Six Sigma programs: An implementation model

Report as inappropriate Your report has been received

Shared by: alpd03l

Tags

Six Sigma, Process improvement, Implementation model, Case study Process improvement Implementation model Case study

Stats

views:: 83
posted:: 6/20/2012
language:: Latin
pages:: 16

Public Domain

Document Sample

ARTICLE IN PRESS
Int. J. Production Economics 119 (2009) 1–16

Contents lists available at ScienceDirect

Int. J. Production Economics
journal homepage: www.elsevier.com/locate/ijpe

Six Sigma programs: An implementation model
Satya S. Chakravorty Ã
Department of Management and Entrepreneurship, Michael J. Coles College of Business, Kennesaw State University, 1000 Chastain Road, Kennesaw,
GA 30144-5591, USA

a r t i c l e in fo abstract

Article history: Despite the pervasiveness of Six Sigma program implementations, there is increasing
Received 19 April 2008 concern about implementation failures. One reason many Six Sigma programs fail is
Accepted 15 January 2009 because an implementation model on how to effectively guide the implementation of
Available online 25 January 2009
these programs is lacking. Using a successful Six Sigma program in a Network
Keywords: Technology company, the purpose of this research is to develop an effective
Six Sigma implementation model which consists of six steps. The ﬁrst step is to perform strategic
Process improvement analysis driven by the market and the customer. The second step is to establish a high-
Implementation model level, cross-functional team to drive the improvement initiative. The third step is to
Case study
identify overall improvement tools. The fourth step is to perform high-level process
mapping and to prioritize improvement opportunities. The ﬁfth step is to develop a
detailed plan for low-level improvement teams, and the sixth step is to implement,
document, and revise as needed. Important for both practitioners and academicians,
implications of our implementation experience along with directions for future research
are provided.
& 2009 Elsevier B.V. All rights reserved.

1. Introduction Delphi Automotive have implemented Six Sigma programs
(Treichler et al., 2002), and claimed that these programs
Many characterize Six Sigma programs as the latest have transformed their organizations. Six Sigma
management fad of improvement tools and techniques programs are heavily promoted in practitioners’ books
(e.g., Watson, 2006). It is well known that Six Sigma on Six Sigma (e.g., Harry and Schroeder, 2000), and in
programs involve a host of critical decisions and many academicians’ books on Operations Management (e.g.,
researchers have contributed to the existing literature. For Jacobs and Chase, 2008). The American Society of Quality1
example, Schroeder et al. (2008) have identiﬁed many offers Six Sigma certiﬁcations; major corporations (e.g.,
critical decisions or elements of Six Sigma programs such General Electric Company, 2005) provide Six Sigma training,
as management involvement, improvement specialists, and a plethora of websites2 advertise Six Sigma solutions.
performance metrics, a systematic procedure, and project Despite the immense popularity and the wide-spread
selection and prioritization. Six Sigma programs improve adoption of Six Sigma, there is an increasing concern
operational performance in order to enhance customer across industries regarding the failure of Six Sigma
satisfaction with a company’s products and services programs. One reason many Six Sigma programs fails is
(Rajagopalan et al., 2004). Over the years, many compa- because an implementation model detailing the sequence
nies, such as General Electric, Allied Signal, Raytheon, and of Six Sigma elements/activities is not available. The
existing literature identiﬁes many elements of Six Sigma

Ã Tel.: +1770 423 6582; fax: +1770 423 6606.
1
E-mail addresses: satya_chakravorty@kennesaw.edu, www.asq.org.
2
schakrav@kennesaw.edu. www.isixsigma.com.

0925-5273/$ - see front matter & 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2009.01.003
ARTICLE IN PRESS
2 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

which does enhance our understanding of Six Sigma facilitating Six Sigma implementation (Gitlow and Levine,
programs. However, the success of Six Sigma programs 2005; Snee and Hoerl, 2003). Antony et al. (2007)
hinges on the sequence of many Six Sigma elements/ emphasize management’s involvement in on-going pro-
activities or a model for implementation. To put it jects for sustainability of Six Sigma programs. Second,
differently, it is well known that one needs many improvement specialists are trained or hired at different
ingredients for Chicken Curry. However, in order to cook Sig Sigma competency levels (e.g., Black Belt or Green
a delicious Chicken Curry, according to Jaffery (2003), the Belt). Their primary responsibility is to provide technical
recipe requires a sequence of ingredients/activities (e.g., expertize and leadership in facilitating a speciﬁc Six Sigma
heat oil, fry onion and garlic, add Indian spices, put in implementation (Pyzdek, 2003). Third, as Keller (2005)
chicken, pour water, and let it simmer). Any unreasonable points out, Six Sigma programs have performance metrics
deviation from the recipe will lead to less than positive and measurements based on cost, quality, and schedules.
experience. In the absence of a recipe or a model, it is not Fourth, Six Sigma implementation uses a systematic
surprising that many implementations of Six Sigma procedure; a ﬁve-step DMAIC (Deﬁne, Measure, Analyze,
programs have failed. A survey of aerospace companies Improve, and Control) methodology. A detailed descrip-
concluded that less that 50% of the respondents were tion of DMAIC methodology can be found in Pyzdek
satisﬁed with their Six Sigma programs (Zimmerman and (2003) or Keller (2005). Fifth, project selection and
Weiss, 2005). Another survey of healthcare companies prioritization is an important element of Six Sigma
revealed that 54% do not intend to embrace Six Sigma programs. The prioritization of projects is determined by
programs (Feng and Manuel, 2007). Companies such as many criteria, such as a cost beneﬁt analysis or the Pareto
3M and Home Depot were not satisﬁed with their priority index (Banuelas et al., 2006).
implementation of Six Sigma programs (Hindo, 2007; While we are familiar with many elements of Six Sigma
Hindo and Grow, 2007). Considering this, many authors programs, we lack an understanding of the sequence
question the return on investment of Six Sigma programs of these elements/activities, or a model for effectively
(e.g., Gupta, 2008). The real question is not whether Six guiding the implementation of these programs. Because
Sigma programs have value, but why do so many Six there is no implementation model, practitioners have
Sigma programs fail? One reason many Six Sigma encountered tremendous difﬁculty in implementing these
programs fail is because we lack a model on how to programs, and there are reports of wide-spread Six Sigma
effectively guide the implementation of these programs failures. Zimmerman and Weiss (2005) found that less
(Wurtzel, 2008). than 50% of the survey respondents from aerospace
Using a successful Six Sigma program in a Network companies expressed satisfaction with their Six Sigma
Technology company, the purpose of this research is to programs. Mullavey (2005) described the top 10 reasons
develop a model to effectively guide the implementation why Six Sigma implementations fail. Berg (2006) reported
of these programs. In the next section, we provide that their Six Sigma program was expensive and did not
the theoretical underpinnings of a Six Sigma implementa- yield expected results. Concerned about Six Sigma’s
tion model. Following a description of our research problems, Sutton (2006) described nine ways to get the
methodology, we present our Six Sigma implementation best out of Six Sigma programs. A national survey of Six
experience. Then, we provide implications of our imple- Sigma programs in healthcare companies revealed that
mentation experience including directions for future 54% do not intend to embrace Six Sigma programs (Feng
research. Finally, we provide the conclusion of our and Manuel, 2007). At 3M, a Six Sigma program that was
research. not correctly implemented almost stiﬂed their creativity
and innovation (Hindo, 2007). Home Depot’s Six Sigma
program negatively affected employee performance, and
2. Literature review yielded Home Depot’s worst Consumer Satisfaction Index
ranking (Hindo and Grow, 2007). Angel and Pritchard
2.1. Six Sigma (2008, p. 41) reported that ‘‘nearly 60% of all corporate Six
Sigma initiatives fail to yield desired resultsy’’. According
Over the years, many researchers have studied Six to Gupta (2008, p. 22), at times, Six Sigma ‘‘yimprove-
Sigma programs and identiﬁed many critical decisions of ment programs cost more than the improvement they
these programs. For example, see the previous research of drive because of incorrect applicationy’’. While reporting
Antony and Banuelas (2002), Coronado and Antony cash ﬂow problems of Six Sigma programs in small
(2002), Lakhavani (2003), Lynch et al. (2003), Mcadam companies, Foster (2007, p. 19) claims that if these
and Evans (2004), Gijo and Rao (2005), Szeto and Tsang programs are not ‘‘skillfully implemented; the beneﬁts
(2005), Ladani et al. (2006), Savolainen and Haikonen of Six Sigma may be marginal’’. According to Chandra
(2007), and Davison and Al-Shaghana (2007). Recently, Zu (2008), one reason Six Sigma programs fail is because
et al. (2008) studied the evolving theory of quality these programs are not correctly implemented.
management and the role of Six Sigma. While deﬁning
Six Sigma programs and uncovering the underlying
theory, Schroeder et al. (2008) identiﬁed ﬁve elements 2.2. Implementation model
of these programs. First is management’s involvement in
performing many Six Sigma functions, such as selecting As discussed before, a model for effectively guiding the
improvement specialists, identifying project selection, and implementation of Six Sigma programs is not available.
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 3

