7/5/2015

3.4 PER MILLION

Use DMAIC to Make Improvement Part of ‘The Way We Work’

by Ronald D. Snee

Faster, better, cheaper. That’s what organizations across almost all major industries must now do to
remain competitive.

Faster embraces everything from product development to throughput to productivity to speed to market.
Better means higher quality, innovation, fewer defects and less variation.
Cheaper translates into reduced costs, less waste, greater efficiency and bottom­line benefits.

Sometimes these goals might seem at odds with each other—for example, better versus cheaper—but meeting
these goals is a must.

And they aren’t just one­time objectives but ongoing imperatives that can be summed up in a single word :
improve. As a result, most managers now have two jobs: doing the work and improving the way work gets done.

Enterprisewide, the long­term challenge is to combine all improvement work into an overall improvement system
and create the management framework to sustain that system.

Such a system can provide a common language for improvement across functions and processes, enable precise
definition and scoping of improvement projects, and offer a method for guiding project teams through problem
solving and process improvement.

In time, improvement becomes a routine managerial process, just like any other, and continuous improvement
becomes an integral part of the way the organization works.

Where to start on such a big job? Experience has shown that adopting define, measure, analyze, improve and
control (DMAIC) as the preferred problem solving and process improvement framework takes an organization a
giant step closer to making improvement part of “the way we work.”

DMAIC Beyond Six Sigma

Although DMAIC is associated with Six Sigma,1,2 there is no reason it cannot be detached from the method and
generalized to a higher level as an overall approach to improvement.

As such, it provides a proven approach to problem solving and guides the application of improvement tools in a
highly structured and sequenced approach; this occurs regardless of whether the specific tools originate within Six
Sigma, lean, Baldrige criteria or some other method. In fact, the use of DMAIC, from its initial definition of the
project forward, can help identify the most effective tools and techniques at each stage of the improvement
process for a particular project.

A simple example demonstrates the many virtues of DMAIC as a general purpose approach to improvement,
divorced from any particular method and, in this case, from business itself.

Tiki Barber, the retired New York Giants running back, could always be counted on to make big yardage gains,
and, in the early part of his career, for making costly fumbles. In four seasons, from 2000­04, Barber fumbled 35
times, more than any other running back in the National Football League.

Understanding what Barber’s fumbles were costing the team, Giants head coach Tom Coughlin analyzed the
problem and looked at a number of possibilities, such as the exchange from the quarterback, excessive shifting of
the ball from one arm to the other, or carrying the ball with the wrong arm. Coughlin concluded that Barber’s style
of carrying the ball—arms extended, ball held horizontal to the ground—was the root cause of the problem.

Coughlin ordered Barber to hold the ball vertically against his chest and to walk around during training camp
holding it that way.

As Barber would head through the locker room carrying the ball, an assistant coach would sneak up on him and
try to slap the fall from Barber’s grasp. Eventually, Barber became acutely conscious of holding onto the ball,
fumbling only nine times in the remaining three years of his career.

Recast in DMAIC terms, this problem solving and improvement process looks like this:

Define: Barber excessively fumbling the ball.

Measure: Calculate the impact of Barber’s 35 fumbles on field position, opposition points on turnovers,
outcome of games and team’s season record.
Analyze: Barber carries the ball horizontally with arms away from body, the root cause of his fumbles.
Improve: Barber should carry the ball vertically, holding it close to his chest.
Control: Have assistant coaches try to cause fumbles in the locker room. Closely observe how Barber
carries the ball in subsequent games and correct, if necessary.

As the example demonstrates, DMAIC is easy to use and understand; it provides a common language for
improvement of any kind. It links and logically sequences procedures and tools. And it is elegant in its simplicity.

http://asq.org/quality­progress/2007/09/process­managementment/use­dmaic­to­make­improvement­part­of­the­way­we­work.html 1/5
7/5/2015

Other Frameworks
Numerous other approaches to problem solving and process improvement claim many of the same virtues, and
they have often served us well in the constant pursuit of improvement.

However, it’s worth asking whether, on inspection, they have the problem­solving firepower of DMAIC. Consider
some of the most prominent of these competing frameworks:

Plan­do­check­act (PDCA) cycle: The four­step PDCA cycle has its origins in the work of great quality pioneers.3

1. Plan: Recognize an opportunity and plan a change.

2. Do: Test the change. Carry out a small scale study.
3. Check: Review the test, analyze the results and identify what you’ve learned.
4. Act: Take action based on what you learned in the study step. If the change did not work, go through the
cycle again with a different plan. If you were successful, incorporate what you learned from the test into
wider changes. Use what you learned to plan new improvements, beginning the cycle again.

