Professional Documents
Culture Documents
Six Sigma
Six Sigma
With increasing competition and emerging global markets, the pressure to provide higher
quality products and better customer satisfaction has increased, hence increasing workload but
necessitating lower costs. Therefore, it is progressively difficult to be a competitive company as
there are fewer resources available. Six Sigma can help organisations learn from and excel at the
challenges they meet. It is a tactical methodology, which decides the best approach for a given
situation or process.
The success of Six Sigma is linked to a set of functional metrics that lead to significant
improvements in customer satisfaction and bottom-line benefits. Some employees of a future
Six Sigma organisation can view Six Sigma quality as a rigorous application of basic and
advanced statistical tools throughout an organisation. They can also believe that it is a group of
statistical components of old Total Quality Management (TQM) programs with the name of
Six Sigma. Others can regard Six Sigma as a refined version of TQM. Six Sigma emphasises
the knowledge of the organisation with established statistical tools to improve both the
efficiency and effectiveness of the organisation in order to meet customer needs and
requirements. Organisations do not necessarily need to use all the measurements described in
this project. It is most important to choose the best set of metrics for a situation, metrics that
yield insight into a situation or process.
Six Sigma is a business process improvement strategy, which essentially checks that a business
is doing what its customers require, that the process employed is the most effective known and
that variation is removed from that process to provide significantly improved consistency. Much
1
of the initial success in the application of Six Sigma was based on manufacturing applications;
however, Six Sigma can also be applied in the services and also banking industries.
As clients place a high value on consistent business processes, which have been proven to be
efficient, Six Sigma would meet their requirements as it is a proven methodology for delivering
consistent incremental improvement. By reducing process variation, Six Sigma allows the
organisation to focus on improving process capability and as Sigma levels increase, the cost of
poor quality decreases and profitability increases.
The principal purpose of this project is to review the Six Sigma implementation process and
emphasise the importance of this methodology in the quality area. For this reason, its main
framework is based on the description of the Six Sigma process. The second chapter describes
the significance of Six Sigma, its statistical interpretation, the team members and their
responsibilities and some of the benefits and limitations of this methodology. The third chapter
illustrates the implementation process and the tools, which can be used in order to achieve highquality results. The fourth chapter is centred on showing the benefits and difficulties that can
occur in the implementation process in some published case studies. In order to show that Six
Sigma can be used in different sectors, the last chapter presents the application of Six Sigma in
the financial services industry, with a focus on banking
Sigma
(a)
DPM
(b)
yield (%)
DPM
(c)
Sigma
(d)
yield (%)
Figure 1. Relationship between sigma, DPM, yield and the number of sigma.
Sigma measures demand a clear definition of what the customer requirements are. This helps
the company to realise what is important and what need to be improved. With their focus on
defects and defect opportunities, sigma measure can be used to analyse and compare different
process inside an organisation or between organisations. Once that have been identified the
opportunity area and the requirement, the organisation can define the defect and the sigma level
which wants to reach. Table 1 lists the sigma level against the number of defects per million of
opportunities, it shows that the number of defects decreases exponentially as the sigma level
increases.
DEFECTS PER
SIGMA LEVEL
MILLION
1
690.000
308.537
66.807
6.210
233
3,4
The output from manufacturing processes is liable to change over time. The mean or setting of
the process could shift or the standard deviation might change increasing the spread of the
process. The result of this is that a part of the tail of the curve would extend beyond the sigma
limits. This means that the sigma level is smaller than the company expects. In practice it is
devalued by 1.5 which is an arbitrary value based on manufacturing experience. If a company
has an X sigma level in the short term it will have an (X 1.5) sigma level in the long term.
Six Sigma is not an exception to this rule, as Figure 2 shows. The blue curve is centred between
the limits, with a mean of zero and a standard deviation of one and represents the short term
process sigma. The green one has exactly the same standard deviation of 1 but the mean has
shifted to the left by 1.5 and so part of the now falls beyond the LSL.
1,5s
LSL
6s
USL
Many people think that 99% quality and therefore 1% defectives is a high level to obtain.
However, the 1% can mean the failure of the company. Table 2 lists a comparison of some
processes evaluated by Six Sigma and by 99% quality. The results show that 99% quality is not
enough in some cases. The table shows that it is necessary to implement Six Sigma as a tool for
reducing defects and increasing customer satisfaction.
99% Quality
Six Sigma
The sigma level for airline company safety is 6.5 sigma, which means less than one-half of a
failure per million flights; in contrast, airline baggage operations work at 3.5 Sigma. People fly
by plane because it is a safe transport medium but a piece of lost luggage is tolerated.
In order to understand better the significance of being a Six Sigma organisation it is necessary to
analyse the connotation of Six Sigma.
Six Sigma is an enterprise process, which enables companies to improve their final
results, design, and supervise their usual activities, minimising waste and resources
and increase costumer satisfaction.
It allows enterprises to reduce the number of defects in all their outputs and
eliminate quality error.
The Six Sigma methodology not only enables errors to be detected and corrected, but it
also contributes methods to create new processes, which ensure that the error does not
occur again.