The existing literature research related to Six Sigma and step after securing management commitment and perso-
other improvement initiatives (e.g., Lean or Theory of nal involvement in a quality improvement initiative was
Constraints) are utilized to isolate steps of implementa- to identify the tools for improvement. According to Keller
tion. Although suggested in different studies, these steps (2005), Six Sigma programs have many tools for improve-
can connect with each other to hypothesize an imple- ment, which include Histograms, Pareto Charts, Statistical
mentation model. In describing a successful lean (e.g., Process Control (SPC), and Analysis of Variance (ANOVA).
manufacturing cells) implementation, Chakravorty and Foster (2007) claimed that a common process for
Hales (2004) found that the ﬁrst step in implementing an implementing improvement tools is the DMAIC metho-
improvement plan was to perform a customer and market dology, which is similar to Edward Deming’s ‘‘Plan-Do-
driven strategic analysis. The purpose of this analysis was Check-Act’’ problem solving approach. Lee-Mortimer
to direct the operational improvement effort to gain a (2006) considered the DMAIC methodology to be essential
competitive position in the market. Schonberger (2008) to Six Sigma programs and appropriate for delivering
points out that the objective of Six Sigma programs is to business improvements. According to Chakravorty and
create a higher perceived value of the company’s products Franza (2009), a form of DMAIC methodology, Deﬁne-
and services in the eyes of the customer. Antony et al. Measure-Analyze-Design-Verify (DMADV), was central to
(2005) indicated that linking Six Sigma to business a new product development experience. Mast and Bis-
strategy and customer needs is critical for successful gaard (2007) considered DMAIC methodology as the
implementation. According to Harry and Schroeder (2000, scientiﬁc method in Six Sigma programs. Keller (2005)
p. 39), Jack Welch said in GE’s 1997 annual meeting: provides a detailed description of each step of DMAIC
methodology, and of various levels of training (e.g., Black
y[Six Sigma implementation] begins not inside the Belt or Green Belt).
business, but outside it, focused on answering the In order to implement Deming’s style of quality
questions, ‘How can we make the customer more management, Hales and Chakravorty (2006) also found
competitive? What is critical to the customer’s suc- that after identifying the tools for improvement to be
cess?’ Learning the answer to the question and then used, the next step was to understand the overall
learning how to provide the solution is the only focus operations, and to set priorities for the project. One way
we need. to understand overall operations is by developing a
process map. According to Pyzdek (2003, p. 252):
Andel (2007) claims that Six Sigma programs should be
implemented with a clear objective of improving compe-
yA process map is graphic representation of a process,
titive positioning and of increasing the company’s value as
showing the sequence of tasks using a modiﬁed
perceived by the customer. All activities related to Six
version of standard ﬂowcharting symbols. The map of
Sigma implementation should be approached from that
a work process is a picture of how people do their
perspective.
work. Work process maps are similar to road maps in
In order to effectively manage bottleneck operations,
that there are many alternative routes that will
Chakravorty and Atwater (2006) found that the second
accomplish the objective. In any given circumstance,
step in that operation was to form a cross-functional team
one route may be better than others. By creating a
to guide the implementation process. Pande et al. (2000)
process map, the various alternatives are displayed and
point out that a cross-functional team is necessary to
effective planning [to improve the process] is facili-
implement Six Sigma programs and the purpose of the
tated.
team is to provide an on-going involvement of manage-
ment in the implementation process. According to Harry
and Linsenmann (2007), the CEO of DuPont committed According to Meredith and Mantel (2003), project
complete management support for implementing Six selection and prioritization is the process of evaluating
Sigma programs, and ensured that management learned projects, and then choosing to implement some set of
Six Sigma methodology by requiring that managers them so that the objectives of the organization are
themselves become Green Belt certiﬁed. At DuPont the achieved. Pande et al. (2000) claim that proper selection
Six Sigma program was not merely a methodology to get of Six Sigma projects and their prioritization could
results, but was a management culture created to ensure substantially improve the potential beneﬁts of Six Sigma
long-term transformation of the business units. Gopal programs. Many authors have proposed a variety of
(2008) found that one reason Six Sigma implementation models of Six Sigma project selection, prioritization
failed in many companies was due to the lack of procedures, and tools (e.g., Breyfogle et al., 2001; Adams
commitment from management. Management simply et al., 2003; Kumar et al., 2008).
pushed Six Sigma programs out to employees, and did Finally, in describing a successful lean (e.g., manufac-
not become personally involved in the implementation turing cells) implementation, Chakravorty and Hales
process. As Mullavey (2005) points out, in order to (2004) describe a detailed plan to successfully orchestrate
successfully implement Six Sigma programs, management activities related to the implementation. The plan should
must understand Six Sigma methodology, must provide include a sequence of execution activities including lower-
leadership, and must guide the implementation process. level team improvement exercises, a training schedule, and a
While implementing Deming’s style of quality man- detailed plan of all necessary changes related to the imple-
agement, Hales and Chakravorty (2006) found that the mentation. The plan provided a basis for implementation,
ARTICLE IN PRESS
4 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

documentation, and revisions needed to assure contin- with participants and stake-holders such as managers,
uous improvement. engineers, and employees. We conducted the interviews
in an open-ended nature, which implied that the
3. Methodology respondents provided objective opinions of the events as
well as insights into certain occurrences. We followed the
3.1. Case study general guidelines provided by Fontana and Frey (2000).
Fourth, qualitative data were collected in an observation
Our case study was conducted in a regional provider of mode involving many decisions during the implementa-
Network Technology services in the United States. These tion. Fifth, since the researcher was involved with decision
regional centers provide provisioning, service assurance, making during the implementation it was possible to
and customer care functions for all the products offered. collect data in a participant-observation mode. Sixth,
While none of these centers create ﬁnished goods from additional quantitative data were collected for mistakes in
raw materials, the locations are often considered small order entry, long setup times, and mistakes in network
internal factories. At the time of implementation, the transfer. These results provided clues to determine the
company held an estimated 5% of the domestic market. reasons for requests for manual activities which were
According to Yin (2008) there are three reasons why occurring in the service operations. During the study the
a case study research methodology is appropriate for researcher kept a research log that documented each
our study. First, when ‘‘how’’ and ‘‘why’’ questions are problem encountered during the implementation, in
investigated, the case study approach provides an ex- addition to the thoughts and insights gained during the
planation of linkages among events and their frequencies process.
of occurrence. This cannot be done through simulation
or statistical models based on survey. Second, the case
study approach is preferred when a real world event is 3.3. Data analysis
examined. Since many companies are actively engaged in
implementing Six Sigma programs, it is a natural way to The primary form of data analysis in case study is the
document emerging models from such an implementation reﬂection by the researcher on his own experience. The
environment. According to Stake (2000), real world researcher identiﬁed patterns and common themes by
studies are valuable for reﬁning theory and suggesting analyzing the experiences of themselves and other
complexities for further investigation. Chakravorty and participants. Content analysis worked well for identifying
Hales (2008) emphasize the need for real world based possible root causes and prioritizing alternative solutions.
research in order to help practicing managers on their In essence, content analysis is the counting of words,
journey to process excellence and prudent decision sentences, or ideas within categories of interest. In this
making. Third, the case study approach is appropriate study, we used it to collect ideas on how Six Sigma was
because this approach makes use of variety of evidence, implemented. Three general guidelines are applicable to
including documents, archival records, interviews, and content analysis. First, two judges were used for perform-
direct observation. According to Meredith (1998, p. 443): ing the analysis so that the consistency of results could be
estimated. Second, the categories of interest must be
ythe importance of direct observation [ﬁrst source
applicable to the research objectives. In this study,
(seeing it oneself) rather than second (speaking or
we collected data speciﬁcally on the implementation
writing to someone who saw or experienced it) or
and use of Six Sigma. Third, the units of analysis must
third, or sometimes no source at all], the role of the
be appropriate for representing the topic under examina-
context in which the phenomenon is occurring, and the
tion (Rosenthal and Rosnow, 1991). In this study, the unit
dynamics of the temporal dimension through which
of analysis was the operational/department level where
the events of the phenomenon unfold help to under-
Six Sigma was designed to be used.
stand the how and why elements of the phenomenon.
To examine the ﬂow of information through the
system; we employed process mapping. Each activity in
3.2. Data collection a business process is shown on a two-dimensional scale.
The processes are then connected with arrows showing
Multiple sources of evidence were used to validate the direction of service ﬂows. Average waiting times and
data. Yin (2008) identiﬁes six major sources of evidence. processing times were estimated for each stage of the
We employed all six in this study. First, qualitative data process. These maps helped identify where services had
were collected through documentation obtained in the breakdowns or long wait times and were processed
form of letters, memoranda, minutes of the meeting, through redundant or unnecessary activities.
progress reports, and strategic planning reports, etc. Each evening the researcher reviewed the information
Second, quantitative data were collected in the form of and data for the current day. They spent several hours
archival records of ﬁnancial data, customer complaint brainstorming ideas and reﬂecting on what had happened
reports, ordering processing, quality reports, purchase during the day’s events with several of the management.
orders, operational data (such as personnel utilization), They grouped the ideas using content analysis and
routing information, performance measurements (such as mapped the order processes which would be addressed
annual sales and responsiveness). Third, additional quali- the next day. The ideas with the greatest potential for
tative data were collected through extensive interviews success and most support from the team responsible for
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 5

