
there is more than one opportunity for defect on each unit. A
single rejected claim could have multiple defects: Wrong physician
Wrong insurance information Wrong test ordered
Only measuring defectives proportion bad or proportion rejected
hides the true defect level. Defective-based metrics the true
effort it will take to get rid of the defects. It makes the process
look much better than it actually is. Whenever possible we should
use defect-based metrics to quantify failure, noncompliance or
error rates.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 28 of 40
Defects vs. Defectives Classification (Activity)
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 29 of 40
DPMO
DPMO stands for Defects per Million Opportunities. We can use
this metric to estimate a long-term or short-term sigma level.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 30 of 40
Common Usage Guidelines
Metrics that do not include opportunity counts are best for
individual projects. This can include percent-defective, defects
per unit, or sigma-level. It may seem logical to utilize
opportunity-based metrics such as DPMO or sigma level as project
metrics, but their most important use is in the comparison of
dissimilar products and services. These opportunity-based metrics
allow for normalizing out the complexity via the opportunity count
of a particular good or service. DPMO offers a way to make a fair
comparison of two organizations quality levels. DPMO and sigma
level are also commonly used to roll-up organizational quality
performance across dissimilar functions and departments.
Opportunity-based metrics are not normally recommended to measure
project performance.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 31 of 40
Section 4: Yield Introduction to Measures of Yield
The concept of differing quality-based metrics (defectives
versus defects) has a corollary in the yield world: Classical Yield
and Rolled Throughput Yield (RTY). Percent defective is the same as
percent yield (Classical Yield). Defects Per Unit corresponds to
percent Rolled Throughput Yield. Defects Per Million Opportunities
corresponds to Sigma Level.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 32 of 40
Yield Brain Teaser
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 33 of 40
Classical Yield Defined
Classical Yield is the number of good units divided by the
number of total units times a hundred percent. A second definition
of classical yield is one hundred percent minus the quantity of the
number of defective units divided by the number of total units.
Either way, we base our calculation on the number of defectives, or
total units minus good units, found at the end of the process.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 34 of 40
Classical Yield Limitations
Using classical yield to measure performance doesnt tell the
whole story. Using classical yield as a metric has some
limitations: It ignores the internal hidden factory rework. It
counts delivered defectives, not in-process defects or defectives.
It hides true performance.
Another way to measure yield is by using Rolled Throughput
Yield. In manufacturing circles, classical yield is sometimes
referred to as First Time Yield, First Pass Yield or Final Test
Yield.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 35 of 40
Rolled Throughput Yield - Introduction
Rolled Throughput Yield is the probability of a unit making it
all the way through a process with no rework. Any unit that
requires rework at any point in the process is considered bad,
whether it results in a good unit or not. Using classical yields
can misrepresent the increased effort and cost due to rework.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 36 of 40
Rolled Throughput Yield
The reason RTY is called a rolled yield is because it rolls the
yield of each process step into the subsequent one. Each step has
its own In-Process Yield, known as IPY. The most accurate estimate
of Rolled Throughput Yield is calculated by multiplying all of the
subordinate in-process yields together. We calculate the Rolled
Throughput Yield as the in-process yield of step one times the
in-process yield of step two times the in-process yield of step
three. Using Rolled Throughput Yield when in-process data is
available provides a true picture of the overall rolled process
yield. It also provides data on where defects are occurring within
a process, not simply a tally of defectives at the end.
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 37 of 40
Rolled Throughput Yield - Analysis
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 38 of 40
Rolled Throughput Yield - Advantages
RTY enables you to: Estimate your true yield performance Allow
for much finer resolution of measuring improvement Pareto-ize your
defects and nail the worst offenders
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 39 of 40
Calculating Six Sigma Metrics (Activity)
-
Six Sigma Metrics
1999-2005 Breakthrough Management Group, Inc. CONFIDENTIAL. Any
Reproduction or Dissemination Strictly Prohibited. All Rights
Reserved September 03, Breakthrough Management Group. Unpublished
proprietary work available only under license. All rights
reserved.
Page 40 of 40
Module Review
In this module, we learned: It is easy to confuse high-level
business metrics with hands-on level primary process metrics. It is
important to measure any potential negative consequences through
defining consequential metrics. How to value the financial impact
of each project through financial metrics. Financial metrics are
the way
we engage leadership to buy into our projects. Measuring the
metrics will require using historical or newly gathered data to
estimate the baseline performance, entitlement, and or benchmark of
a process. The entitlement and benchmarks will allow us the ability
to set reasonable targets for the magnitude of the improvement that
we will seek. Defectives and classical yield tend to hide true
performance. Instead, defect-based measures, or Rolled Throughput
Yield, should be used as primary process metrics when dealing with
quality based project objectives.