The Six Sigma is a business strategy, which seeks to identify and eliminate causes
of error or defects or failures in business processes by focusing on outputs which
are critical to the customer.
Once Six Sigma have been defined is necessary to understand the importance of implementing
Six Sigma, the different models which can be used and the teams structure and responsibilities
of the team members which is analysed in the next stage.
In the 1980s, Total Quality Management (TQM) was very popular but its impact has begun to
wane. A new methodology was necessary and Six Sigma is emerging as the new way.
Six Sigma focuses on reducing process variation and then on improving process capability and
in order to achieve this uses two improvement models based on the PDCA cycle, introduced by
Shewhart and applied the first time for Edwards Deming, which describes the logic of data
process improvement. The PDCA cycle is based on the next phases:
The two Six Sigma models, which are used in order to reduce the number of defects and to meet
with the customers requirements, are DMAIC and DMADV. The Design for Six Sigma process
is known as DMADV - Define, Measure, Analyse, Design and Verify. This process is generally
used for product design before manufacturing. The DMAIC process is used to refine the
manufacturing process through reduction of common cause variation. The DMADV process is
designed to remove variation between what the customer wants and what actually was designed.
The DMADV methodology should be implemented when a service or product needs to be
developed in a company where does not already exist or it does not reach the specific sigma
level. The phases of DMADV are:
5.CONTROL
1.DEFINE
4.IMPROVE
2.MEASURE
3.ANALYZE
Selecting the right people is one of the key aspects of the implementation process. It is
necessary to realise the importance of finding the correct team, which is going to be involved in
the Six Sigma initiative. In order to achieve their desired goals, the team may need to change the
way in which they have been working and thinking. A Six Sigma project requires a leader who
work-guide the team from the old way of working to the Six Sigma way. Everyone involved in
Six Sigma is a leader (Champions, MBBs, Black Belts, and Green Belts). Managers should be
trained in using Six Sigma to help run the organisation. The function of the leadership team
11
depends on the size of the company and the project. In large companies, there is a leadership
team at the corporate level and other leadership teams for each part of the project unit and
function. Eckes[2003] indicates that a good team requires different people and abilities.
Workers who use the process and work closely with customers.
Intelligent and dynamic people should form the team but they must work well together to
achieve the goals of the organisation. In the process of selecting people for the team, it is
necessary to keep in mind some of the characteristics that the members of the team should
possess i.e. they should be:
A team generally comprises between 3 and 12 members but smaller teams are preferred because
generally they work faster and produce results more quickly. Larger ones tend to require
additional time and sometimes it is necessary to form sub teams in order for them to operate
more effectively. In large enterprises, it is more common to have a design team of 5 or fewer
people as a subset of a larger team. This structure helps the design team to establish faster
results. Having considered the size of the team, the next issue is its structure.
12
The leadership team is the responsible for identifying the improvement opportunities and
leading the project to achieve the teams goals. In the case of a manufacturing process, the
leadership is the plant manager and the members of his or her department. In the case of the
finance function, the leadership team would be the CFO. A champion, Master Black Belt, Black
Belts and Green Belts, should form the team.
Each team should have a champion who is the business and political leader of the Six Sigma
effort. He or she uses to be a member of the unit leadership team and is the maximum
responsible of the project. Some of the responsibilities which a champion should have, are:
The Master Black Belts are technical leaders of the team. They have strong leadership qualities
and help the organisation to integrate Six Sigma within its operations. The MBB helps the
Champion to select the projects and to review the progress, they also train Black Belts, and
sometimes Green Belts. The MBB should help the champions in the improvement process and
implement the changes around the organisation. The Project Champion has the responsibility to
see that any problems or barriers identified are resolved.
Black Belts are responsible for driving organisational vitality and process consolidation, for
rationalising project activities, and improving service levels through quality leadership,
mentorship, and application of the Six Sigma methodology. Black Belts will work closely with
13
champions and the leadership team to ensure devotion to the quality methodology and to the
tools. This includes, but is not limited to, quality reviews, validation of adherence to quality
methods, strategy, and objectives.
A Green Belt sometimes leads a project with the supervision of a Champion or MBB. Green
Belts work part-time, devoting typically 25 percent of their time to the project. Green Belt
projects are more focused than are Black Belt projects. Since Green Belts work on improvement
projects in addition to their existing job responsibilities. Some companies have an objective to
form all professionals will be at least Green Belts such as GE. Some Green Belts will become
Black Belts. The different departments of the organisation should:
Provide expertise.
The team needs to include people who are familiar with the process, so that they can contribute
to identifying the solution. They will be involved in the process implementation. The Black
Belt, Project Champion, and managers are required to prepare the future team members report.
The process of forming the team is:
Identify any barrier, which do not help to achieve the teams goals.
14
It is important to inform the team about their role in the project and the benefit that they provide
to it so that they can appreciate their importance within the organisation. The new Six Sigma
organisation needs to realise the importance of re-evaluating existing work programmes, and to
re-prioritise how they utilise their resources. Six Sigma is usually deployed using existing
resources.