changes or from other company executives were prior- occurred and manual intervention was required, some
itized for implementation. Results of the ideas implemen- of the beneﬁts of the system were lost due to costs
ted during the current day were reviewed for associated with manual repair. If the RMAs were reduced,
performance. Those that provided improvement were these expenses would be avoided and customer satisfac-
noted, as well as our preliminary insights into why they tion would be improved by reducing the circuit provision-
worked. Other areas in the order process that could ing interval. Therefore, it was very important to the BNISC
beneﬁt from the same solutions were identiﬁed for to reduce RMAs to decrease the variability, meet or exceed
implementation on the following day. Each morning, customer expectations, and achieve the original purpose
the team chose the next problems to attack based on and savings of deploying an automated solution.
consensus. The planned solutions were implemented The company adopted the Six Sigma program and the
during the ﬁrst shift operations of the company. During Chief Broadband Ofﬁcer (CBO) championed the program.
this period, the researcher focused solely on assisting with The program was developed in cooperation with internal
the implementation efforts and documenting the process. resources and by levering best practices from the private
These processes occurred on an iterative basis throughout industry. The program included application of statistical
the study. methods to bring rigor into daily business decision
making. The program was an extensive deployment
3.4. Limitation of case study approach within the company as a ﬁrst attempt to reduce variation
or a form of waste throughout the company. The CBO was
aware that a company-wide improvement strategy ex-
There are three limitations of case study based
tending over several months had to be developed. While
research. First, case study fundamentally assumes that
the team had considerable engineering experience and
an espoused theory should adequately specify action,
qualiﬁcations, cumulatively they did not have the experi-
which is rarely the case (Kemmis and McTaggart, 2000).
ence or the background to deploy a Six Sigma implemen-
Rather, the best theories are parsimonious and do not
tation program and execute the program throughout the
claim to replicate reality. Second, the participants are
company.
exposed initially to the ideas of the researcher which can
The CBO facilitated improvement efforts and estab-
create bias when identifying root causes or explaining
lished change objectives. The researcher-client agreement
events. Three, the conclusions from a single study may
stipulated: (1) that the client was committed to improving
have limited generalizability, and therefore, contributing
existing order processes and was willing to provide all
little to developing or informing a theory. Other research-
pertinent data to the researcher, (2) that the researcher
ers are encouraged to test these ﬁndings by conducting
would document the application improvement solutions
further research in other environments.
in a research log and provide ﬁndings to managers as
requested, (3) that the researcher would provide training
4. Description of implementation experience and participate in improvement efforts in exchange for
publishing the results, and (4) that management would be
4.1. Perform strategic analysis available for at least two hours per day to reﬂect and
provide feedback to the researcher. Following the initial
When customers use their Automated Teller Machine discussion between the researcher and management, a
(ATM) card, several transactions need to be completed cross-function team was established.
simultaneously. These transactions are completed by
Broadband Network Infrastructure Support Center (BNISC)
by performing provisioning translations for ATM custo- 4.2. Form cross-functional improvement team
mers. These translations help complete the transactions
quickly and include assigning logical layer (software) With approval from the CBO, a cross-functional
information for use with the physical layer (equipment) of team was formed, consisting of managers, engineers, and
the desired circuit. In an effort to improve operational consultants. This team was called the ‘‘Management
efﬁciency, an automated system was designed and System of Operational Change’’ (MSOC). While leadership
implemented entitled ‘‘Broadband Network Management skills were considered, eight members of MSOC were
System’’ (BBNMS). This internally developed application primarily chosen for their technical skills. This included
was designed to automate the ﬂow of information from participants from the customer service organization (non-
the customer’s service order through the provisioning access), the circuit fulﬁllment group (access), and the
process. This was a completely manual process prior to application development team that originally designed
implementation. Thus, the primary intent of the BBNMS BBNMS. By deﬁnition, ‘‘access’’ orders were only those
was to eliminate manual work, improve operational circuits that require any type of long distance connection
efﬁciencies, and improve order accuracy. which crosses one or more footprints and do not
One of the features of the BBNMS system was a require a long distance connection. The BNISC had the
functionality set known as ‘‘Request for Manual Activity’’ responsibility to prepare translations for both access/non-
(RMA). An RMA occurred when an order failed to access ‘‘ﬂow-through’’ and access/non-access ‘‘non-ﬂow-
automatically ﬂow through the provisioning system. This through’’ orders. MSOC decided that the ‘‘non-ﬂow
indicated that an error occurred somewhere in the through’’ orders impacted circuits managed out BBNMS
process, thereby indicating a defect. Any time a defect and were therefore not included as a part of initial Six
ARTICLE IN PRESS
6 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

Sigma implementation (deployment). Initially, Six Sigma improvement ideas, such as Lean and Theory of Con-
implementation was to identify ‘‘ﬂow through’’ access and straints principles, were included to enrich the training
non-access orders that generated one or more RMAs, programs (Appendix B). To reinforce understanding
measure the RMAs and their causes, and reduce/eliminate among managers, engineers, and associates in the appli-
RMAs using the Six Sigma tool kit. cation of DMAIC methodology, we developed examples of
MSOC was responsible for developing a charter for all DMAIC applications, and supervised many sessions where
the improvement initiatives, a timeline for carrying out the managers and the workers applied the methodology
the implementation, and a budget to support the changes. to reduce variation or waste from the process.
MSOC was also responsible for analyzing the existing
process, planning the change with active participation
4.4. Execute high-level process mapping and
from all departments, implementing the change in
prioritize improvement
process, establishing performance matrices, and following
up the effects on performance. MSOC began their analysis
MSOC performed a high level process mapping of
by interviewing personnel (customers, engineers, and
information ﬂow with the objective of identifying and
technicians) from different departments in existing pro-
prioritizing improvement opportunities. This was de-
cess and by examining customer complaint reports, cost
signed to understand the existing ﬂow of information
reports, types of RMAs, etc. MSOC initial assessment was
and its associated value creation. While performing
that the deteriorating performance was due to variation in
process mapping, the MSOC discovered numerous exam-
the process and/or many wasteful activities at each stage
ples of variation (wasteful) activities throughout the
of the existing process. The company was determined to
entire order process. Speciﬁcally, MSOC found outdated
ﬁnd a long-term solution that would improve their
information in the systems database, redundant proce-
operations and satisfy their customers. The main objec-
dure, duplicate and confusing paper work, long setup
tives were to improve customer service by eliminating
times for the network, mistakes in order entry, mistakes in
waste from the existing process.
network transfer, etc. The MSOC could not effectively
address all of the problems simultaneously and struggled
deciding which ones to address ﬁrst. MSOC documented
4.3. Choose improvement tools
many wasteful activities from different departments and
segregated them using content analysis. After several
MSOC deliberated to adopt an improvement tool
iterations, they were able to agree to segregate the
and an instrument to document and communicate the
problems based on possible outcomes, i.e. cost savings
implementation of process improvement ideas. The
and/or increased revenue generation. When applicable,
company adopted Six Sigma process improvement in-
the potential cost savings were computed in terms of
itiatives and was keen on a uniﬁed deployment of
reduction in duplicate activities, reduction in returns,
improvement initiative. In the past, the company had
reduction in payroll, etc. When applicable, the revenue
initiated many process improvement or change programs
generation ideas were calculated in terms of increased
(e.g., Total Quality Management, Business Process Reen-
sales, faster response time, and better utilization of
gineering, Team Building, etc.). One problem with these
network bottlenecks. Using the two potential outcomes,
initiatives was that they were isolated in different
MSOC prioritized Six Sigma variation (waste) reduction
departments and rarely worked together to improve the
projects based on those with the strongest support from
performance or the overall value creation process. One
the content analysis (i.e. those that were mentioned
objective of this implementation was to integrate all
most by engineers and managers), keeping in mind the
of these strategies into one simple core strategy to drive
company’s long-term objectives. MSOC then held a meet-
operational excellence from the bottom. MSOC recognized
ing with all managers, including the President, for their
the importance of statistical tools such as Histograms,
input and approval. After deliberations, MSOC developed
Pareto Charts, Statistical Process Control, Regression, and
the ﬁnal priority of all Six Sigma projects to reduce
Design of Experiments (DOE), etc. The general census
variation from the process.
was that the vast majority of the engineers did possess
requisite analytical and/or mathematical skills to fully
understand and implement such tools. 4.5. Develop detailed implementation plan
MSOC worked diligently to develop the contents of
training program with illustrations, examples, and inter- Using the implementation priority, a detailed imple-
active statistical software to provide training in the basics mentation plan consisting of six major activities was
of process improvement. The researcher recommended developed. First, each department was designated to
that Six Sigma training be provided in four different levels setup a lower level improvement team from the depart-
(Appendix A). First level was designed for executive ment. After meeting and discussing with the supervisor, a
overview of Six Sigma for upper level management and team (6–10 engineers and associates) of shop workers
engineers for one day. Second level was designed for (and/or process owners) was established to carry out the
directors and managers working directly as champions of implementation. The entire company was divided into
Six Sigma projects. Third level was designed at the Green 30 teams. Second, once teams were identiﬁed, the next
Belt level, and fourth level was designed at the Black Belt step was to select supervisors or champions for managing
level. In addition, several concepts from popular process the teams during implementation. MSOC spent inordinate
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 7