As an iterative process, PDCA seeks to move closer to the best solution for producing improvement. In
experienced hands, it can often produce significant improvement. However, such success often occurs because
those experienced practitioners translate PDCA’s somewhat nebulous four steps into more precise terms, often
much like the far clearer and more explicit terms of DMAIC. As a general improvement framework, PDCA is
perhaps too broad, lacking the rigor and specificity of DMAIC.

The seven­step process: This approach also goes under the name of plan­do­study­act (PDSA), Deming’s later
elaboration of PDCA. It encompasses the following seven steps:

Define the system.

Assess the situation.
Analyze the cause.
Try out an improvement theory.
Study the results.
Standardize improvements.
Plan for continuous improvement.

Like PDCA, the seven­step process is iterative, seeking to get closer to the root causes of a problem with each
succeeding application of its methods. Also like PDCA, in experienced hands it can produce significant
improvement.

And like DMAIC, it includes steps for definition, analysis and improvement. However, also like its predecessor,
PDCA, it remains somewhat vague in its formulation, operating at too high a level of generality in steps like “study
results” and “plan for continuous improvement.”

8D: Originally developed by the Ford Motor Co. and introduced in 1987 in a company manual titled Team Oriented
Problem Solving, 8D refers to the eight disciplines the company follows in the identification, investigation and
resolution of problems.4 The company describes the process this way:

1. Team contact: This is reference information about the initiator and assignee of the 8D.
2. Problem description: This refers to a statement description of the actual concern (problem). Also, the
source that it came from and severity, or how bad it is.
3. Root cause analysis: This is the most important element of the whole process. The primary reason the
concern happened must be analyzed and a solution brainstormed. In succession, keep asking why until the
root cause is found.
4. Action taken (interim): This is the first step you take to isolate the customer from the problem,
immediately.
5. Action taken (permanent): This is the action taken to change the method, system or process on permanent
basis to keep it from recurring.
6. Verification of action: This describes how you know your actions are effective. It includes inspection results,
large sampling, statistical process control charts or other measurables.
7. Prevention: After you’ve found the root cause and taken actions, analyze and state at this step what could
have been done to prevent the problem from happening in the first place.
8. Team congratulations: Always recognize your people for their accomplishments in resolving the concern
and working as a team.

This team oriented approach has proven effective in helping solve recurring problems, especially problems of
quality that ultimately affect customers. As steps four and five suggest, it focuses on both short­term firefighting
and long­term correction.

However, as a general approach to problem solving, 8D does not specifically ask for the use of data like DMAIC
does. Arguably, 8D isn’t as specific on implementing control plans to hold improvement gains.

Lean: Lean principles help achieve remarkable efficiency in processes by eliminating wasted time, materials and
expenses; reducing nonvalue added activities; and increasing the speed and output of the operation. However, it
lacks a general approach to problem solving.

In fact, DMAIC can provide precisely the kind of overall approach to problem solving that lean needs to focus and
guide improvement efforts. (See “On the Way to an Overall Improvement Framework,” p. 54)

http://asq.org/quality­progress/2007/09/process­managementment/use­dmaic­to­make­improvement­part­of­the­way­we­work.html 2/5
7/5/2015

Putting DMAIC to Work

Once DMAIC has been introduced as a general purpose problem solving and process improvement framework, it
can be intially used to give managers an easy­to­use method of evaluating process studies by asking some simple
questions at each stage:

Define: What’s the issue and goal?

Measure: Where’s the data?
Analyze: How did you analyze the data and what root causes did you find?
Improve: What improvements did you consider and reject? How do you test the improvements and
determine that the improvement will work?
Control: What did you put in place to hold the gains?

As managers become familiar with the approach, they can begin to become familiar with the tools that are
appropriate at each step. As previously noted, DMAIC can be divorced from Six Sigma to provide a general
purpose improvement approach. The tools to be used at each step then can be drawn from any of numerous
methods—for example, Six Sigma, lean, lean Six Sigma, Baldrige criteria and ISO 9000—depending on the nature
of the problem.

In the Tiki Barber example, videotaping—a technique that lean often employs—would be used in the analyze
phase to get at the root cause of Barber’s fumbles and in the control phase to help sustain the improvement.
Thus, DMAIC can bring together the tools from various methods and integrate them in a holistic framework, as
follows:

Define: Identify the problem clearly, determine its financial impact and select and organize the right people
to solve it. Some key tools in the define stage might include: project chartering, value mapping and voice
of the customer.
Measure: To better understand a process, a project team might use any of various techniques of
measurement, including: process
mapping or flow charting, cause and effect matrix, videotaping, measurement system analysis, Pareto
charting and capability analysis.
Analyze: Key tools for getting at the root causes of problems include: more videotape analysis, failure
modes and effects analysis, and multivari studies.
Improve: To identify, test and implement solutions, the team—depending on the nature of the root causes
uncovered in the analyze phase, of course—might use: design of experiments to resolve any remaining
ambiguities about root causes and to quantify cause­effect relationships; production smoothing to distribute
the flow and mix of work more evenly over time; and kaizen events to generate ideas for improving
processes. Within the DMAIC framework, kaizen events will be even more powerful because the root
causes of the problem will already be known when the workers brainstorm solutions.
Control: Key tools for sustaining the gains achieved by the improvements implemented in the improve
phase include: a control plan to identify the variables that process owners must monitor; statistical process
control to provide early warning of anything that might warrant attention; standard work procedures to
eliminate human error; 5S to keep the workplace organized for maximum efficiency and minimal waste;
and mistake proofing to introduce in­line methods of quality testing to ensure quality at the source.