amounts of time selecting supervisors and champions. associate’s questions openly and honestly and solicited
MSOC believed that the success of the implementation suggestions to improve the process. After several weeks,
was dependent on the champions and their ability to MSOC and the workers were freely exchanging ideas and
manage the teams. While technical skills of champions working with greater cohesiveness.
were required, their human skills were more important
for this role. They came with varying levels of experience
(5–25 years), professional qualiﬁcations (e.g., Electronics 4.6. Implement, document, and revise
Engineering to Human Resources to Sales or Accounting),
and education (e.g., no college degree to Masters in While implementing, MSOC encountered two major
Engineering Management). Following many days of ﬁndings that impacted the implementation of Six Sigma
deliberations, MSOC selected 10 champions or supervisors and its DMAIC methodology. First, when the problem
to manage 30 teams (average three teams per champion). was easy and clearly identiﬁed, DMAIC proceeded in a
After establishing communication protocol with MSOC, sequential and rational manner—beginning with phase
the champions or supervisors were responsible for one and ending with phase four. Although the amount of
implementing Six Sigma programs with the teams. Third, time spent in each stage varied, MSOC experienced the
teams were formed to become completely familiar with succession from one stage to another smoothly. After each
the department and its operations. Each team was iteration of DMAIC, MSOC documented the process,
responsible for obtaining the latest department functional interacted to reﬂect on why the solution worked (or did
ﬂow drawings, the department’s scope or boundaries not work), and tried to identify other areas where the
(where it begins and where it ends), and information ﬂow solution might apply. On reﬂection, the researcher
throughout the department. These teams would walk believed that when the problem was clearly identiﬁed,
through the process and observe information ﬂow and the MSOC and the departments felt that everyone behaved in
use of various sub-systems and tools. Fourth, once a a predictable and rational manner and the company’s
‘‘good’’ understanding of a department and its operations resources (money and personnel) were not wasted. One
were developed, we needed to verbalize the possible example of a Six Sigma project is provided in Appendix C.
problem statement. Typically, the ﬁrst time a problem Second, when a problem was difﬁcult and not clearly
statement is written there is no clear agreement among identiﬁed early in the process, DMAIC proceeded in a
the team members. Fifth, a room was secured for meeting, cyclic and reﬂective manner. In other words, the problem
invitations were sent to management to attend the event, deﬁnition (Deﬁne Stage) may not be complete until
and suggestion boxes were set up near the department. A Analyze Stage. This process appeared counterintuitive
room was secured close to the target department in order and irrational to participants and observers in the process.
for engineers and associates to meet frequently and During this period, the amount of time spent in each stage
openly discuss ways to improvement the process. Sixth, varied greatly. In addition, the succession from one stage
as we know people learn differently (some learn better to other was not smooth. In fact, when consensus could
with text and others learn better with visuals), the not be reached on either the problem or the solution,
training was delivered by MSOC in an interactive manner MSOC proceeded through DMAIC cyclical process (i.e. they
using presentation slides with audio and video enhance- painfully jumped from one stage to other with little
ments. rationale). This resulted in speculation on root causes of
MSOC was aware that in the past many improvement problems and proposed solutions being implemented
initiatives had failed, not because the programs were through ‘‘trial and error’’ with little solid justiﬁcation.
ﬂawed, but because the company did not identify The researcher attempted to guide the efforts; however,
improvement champions and failed to communicate with some members became emotional and would not at ﬁrst
the shop ﬂoor employees. In other words, there was no accept guidance.
general agreement on what to do, which program to use, To stabilize the process, we recommended initially
no adequate training, no communication of what was to delaying deliberation on problems which were not fully
be done, no development of trust between workers and identiﬁed, in favor of addressing those for which the
managers, and no visual support of top management. problem was easier to recognize. During the daily review
The MSOC resolved these critical problems to prepare the sessions, we proposed a revision to the DMAIC training
company for the changes. For example, MSOC knew that program. We suggested revising the DMAIC training to
the workers did not trust the management’s commitment include both sequential methods as suggested in the
and were not interested in participating and giving their literature, as well as cyclical methods. In other words, we
honest responses (or suggestions) to improve the process. proposed an additional one or two-day training session
Over four weeks MSOC held a series of formal and where we would discuss how to address a difﬁcult event
informal meetings with workers from different depart- where the root cause was difﬁcult to identify. Although
ments. During the last week of the meetings, MSOC not speciﬁcally for the DMAIC process, previous research
provided a week-long training session on process excel- indicates that revised training of a cyclical nature does
lence through Six Sigma. By the time they completed their reduce average time to solve problems and drive process
training program, MSOC and workers had met twenty improvement (e.g., Chakravorty et al., 2008).
times. Upper managers were involved at every stage Fig. 1 shows the Six Sigma implementation model.
and exhibited visual commitment to the effort. For the First four steps were: Strategic Analysis; Form High-Level
ﬁrst time in the company’s history, managers answered Cross-Functional Team; Establish Improvement Tools; and
ARTICLE IN PRESS
8 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

Strategic Analysis
Start Market/Customer Driven

Change objectives

Form High-Level Cross
Functional Improvement Team

Develop a charter to guide the improvement initiative
Develop timeline to complete all improvements Strategic Decisions
Establish budget for all the implementation
Establish performance matrices

Establish Overall
Improvement Tools

Develop appropriate training material

Perform a High-Level Process Mapping
Prioritize Improvement Opportunities

Establish communication with employees
Identify improvement champions
Tactical Decisions

Detailed Plan and Form Implementation, Documentation
Low-Level Improvement Teams and Revision

Change training material
Revise implementation plan
Continuous Improvement

Fig. 1. Six Sigma implementation model.

Perform high-Level Process Mapping and Prioritization of (2005, p. 65) point out, that ﬁrst step in Six Sigma
Improvement Opportunities. These steps were considered implementation:
strategic decisions, implying a top down approach, where
management was primarily involved in decision making. yis voice of customer. Yet, in practice is focus often
Last two steps were: Detailed Plan and Form Low-Level perverted or diluted. A survey conducted by Greenwich
Improvement Teams; and Implementation, Documenta- Associates. Strangely absent from most user responses
tion, and Revision. These steps were considered tactical was any mention of the customer. When asked on
decisions implying bottom up approach where engineers an unaided basis to deﬁne a successful project, only
or technicians were primarily involved in decision mak- three out of the 13 companies mentioned customers
ing. as critical success factors, despite the ﬁrst rule of Six
Sigma—listening to the voice of customer.

5. Implications of Six Sigma implementation In our implementation experience, customer expecta-
tions were fairly evident. Often, this may not be the case,
There are several important points worth discussing and we may have to employ rigorous techniques, such as
about the implementation model. The ﬁrst step of the Quality Function Deployment (QFD) or traditional orga-
model is to perform Strategic Analysis, which needs to be nizational assessments. According to Evans and Lindsay
market/customer driven. Our implementation experience (2008), the QFD approach incorporates customer require-
shows that the reason for Six Sigma implementation was ments and performs competitive evaluation to identify
to improve customer expectations through operational design requirements which drives the design of produc-
excellence. Many Six Sigma programs are implemented to tion systems. Many companies such as Toyota, General
gain operational efﬁciency. Unfortunately, many of these Motors, Ford, and Mazda have utilized a QFD approach
operational gains do not directly provide enhanced to plan and design their products. Recent case studies
customer satisfaction or value. Bendell (2006) claims (e.g., Hsieh et al., 2007; Edgeman et al., 2005) involving
that Six Sigma is a strategic approach and improvement real world application, have found that a QFD approach
projects should be selected based on improving customer could be applied to orchestrate Six Sigma implementation.
satisfaction and operational efﬁciency. In reality, a More research is necessary to connect the QFD approach
majority of the improvement projects are selected based to pinpoint customer expectations, and Six Sigma im-
on cost minimization and, therefore, the approach plementation. In addition, traditional organization assess-
becomes suboptimal. According to Andel (2007, p. 1) the ment-Strengths, Weaknesses, Opportunities, and Threats
cost minimization approach usually translates into a (SWOT) analysis can be performed to realize optimum
cutting headcount exercise and when that happens, ‘‘any alignment in a business situation (Andrews, 1971). Cheng
program, Lean or Six Sigma is bound to meet stiff, (2006) claims that by performing organization assess-
self-protecting opposition.’’ As Zimmerman and Weiss ment or SWOT analysis, companies can assess their
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 9

current situation and develop a strategy for Six Sigma analytical and/or mathematical skills to fully understand
implementation. More research is necessary to support and implement such tools in improving operations.
this claim. The projects required application of statistical tools. It
The second step of the model is to Form High-Level is important to understand that often improvement
Cross-Functional Improvement Team. Our implementation projects do not require rigorous statistical tools. As Gopal
experience shows that a high-level cross-functional team (2008, p. 1) says:
was effective in guiding Six Sigma initiatives and keeping
management involvement. Previous research supports our Although Six Sigma as a methodology boasts a multi-
implementation experience. According to Jepperson tude of robust statistical tools, one needs to be wary of
(2005) Raytheon launched their Six Sigma transformation getting trapped by these tools. The practitioner needs
by forming a team consisting of experts from different to use these tools to drive improvement and not the
parts of the organization. While implementing a Six Sigma other way round. In fact, according to one survey, 60 to
program in an Automotive company, Kumar et al. (2007, 70 percent of improvement projects do not require
p. 862) found that: advanced statistical tools. While driving projects, the
larger goal of improvement and the essence of the
yThere was also a need to have effective organiza- methodology should be kept intact. Using tools just for
tional infrastructure in place to support the Six Sigma the sake of using them because they are part of the Six
introduction and development program within the Sigma toolkit is not smart or efﬁcient in terms of time.
organization. Thus, a cross-functional deployment Consistent with our implementation experience, there
team comprising people from middle-level manage- is a good bit of confusion among companies attempting to
ment (executives from Department, Quality Assurance select a particular improvement tool such as Six Sigma,
Department, Sales and Marketing Departments, etc.) Lean, or Theory of Constraints for their speciﬁc business
and workers involved on the ﬂoor were formed. The Six environment. The problem is that managers waste their
Sigma initiative was led by Divisional Manager of the time in adopting the newest improvement tool, using
Foundry Shopy . This was followed by formation of y consultants, and spending very little time in driving
other team members who represented different de- process improvement from the bottom. In other words,
partments of the organization. they do not develop deep problem solving capability
inside their companies. Hayes et al. (2004) argue that
There are two points worth mentioning regarding a improvement tools fail to show a signiﬁcant increase in
high-level, cross-functional team to guide Six Sigma performance because prior to implementing them; most
implementation. First, in choosing members of the team; companies do not develop deep problem-solving capabil-
leadership skills were considered, but no formal psycho- ities in their employees. They refer to the development of
logical proﬁles were considered. There are many such these capabilities as ‘‘Learning by Doing’’. As the name
tests available. See, for example, Stevens and Campion’s implies, managers and workers must practice the art and
(1994) ‘‘Knowledge, Skill, and Ability Test,’’ Kembel’s science of problem solving to use it effectively. Further,
(1996) ‘‘Rational, Organized, Loving, and Energized Test,’’ they write that ‘‘yyou are not likely to be able to make
and Kolbe’s (1994) ‘‘Measure for Instinctive Behavior of substantial progress through analysis and detailed im-
Individuals,’’ which could assist in quick decision-making. provement programs, no matter how many consultants
While theoretical research (e.g., Askin and Huang, 2001) you hirey’’ There is some research on how to develop
provides strong argument for the consideration of psy- deep problem-solving capability in various real-world
chological factors when introducing change programs; settings (e.g. Chakravorty et al., 2008) and why this
additional insights are necessary in this area. Second, the capability is helpful in implementing Lean (Chakravorty
researcher used his implementation experience to deter- and Hales, 2008) and Six Sigma (Chakravorty and Franza,
mine a maximum size of eight to lead Six Sigma 2009). Instead of developing sophisticated statistical
implementation. There is no research to suggest the tools; future research needs to develop methods to
optimal number of team members in a steering commit- implement deep problem solving capability inside com-
tee. We believe that many factors such as the business panies.
environment, the complexity of implementation, and the The fourth step of the model is to Perform a High-
skill levels of teams will play an important role in Level Process Mapping and Prioritize Improvement Opportu-
determining the optimal number of team members. More nities. Previous research supports our implementation
research is necessary to support this claim. A previous experience. For example, according to Mader (2008) a
claim (e.g., Kleasen, 2007) has emphasized the need to company’ process converts inputs (e.g., material, informa-
conduct research on the role and the appropriateness tion) into outputs desired by the customers (products
of cross-functional improvement team for Six Sigma or services). A typical approach is to ﬁrst identify
implementation. processes within the company and then perform project
The third step of the model is to Establish Overall selection and their prioritization. Thiraviam (2006, p. 40)
Improvement Tools. In our implementation Six Sigma said that:
training program for employees included rigorous statis-
tical tools such as Statistical Process Control, etc. Since The ﬁrst tool we used in our case was process mapping.
this was primarily an engineering company, vast majority In our Six Sigma based technique, the ﬁrst step was to
of the engineers or technicians did possess requisite understand, identify and deﬁne all opportunities
ARTICLE IN PRESS
10 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

for failure in a given manufacturing process. and Grow, 2007). According Angel and Pritchard
We mapped processes from incoming inspection (2008, p. 41):
through shippingy .
Examples like Home Depot and 3 M show that
We experienced considerable difﬁculty in identifying companies cannot focus on implementing Six Sigma
and prioritizing Six Sigma projects in alignment with in isolation. Clearly Six Sigma is set of process tools
overall process improvement objectives. In order to that should be part of a more holistic process
prioritize our projects, we used cost reduction and improvement strategy. For any of these tools to be
revenue generation (or throughput based) potential, along used effectively, employee behavior change must be an
with input from workers and management. It is important integral part of the programy . A behavior-focused
to learn more about how to identify projects and how to approach makes change sustainabley Further, it keeps
prioritize them. One reason many Six Sigma improvement us ever aware that a technically sound change designed
programs fail is because improvement projects are not by Six Sigma, lean or similar applications could be at
correctly identiﬁed and prioritized (Zimmerman and risk of failing unless supported by the appropriate
Weiss, 2005). Over the years, many researchers have behavior change.
worked on prioritizing improvement projects by mixing
tools such as Six Sigma, Quality, Lean, or Theory of Previous research (e.g., Chakravorty and Hales,
Constraints tools. For example, Chakravorty and Atwater 2008) has shown that effective management of human
(1998) showed how to prioritize quality improvement aspects is critical for a successful lean implementation.
projects using Theory of Constraints. Chakravorty and Zimmerman and Weiss (2005) point out companies
Sessum (1995) showed how to prioritize Lean improve- need to pay attention to the human side of Six Sigma
ment projects using Theory of Constraints. Chakravorty implementation. The human side of Six Sigma implemen-
(1996) mixed Lean and Theory of Constraints concepts to tation is an important area for future research.
improve the performance of manufacturing operations. This research will be greatly helpful for practicing
Recent empirical research (e.g., Banuelas et al., 2006) managers wanting to effectively implement Six Sigma
found that companies prioritize improvement initiatives programs to achieve sustained results in their business
by mixing these tools. More research is necessary on how environment.
to mix these tools to correctly identify and prioritize
improvement projects. 6. Conclusion
The last two steps of the model are to Develop Plan
and Form Low-Level Improvement Teams; and, Implementa- Using a successful Six Sigma program in a Network
tion, Documentation, and Revision. In addition to having Technology company, the purpose of this research was to
a detailed plan for implementation, our Six Sigma develop an implementation model which consists of six
implementation was successful because we developed steps. The ﬁrst step is to perform a strategic analysis
an appropriate human infrastructure to drive improve- which is customer/market driven. The second step is to
ment from the bottom up. Our approach was to establish a high-level, cross-functional team to drive the
establish low-level improvement teams in each depart- improvement initiative. The third step is to identify the
ment and choose champions or supervisors to manage overall improvement tools. The fourth step is to perform
the teams. In addition to communicating with the high-level process mapping and to prioritize improve-
management; champions or supervisors worked with ment opportunities. These four steps are considered
the improvement teams on various Six Sigma projects. strategic decisions implying a top down approach where
This approach is well supported in the literature. management was primarily involved in decision making.
For example, Carlson and Wilmot (2006) claim that The ﬁfth and sixth steps are to develop a detailed plan and
low-level improvement teams and champions are form low-level improvement teams and to implement,
essential to drive improvement projects from the bottom document, and revise as needed. The last two steps are
up. According McManus (2007, p. 20) for Six Sigma considered tactical decisions, implying a bottom up
implementation: approach where engineers or technicians were primarily
Every organization should have a process improvement involved in decision making. In addition, important for
team approach in place that engages a high percentage both practitioners and academicians, several areas of
of the work force. Such an approach is mandatory for future research are also discussed regarding the imple-
sustained progress toward process excellencey If you mentation model.
want to sustain your improvement efforts over timey Lastly, this research provides a model to effectively
give a training [to the improvement teams], and turn guide the implementation of Six Sigma programs
teams loose. to reduce variation or waste from the operations.
This is particularly relevant because today’s competitive
There is growing concern that Six Sigma or other environment demands that companies reduce variation
process improvement programs fail because they (waste) to meet or exceed efﬁciency and responsi-
do not consider the human side of implementation. veness requirements of customers. There is increasing
For example, Six Sigma implementation negatively pressure to pursue new ways of thinking as a source
affected employee morale at Home Depot and stiﬂed of competitive advantage. More research in this area is
creativity and innovation at3M (Hindo, 2007; Hindo necessary to contribute to the science and practice of
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 11

implementation of Six Sigma or any other process MEASURE
improvement model, to reduce waste and create Establish performance metrics
Develop data collection and sampling plan
value.
Conduct measurement systems analysis
ANALYZE
Appendix A. Six Sigma training Benchmark the process or product
Analyze the process map theory of constraints
Determine and summarize root cause
IMPROVE
Executive overview for Six-Sigma Determine breakthrough solutions, costs and beneﬁts
Maximum of 5 attendees—1 day course Implement and measure solution effectiveness
This course is for people from service backgrounds. It is designed for CONTROL
executives working directly with Champions, Black Belts, and Green Determine type of control needed
Belts on Six-Sigma projects. Implement and validate controls