On the Way to an Overall Improvement Framework

Short of a full scale improvement framework in which all improvement tools are subsumed under define,
measure, analyze, improve and control (DMAIC), some organizations might take the intermediate step of simply
applying DMAIC to individual tools with which they might already be familiar.

For example, an organization that uses lean tools might begin by using DMAIC to guide the use of a tool from start
to finish, as follows:

DMAIC Kaizen Event

Define

Select kaizen opportunity.

Select team.
Determine objectives.
Hold team meeting.
Communicate goal and timeline.
Conduct training.
Assign responsibilities.

Measure/analyze

Validate current state.

Identify waste.
Brainstorm ideas.

Improve

http://asq.org/quality­progress/2007/09/process­managementment/use­dmaic­to­make­improvement­part­of­the­way­we­work.html 3/5
7/5/2015
Test ideas.
Implement ideas.
Fine tune.

Control

Prevent roll­back.
Document new process.
Train.
Conduct kaizen management presentation.

DMAIC Single Minute Exchange of Die—An Example

Define: Reduce setup time to less than 10 minutes.

Measure: Gather data on current approach to changing setups and time required. Typical setup time was 4 hours.

Analyze: Identify the source of operator setup activities and time requirements.

Improve: Change work methods and implement. Measure effect of new procedures. Deploy approach to other
processes. Setup time reduced from 4 hours to 2 minutes, 38 seconds.

Control: Make physical changes to the machines for operators to work on setups only in the new way and not the
old ways.

DMAIC 5S
Define

1. Plan a course of action.

2. Select and educate team.

Measure and analyze

3. Evaluate the work area.

Improve

4. Initiate the 5S’s.

5. Measure the results.

Control

6. Maintain 5S activities.

7. Deploy to other areas.

As DMAIC and the tools associated with each of its steps are woven into the entire fabric of an organization’s
problem solving and improvement activities, significant cultural benefits multiply. All personnel begin to think in
terms of processes, which are not only the way work gets done but also the only way improvement can take
place.

In addition, people become more adept at using data, clearly defining problems, and controlling and sustaining
improvements. Once accustomed to applying DMAIC to problems, people begin to continually look for better ways
of doing things.

Arriving at this ultimate state, in which continuous improvement is simply part of the way the organization does its
work, might take an investment of several years.

But by starting with DMAIC as the general framework for problem solving, organizations can significantly
accelerate the journey by establishing a firm foundation for building each successive element of an improvement
infrastructure. This will culminate in an enterprisewide, self­sustaining improvement system that makes faster,
better, cheaper a repeatedly attainable reality.

© Ronald D. Snee, 2007

REFERENCES
1. R.D. Snee and R.W. Hoerl, Leading Six Sigma—A Step­by­Step Guide Based on the Experience with
General Electric and Other Six Sigma Companies, FT Prentice Hall, 2003.
2. R.D. Snee and R.W. Hoerl, Six Sigma Beyond the Factory Floor: Deployment Strategies for Financial
Services, Health Care and the Rest of the Real Economy, Pearson Prentice Hall, 2005.
3. Nancy R. Tague, The Quality Toolbox, second edition, ASQ Quality Press, 2004.
http://asq.org/quality­progress/2007/09/process­managementment/use­dmaic­to­make­improvement­part­of­the­way­we­work.html 4/5
7/5/2015
4. Ford Motor Co., 8D Problem Solving, 2002.

RONALD D. SNEE is principal of performance excellence and lean Six Sigma initiative leader at Tunnell Consulting
in King of Prussia, PA. He has a doctorate in applied and mathematical statistics from Rutgers University in New
Brunswick, NJ. Snee has received the ASQ Shewhart and Grant medals and is an ASQ fellow.

Copyright © 2005­2014 American Society for Quality. All rights reserved.

Republication or redistribution of ASQ content is expressly prohibited without the prior written consent.
ASQ shall not be liable for any errors in the content, or for any actions taken in reliance thereon.

http://asq.org/quality­progress/2007/09/process­managementment/use­dmaic­to­make­improvement­part­of­the­way­we­work.html 5/5