This course includes an overview of Six-Sigma concepts, including: Black Belt—process excellence for Six-Sigma
What is Six-Sigma? 5–10 attendees—2 week course
How Six-Sigma ﬁts into your company?
Integration of Lean manufacturing into Six-Sigma deployment
DMAIC road map
Expected beneﬁts of Six-Sigma This course focuses on product production excellence. It is designed for
Management infrastructure for Six-Sigma those individuals working directly on Six-Sigma projects as Black or
Business goals, objectives, and dashboards for Six-Sigma Green Belts.
DMAIC breakthrough methodology This course covers Six-Sigma philosophy and tools. Students will learn, at
Implementation issues and timelines both a basic and statistical level, the DMAIC methodology and apply it to
Six-Sigma deployment model their manufacturing projects.
Project selection and Theory of Constraints
Black Belt selection
Reviewing progress and maintaining accountability DEFINE
Next steps Deﬁne resource/stakeholder analysis
Project leadership
MEASURE
Champion training for Six-Sigma Determine critical Xs and Ys
5–10 attendees—2 day course Understand foundation of statistics
This course is for people from service backgrounds. It is designed for ANALYZE
directors and managers working directly as champions of Six-Sigma Establish causal relationships using regression analysis
projects. Visualize the problem and determine root cause
IMPROVE
Measure solution effectiveness
Rapid DOE (Design of Experiment)
This course includes an overview of Six-Sigma concepts, including: CONTROL
What is Six-Sigma? Statistical process control
Expected beneﬁts of Six-Sigma Develop transfer plan and close project
DMAIC breakthrough process
Management and organizational infrastructure for Six-Sigma (1) Keller, P., 2005. Six Sigma: Demystiﬁed. McGraw-Hill. New York, NY.
Project identiﬁcation and prioritization (2) Pyzdek, T., 2003. Six Sigma. Handbook. McGraw-Hill. New York, NY.
Financial analysis
Consider painful processes
Examining areas of waste
Look at the cost of quality
Customer satisfaction
Appendix B. Training program enhancements
Link projects to strategic objectives
Prioritizing projects and Theory of Constraints
DMAIC breakthrough methodology and deliverables
Black Belt and Green Belt selection and management Lean Operations principles categorize activities into value added and
Six-Sigma implementation roadmap non-value added activities. The focus of these principles is on
eliminating all forms of non-value added activities. There are seven
forms of non-value added (or waste): waste of over production, waste of
inventory, waste of defects, waste of motion, waste of processing, waste
Green Belt—process excellence for Six-Sigma
of waiting, and waste of transportation. Lean Operations tools include:
5–10 attendees—1 week course
This course focuses on product production excellence. It is designed for
those individuals working directly on Six-Sigma projects as Green Belts. Value stream mapping
Lean events
SMED (a.k.a quick changeover)
Manufacturing cell design
This course covers the Six-Sigma philosophy and tools. Students will Poka-Yoke (a.k.a. mistake prooﬁng)
learn the DMAIC methodology at a basic level and apply it to their 5S plus one (Sort, Straighten, Shine, Standardize, Sustain, and Safety)
manufacturing projects.
Theory of Constraints (TOC) challenges the belief of functional
independence with the view of an organization as a chain of dependent
DEFINE activities. According to TOC, every organization has a constraint or
Deﬁne problem, objective, goals, and beneﬁts bottleneck and the existence of a bottleneck provides an opportunity for
Develop project plan and map the process to identify waste process improvement priority. In addition, TOC provides a set of
ARTICLE IN PRESS
12 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

measures that link local actions to the measures of organization wide the service orders and due to critical role in the process
measures. TOC tools include: also had two other RMA types that would be in scope:
Bottleneck analysis cirmisact and unknactivity. The second highest overall and
Drum buffer rope
Buffer management the highest RMA for Systems Management (SM) was
Buffer sizes and locations interprocesserr. Finally, the largest RMA for Service Order
Throughput maximization Management (SOM) was rcidnotfnd. This rcidnotfnd type
VAT operations analysis was later eliminated from consideration due to a system
dependency. Two RMA types were deemed redundant
(1) Womack, J.P., & Jones, D.T., 2003. Lean Thinking. Free Press, New York, (ddmissed and dvamissed) and other there were thrown
NY. out due to small number of observations. Table 1 provides
(2) Goldratt, E.M., & Cox, J., 2004. The Goal. North River Press, Great the detailed log of RMAs for the weeks speciﬁed and the
Barrington, MA.
Pareto Chart that summarizes the observation data over
the same period (Fig. 2).
The team initially reviewed raw RMA data but after
considering counting constraints it was decided that
Appendix C. Implementation of a Six Sigma project simplifying the measure to RMA per Service Order
(RMA/SO) would produce the most meaningful results.
C.1. Deﬁne The project objective was changed to reduce RMA/SO by
50% by December 31. The team calculated the expected
During scope deﬁnition, the project team evaluated all objective by averaging four weeks of data after all system
Requests for Manual Activity (RMA) categories and related functionality was put into place. The project used a
sub-groups which deﬁne the group owner. After tracking baseline of September 29. The team proceeded by
and recording RMA events over a week period, the team counting two weeks past the baseline order to ensure
was ready to review the results to determine the proper that the numbers included all the data. This was required
course of action to drive corrections. The team elected to since RMA/SO are measured on service order due date
pursue the top RMA generates for each sub-group and the which typically occurs within two weeks of the service
next two RMA types with the highest contribution. order submission. The resulting baseline was 4.6 RMA/SO
Overall, the highest RMA was missﬁd which impacted calculated between 10/13 and 11/7. Therefore, the objec-
the Service Order Processing (SOP) team. This team wrote tive was established at 2.3 RMA/SO.
Simple count, sore, and Pareto analysis were use to
identify the RMA that occurred most frequently. In
Table 1
Initial RMA observation count. addition, ﬁeld experience was a determining factor in
evaluating which RMA took the longest to resolve. This
RMA type RMA sub-group 9–22 9–23 9–24 9–25 9–26 Total practice was input to RMA priority which then honed
further using root cause analysis. After the root causes
missfd SOP 214 176 510 308 264 1472
interprocessserr SM 85 48 72 76 66 347 were isolated, and improvement plan was designed
ddmissed Date 0 85 134 83 57 359 based on the priority and system capabilities. The project
unknactivity SOP 104 63 47 41 38 293 was subject to following complications and constraints
pvcpserr SOM 14 29 19 5 14 81 (Table 2).
cirmisact SOP 72 38 36 67 18 231
rcidnotfnd SOM 25 28 46 30 30 159
dvamissed Date 0 13 31 14 17 75 C.2. Measure
socpxfail SOM 40 26 22 22 28 138
iportpserr SOM 2 5 8 4 8 27
.... .... .... .... .... .... .... .... In order to fully understand the RMAs, the team
Total ALL RMAs 739 669 1099 807 647 2489 prepared BBNMS process ﬂows. This permitted the team
to conceptualize breakdowns and refer to the correction

3000 1.00
0.90
2500 0.80
2000 0.70
0.60
1500 0.50
0.40
1000 0.30
500 0.20
0.10
0 0.00
d

ity

d
rr

rr
ai
ac
sf

tfn

se
se

se
iv

ps
is

is
is

is
no
t
ss

rtp
cp
ac
m

c
m

am
ce

pv
so

o
kn
dd

ip
rc

dv
ro

un
rp
te
in

Fig. 2. Pareto chart.
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 13

Table 2 time recorded for repair. The team then determined which
System complications and constraints. RMAs to measure by examining the RMA histogram. Upon
reviewing the special causes, it was fund that interpro-
Condition Description
cesserr RMA contained embedded error codes. The other
Redundant RMA counts ddmissed (due date missed) and three RMA types (missﬁd, cirmisact, and unknactivity) did
dvamissed were that were not contain this information and manual investigation
generated because of other RMAs was required.
Initial delay due to system All states were not initially
unavailability represented
The team then conducted a detailed review of each of
Total service orders not counted by Sigma number unavailable the four remaining RMAs. The sample size was deter-
OMS (Order Management System) mined by taking each RMA, beginning and missﬁd,
RMA/SO numbers based on RMA To be determined calculating the number of service orders with that RMA
due date (dd) vice when the RMA
in one week, and comparing that total with the total
initially falls out
Team member availability Workload dependent and not 100% number of service orders. Table 3 shows the sample size
Accurate time measurements Extreme variability in manual (n) was derived.
rework Each of the RMA data sets was randomly sampled by
Existing records issue Addressed in improve phase computer for 62 unique entries for missﬁd, cirmisact, and
interprocesserr. Since the data set was small for unknac-
tivity only 40 samples were taken. Once the service order
Table 3
Sample size calculations. samples were identiﬁed, the results were tabulated
in excel. The summarized results from unknactivity are
Week Total# SOs #missﬁd SOs Z value (90%) p c n shown in Table 4.
The interprocesserr RMA differs from the service order
11–3 2860 1033 1.64 0.36 0.1 62
management and processing RMAs. This type occurs when
BBNMS encounters an unexpected event, something that
Table 4 was not previously deﬁned. Since these RMAs related to
Unknactivity results. developed software, the developer devised a method for
identiﬁcation and tracking of these speciﬁc cases. In a
Order problem Quantity Percentage (%)
similar fashion as the other RMA type, the team collected
Circuit 6 15.00 and prioritized the data by frequency of receipt. The total
Records 11 27.50 service orders were available for interprocesserr RMA so U
Special assembly 10 25.00 Charts were generated based on a two week period
Missing action 13 32.50
Total 40 100.00
calculation. Only the top ten interprocesserr error codes
included amounting for 86.99% of all errors (Figs. 3–6).

7
6
5
4
3
2
1
0
70-153 154-237 238-320 321-404 405-488
missfid RMAs 10/13-11/7

Fig. 3. Missﬁd histogram.

9
8
7
6
5
4
3
2
1
0
32-65 66-98 99-132 133-165 166-199
cirmisact RMAs 10/13-11/7

Fig. 4. Cirmisact histogram.
ARTICLE IN PRESS
14 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

12
10
8
6
4
2
0
19-65 66-112 113-158 159-205 206-252
intrprocesserr RMAs 10/13-11/7

Fig. 5. Interprocesserr histogram.

9
8
7
6
5
4
3
2
1
0
2-16 17-30 31-45 46-59 60-74
unknactivity RMAs 10/13-11/7

Fig. 6. Unknactivity histogram.

0.080
0.070
UCL
0.060
0.050
0.040
0.030
0.020
0.010
LCL
0.000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
endpoint collision 11/17-12/12

Fig. 7. End point U chart.

Table 5 condition/constraint and was caused by BBNMS procedur-
Improvements. al change in how a service order is prepared. It was
determined that this information would be noted as a
RMA type SS team Other Total
initiated initiated improvements change to the order preparation process. This ﬁeld is
necessary input into the automated ﬂow through process
missﬁd 7 3 10 and will be listed as a required ﬁeld. This change would
cirmisact 6 0 6 only beneﬁt future orders. The other RMA categories
unknactivity 5 0 5
interprocesserr 5 3 8
include in-ﬂight software code changes (36.33%), pending
future change requests (9.34%), and training issues
(12.46%).
The RMA cirmisact was similar to missﬁd as 38.81% of
C.3. Analyze the samples were records related. In this instance, a
greater portion of the sample of RMAs has either had edits
Upon analyzing the missﬁd RMA type and categorizing completed or in-process (44.78%). The remaining issues
the results, the data indicated that the majority of orders included training (5.97%), future software release (4.48%),
(41.87%) produced RMAs due to an existing records issue. parser sequencing (2.99%), and unidentiﬁed problems
This issue was originally identiﬁed as a pre-existing (2.99%).
ARTICLE IN PRESS
S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16 15

Table 6
Cost of quality.

08/06-01/02 RMA RMAs 10/ RMAs # Sos Min/ Total Total Total Avg cost Cost 10/ Cost 12/ Six Sigma Annualized Six
Total 13-11/7 12/8-1/2 (Tot/ Ord Min ($) Hr ($) cost ($) RMA ($) 13-11/7 ($) 8-1/2 ($) savings ($) Sigma savings ($)
2.85)

cirmisact 4634 1965 289 1626 12 19,512 325 14,006 5,939 873 5,066 20,263
interprocesserr 7399 1558 651 2596 25 64,904 1,082 46,590 9,810 4,099 5,711 22,845
missﬁd 15007 4571 1890 5266 12 63,187 1,053 45,358 13,816 5,712 8,103 32,413
unknactivity 289 492 285 1017 12 12,206 203 8,762 1,487 861 626 2,503
Grand total 27329 8586 3115 10505 61 159,809 2,663 114,716 31,052 11,547 19,506 78,023

The RMA unknactivity was also comparable to missﬁd References
and cirmisact where the records issue was signiﬁcant
accounting for 27. Zero percent of the random sample. The Adams, C., Gupta, P., Wilson, C., 2003. Six Sigma Deployment. Butter-
remaining issues included training (32.5%), special as- worth-Heinemann, Oxford, UK.
Andrews, K.R., 1971. The Concepts of Corporate Strategy. Dow Jones-
sembly (25%), and circuit order-related (15%). Irwin, New York, NY.
The interprocesserr RMA analysis evaluated the histo- Andel, T., 2007. Lean & Six Sigma Traps to Avoid. Material Handling
grams from the top ten error codes. As seen in other Management, March /http://www.mhmonline.com/viewStory.asp?
nID=5423S.
RMA types, a large amount of special cause variation was
Angel, D.C., Pritchard, C., 2008. Behavior tests Six Sigma. Industrial
evident. This variation was captured using Poisson U Engineer (August), 41.
Charts. The data was then analyzed per error code and a Antony, J., Antony, F.J., Kumar, M., Cho, B.R., 2007. Six Sigma in service
organisations: beneﬁts, challenges and difﬁculties, common myths,
sample U Chart from the most frequently reported error,
empirical observations and success factors. International Journal of
end point collision, is shown in Fig. 7. Quality & Reliability Management 24 (3), 294–311.
Improve: One of the biggest challenges during project Antony, J., Banuelas, R., 2002. Key ingredients for the effective
execution was managing change requests to incorporate implementation of Six Sigma program. Measuring Business Excel-
lence 6 (4), 20–27.
new edits, defects, and other desired improvement Antony, J., Kumar, M., Madu, C.N., 2005. Six Sigma in small- and medium
actions. Table 5 summarizes the types of changes and sized UK manufacturing enterprises: Some empirical observations.
references where they were initiated or inﬂuences by the International Journal of Quality & Reliability Management 22 (8),
860–874.
Six Sigma project team or from an external source. All Askin, R.G., Huang, Y., 2001. Forming effective worker teams for cellular
changes summarized in this table were acted upon so 29 manufacturing. International Journal of Production Research 39 (11),
total changes were introduced from this initiative. Overall, 2431–2451.
Banuelas, R., Tennant, C., Tuersley, I., Tang, S., 2006. Selection of Six
the project delivered a 52.08% process improvement from Sigma projects in the UK. The TQM Magazine 18 (5), 514–527.
baseline to completion. Bendell, T., 2006. A review and comparison of Six Sigma and the lean
Control: While the project ended, the team instituted a organizations. The TQM Magazine 18 (3), 255–262.
Berg, M., 2006. Six Sigma shortcomings. Industrial Engineer (October),
lifecycle improvement process to monitor activities for
45.
future identiﬁcation and elimination of RMAs. A workﬂow Breyfogle, F., Cipello, J., Meadows, B., 2001. Managing Six Sigma. Wiley
analyst position was created to track RMA numbers Inter-Science, New York, NY.
Coronado, R.B., Antony, J., 2002. Critical success factors for the successful
on a daily basis and identify and investigate when
implementation of Six Sigma projects in organisations. The TQM
RMA numbers fall outside the control limits. In addition, Magazine 14 (2), 92–99.
a lifecycle manager was given responsibility to lead Carlson, C.R., Wilmot, W.W., 2006. Innovation: The Five Disciplines for
regular recurring service order meetings to manage the Creating What Customer Want. Random House, Inc., New York, NY.
Chakravorty, S.S., 1996. Robert Bowden, Inc.: a case study of cellular
work effort related with the identiﬁcation and correction manufacturing and drum buffer rope implementation. Production
of RMA issues. The project enabled the lifecycle team with and Inventory Management Journal 3rdQtr, 14–18.
many new tools in the form of histograms, Pareto Charts, Chakravorty, S.S., Atwater, J.B., 1998. Implementing quality improvement
programs using the focusing steps of theory of constraints.
Statistical Process Controls charts, etc. to assist with this International Journal of Technology Management 16 (4/5/6/),
analysis. 544–555.
Cost of quality: By implementing improvements across Chakravorty, S.S., Atwater, J.B., 2006. Bottleneck management: theory
and practice. Production & Planning Control 17 (5), 441–447.
for RMA types from 29 requests, this project determined Chakravorty, S.S., Hales, D.N., 2004. Implications of cell design imple-
the Total cost of Quality as $318,635 which reﬂects mentation: a case study and analysis. European Journal of Opera-
prevention, appraisal, and internal failure costs (no tional Research 152, 602–614.
Chakravorty, S.S., Hales, D.N., 2008. The evolution of manufacturing cells:
external failure numbers were available). This includes
an action research study. European Journal of Operational Research
costs associated the original process through the comple- 188, 153–168.
tion of the project. Many of these numbers were estimates Chakravorty, S.S., Hales, D.N., Herbert, J.I., 2008. How problem solving
really works. International Journal of Data Analysis Techniques and
since no individual data was recorded. To reﬂect the
Strategies 1 (1), 44–59.
Annualized Total Cost of quality ($479,237), the above Chakravorty, S.S., Franza, R.M., 2009. Implementation of design for Six
number was adjusted in scale from ﬁve months originally Sigma (DFSS): a development experience. International Journal of
represented into 12. The saving generated from this Product Development, in press (Special Issue on: Transformations in
Industrial Design: Emerging Trends, Approaches, and Challenges).
initiative was $19,506 and would be $78,023 if annualized Chakravorty, S.S., Sessum, J.L., 1995. Developing effective strategies to
(Table 6). prioritize set-up reduction in multi-machine production system.
ARTICLE IN PRESS
16 S.S. Chakravorty / Int. J. Production Economics 119 (2009) 1–16

International Journal of Operations and Production Management 15 Kemmis, S., McTaggart, R., 2000. Participatory action research. In:
(10), 103–111. Denzin, N.K., Lincoln, Y.S. (Eds.), Handbook of Qualitative Research,
Chandra, A., 2008. Five Six Sigma deployment mistakes- and how to second ed. Sage Publications, Thousand Oaks, CA, pp. 567–606.
avoid them. /http://www.isixsigma.com/library/content/c080414a. Ladani, L.J., Das, D., Cartwright, J.L., Yenker, R., Razmi, J., 2006.
aspS. Implementation of Six Sigma system in celestica with practical
Cheng, J.L., 2006. DMAIC integration necessary for success. ASQ SIX examples. International Journal of Six Sigma and Competitive
Sigma Forum Magazine 5 (4), 22–28. Advantage 2 (1), 69–88.
Davison, L., Al-shaghana, K., 2007. The link between Six Sigma and Lakhavani, S.T., 2003. Six Sigma implementations: trials, tribulations,
quality culture—an empirical study. Total Quality Management 18 and lessons learned. ASQ’s Annual Quality Congress Proceedings 57,
(3), 249–265. 643–647.
Edgeman, R., Bigio, D., Ferleman, T., 2005. Six Sigma and business Lee-Mortimer, A., 2006. Six Sigma: a vital improvement approach when
excellence: strategic and tactical examination of IT service level applied to the right problems, in the right environment. Assembly
management at the ofﬁce of the Chief Technology Ofﬁce of Automation 26 (1), 10–18.
Washington, DC. Quality and Reliability Engineering International Lynch, D.P., Bertolino, S., Cloutier, E., 2003. How to scope DMAIC projects.
21, 257–273. Quality Progress 36 (1), 37–41.
Evans, J.P., Lindsay, W.M., 2008. Managing for Quality and Performance Mader, D.P., 2008. Lean Six Sigma’s evolution. Quality Progress 41 (1),
Excellence, seventh ed. South-Western, Mason, OH. 40–48.
Feng, Q., Manuel, C.M., 2007. Under the knife: a national survey of Six Mast, J.D., Bisgaard, S., 2007. The science in Six Sigma. Quality Progress
Sigma programs in US healthcare organizations. International 40 (1), 25–29.
Journal of Health Care Quality Assurance 21 (6), 535–547. Mcadam, R., Evans, A., 2004. Challenge to Six Sigma in a high technology
Fontana, A., Frey, J.H., 2000. The interview: from structured questions to mass-manufacturing environments. Total Quality Management 15
negotiated texts. In: Denzin, N.K., Lincoln, Y.S. (Eds.), Handbook of (5/6), 699–706.
Qualitative Research, second ed. Sage Publications, Thousand Oaks, McManus, K., 2007. The trouble with teams. Industrial Engineer 39 (10),
CA, pp. 567–606. 20.
Foster, S.T., 2007. Does Six Sigma improve performance? The Quality Meredith, J., 1998. Building operations management theory through case
Management Journal 14 (4), 7–20. and ﬁeld research. Journal of Operations Management 16, 441–454.
General Electric Company, 2005. Six Sigma Customer Black Belt Training: Meredith, J., Mantel, S., 2003. Project Management: A Managerial
Six Sigma Customer Solutions. Cleveland, OH. Approach. Wiley, New York, NY.
Gijo, E.V., Rao, T.L., 2005. Six Sigma implementation—hurdles and more Mullavey, F., 2005. Shackled by bad Six Sigma. Quality Digest, September
hurdles. Total Quality Management 16 (6), 721–725. /http://www.qualitydigest.com/sept05/articles/03_article.shtmlS.
Gitlow, H.S., Levine, D.M., 2005. Six Sigma for Green Belts and Pande, P.S., Neuman, R.P., Cavanagh, R., 2000. The Six Sigma Way: How
Champions: Foundations, DMAIC, Tools, Cases, and Certiﬁcation. GE, Motorola, and Other Top Companies are Honing their Perfor-
Prentice-Hall, Upper Saddle River, NJ. mance. McGraw-Hill, New York, NY.
Gopal, R., 2008. 10 ways to failure for a new Six Sigma program. /http:// Pyzdek, T., 2003. The Six Sigma Handbook: A Complete Guide for Green
www.isixsigma.com/library/content/c071203a.aspS. Belts, Black Belts, and Managers at All Levels. McGraw-Hill, New
Gupta, P., 2008. Reducing the cost of failures. Quality Digest 47 (1), 22. York, NY.
Hales, D.N., Chakravorty, S.S., 2006. Implementation of Deming’s style of Rajagopalan, R., Francis, M., Suarez, W., 2004. Developing novel catalysts
quality management: an action research study. International Journal with Six Sigma. Research—Technology Management 47, 13–16.
of Production Economics 103, 31–148. Rosenthal, R., Rosnow, R., 1991. Essentials of Behavioral Research.
Harry, M.J., Linsenmann, D.R., 2007. The Six Sigma Fieldbook: How McGraw-Hill Publishers, Boston, MA.
DuPont Successfully Implemented the Six Sigma Breakthrough Savolainen, T., Haikonen, A., 2007. Dynamics of organizational learning
Strategy. Doubleday Business of Random House, Inc., New York, NY. and continuous improvement in Six Sigma implementation. The
Harry, M.J., Schroeder, R., 2000. Six Sigma: The Breakthrough Manage- TQM Magazine 19 (1), 6–17.
ment Strategy Revolutionizing the World’s Top Corporations. Ran- Schonberger, R.J., 2008. Best Practices in Lean and Six Sigma Process
dom House, Inc., New York, NY. Improvement. John Wiley & Sons, Hobokon, NJ.
Hsieh, C.T., Lin, B., Manduca, B., 2007. Information technology and Six Schroeder, R.G., Linderman, K., Liedtke, C., Choo, A.S., 2008. Six Sigma:
Sigma implementation. The Journal of Computer Information deﬁnition and underlying theory. Journal of Operations Management
Systems 47 (4), 1–10. 26 (4), 536–554.
Hayes, R.H., Pisano, G.P., Upton, D.M., Wheelwright, S.C., 2004. Opera- Snee, R.D., Hoerl, R.W., 2003. Leading Six Sigma. Wiley, Hoboken, NJ.
tions, Strategy, and Technology: Pursuing the Competitive Edge. John Stake, R.E., 2000. Case studies. In: Denzin, N.K., Lincoln, Y.S. (Eds.),
Wiley & Sons, Indianapolis, IN. Handbook of Qualitative Research, second ed. Sage Publications,
Hindo, B., 2007. At 3M, a struggle between efﬁciency and creativity. Thousand Oaks, CA, pp. 435–454.
Business Week, June 11. Stevens, M.J., Campion, M.A., 1994. The knowledge, skill, and ability
Hindo, B., Grow, B., 2007. Six Sigma: so yesterday. Business Week, June requirements for teamwork: implications for human resource
11. management. Journal of Management 20 (2), 303–530.
Jacobs, F.R., Chase, R.B., 2008. Operations and Supply Chain Manage- Sutton, C., 2006. Getting the most out of Six Sigma. Quality Management
ment: The Core. McGraw-Hill Irwin, New York, NY. (March), 46–48.
Jaffery, M., 2003. Indian Cooking. Barron’s Educational Series, Inc., Szeto, A.Y.T., Tsang, A.H.C., 2005. Antecedent to successful implementa-
Hauppauge, NY. tion of Six Sigma. International Journal of Six Sigma and Competitive
Jepperson, M., 2005. Focused journey of change. Industrial Management Advantage 1 (3), 307–322.
(July/August), 8–13. Thiraviam, A.R., 2006. Simple tools for complex systems. Quality Progress
Keller, P., 2005. Six Sigma: Demystiﬁed. McGraw-Hill, New York, NY. 39 (6), 40–44.
Kembel, J., 1996. Role Play—Personalities in Action. Life Change Partners, Treichler, D., Carmichael, R., Kusmanoff, A., Lewis, J., Berthiez, G., 2002.
Boston, MA. Design for Six Sigma: 15 lessons learned. Quality Progress 35, 33–42.
Kleasen, K., 2007. Building human resources strategic planning, process Watson, G., 2006. Building on Six Sigma effectiveness. ASQ Six Sigma
and measurement capability: using Six Sigma as a foundation. Forum Magazine 5 (4), 14–15.
Organizational Development Journal 25 (2), 37–41. Wurtzel, M., 2008. Reasons for Six Sigma deployment failures. BPM
Kolbe, K., 1994. Pure Instincts. Random House, New York, NY. Institute. June /http://www.bpminstitute.org/articles/article/article/
Kumar, U.D., Nowicki, D., Ramirez-Marquez, J., Verma, D., 2008. On reasons-for-six-sigma-deployment-failures.htmlS.
the optimal selection of process alternatives in a Six Sigma Yin, R.K., 2008. Case Study Research: Design and Methods. Sage
implementation. International Journal of Production Economics 111, Publications, Thousand Oaks, CA.
456–467. Zimmerman, J.P., Weiss, J., 2005. Six Sigma’s seven deadly sins. Quality
Kumar, M., Antony, J., Antony, F.J., Madu, C.N., 2007. Winning customer 44, 62–66.
loyalty in an automotive company through Six Sigma: a case Zu, X., Fredendall, L.D., Douglas, T.J., 2008. The evolving theory of quality
study. Quality and Reliability Engineering International 23, management: the role of Six Sigma. Journal of Operations Manage-
849–866. ment 26, 630–650.