During the implementation process the team needs to realise that they can find out some
problems because the own Six Sigma methodology has its benefits but also its limitations. Some
of these are described on the next section.
15
Six Sigma is based on tools and techniques that have to be followed to achieve
the goals. Each has its place and reason to be applied to obtain the final objective of
the business.
Six Sigma establishes an internal structure within the business, enterprise which
leads, deploys, and implements the approach. This usually takes the form of:
1. Master Black Belts (MBBs).
2. Black Belts (BBs).
3. Green Belts (GBs ).
Six Sigma is based on the importance of accurate data and decision making
rather than on assumptions.
Six Sigma utilises statistical concepts, tools, and techniques for reducing the
number of defects based on variability reduction.
Six Sigma like any other quality improvement initiative has its limitations.
Sometimes, there are many problems in applying the change and finding the data
to begin with can be difficult. This could take most of the project time.
One of the most important problems associated with implementing the Six Sigma
methodology is determining the right selection and prioritisation of projects. In some
enterprises, the process of prioritisation is completely subjective.
The statistical definition of Six Sigma is 3.4 defects per million of opportunities;
on the one hand, this can result in dissatisfied customers, but on the other it cannot
assume that all defects have the same level of severity. For instance, for an airline
company a tragic accident is more serious than a cold meal.
16
Six Sigma can easily digress into a bureaucratic exercise between Master, Black
and Green Belts.
The relationship between Cost of Poor Quality (COPQ) and Process Sigma
Quality Level requires more justification.
In this chapter has been explained the necessity of implementing Six Sigma, the role of the
different members of the team and some of the benefits and limitations that this methodology
imply. The principal objective of the next chapter is to approach, in more detail, to the tools and
methods used during the implementation process.
17
When an organisation begins to grow and develop in size and structure a lot of inadequate and
inefficient processes emerge. Most of the employees of a business realise that a change is needed
but are unable to carry out improvement. Projects teams, for implementing Six Sigma, need
extensive training. The best way to start a project is to define a dynamic document, which sets out
the responsibilities, goals, roles, and aims of the team.
Everything that is done in life follows a process with a series of steps. In some cases these steps are
easy to recognise and sometimes there is an output .When a process map is constructed it is
necessary to identify the inputs and the outputs at each step. Process maps commence at a start
point and finish at a stop point, and pass through different process steps with decision points and
inefficient delays as shown in Figure 4.
Rework
STEP
DEL
AY
DECI
SION
STEP
START
DECI
SION
STOP
MOVE
DEL
AY
STEP
MOVE
The high level of the process map is called SIPOC (Suppliers, Inputs, Process, Outputs, and
Customers) demonstrates areas of inefficiency and waste. The Six Sigma Methodology is
19
focused in understanding the relationships between dependent and independent variables. This
is fundamentally different to what has been done in the past which was to focus on the outcomes
or the Ys. Now the emphasis is placed on understanding the vital inputs or Xs. Controlling these
parameters will bring success.
Y (output) = function of (inputs X1, X2, X3 Xn).
Six Sigma is aimed at producing Ys within a specific range by reducing the variation in the key
influential Xs. When Y is centred about the targeted (desired) value and the variation in the Ys
is 33 percent of the spread of the upper and lower control limits, we have a Six Sigma process.
Six Sigma uses statistics. It is necessary to use statistical tools to characterise the influence of
the Xs on the Ys. It can be quite difficult when there are interactions occurring among the
multiple Xs that influence Y and when there is a system that cannot be measured.
A process map is fundamental to the Six Sigma method. It is the foundation of measure. In
manufacturing, a good process map can be facilitated only with a knowledgeable and
participative team of operators, maintainers, process engineers, supervisors and engineers.
Process mapping allows the identification of core and enabling processes, as well as added
value and enabling process steps because customer are only interested in processes which add
value to the product. The main core processes are those, which constitute essential operations,
add value for the customer, and map well onto customer perceptions of the delivery of the
product or service. The core processes are those for which the customer is prepared to pay. They
can be divided into two or three more processes. However, there are other processes, which
from the point of view of the customer, do not add value to the product, these are enabling
processes.
20
The team needs to identify who is the customer and what are the customers requirements. In order
to achieve a good project selection the team needs to understand the classification of the customer,
the necessity of customer satisfaction, and the critical to quality characteristics. The customer can
be classified as external and internal customer.
A external customer is any person who is not related with the organisation and tastes the product or
the service of the organisation is an external customer. The most direct customer is the one who
consumes the product or service. The team can identified three types of possible customers:
1. Customers are those who have been consuming the products or the services of an organisation
during a determined period.
2. Not customer is existing or past customer who has gone elsewhere and now is not customer of
the organisation. These customers are important for the organisation because they were customers
in the past and now they are not. They are dissatisfied customers and the company needs to analyse
which are the reason to be not customer.
3. Non-customers are potential customers. Each class is important to an organisation but the most
important are the last two, they have more to say about poor quality and added value.
Sometimes the output of one process is the input of another and so if is an output is a defect it could
cause more defects in the final product or service. The internal customer is of equal importance to
the running and smooth operation of a company .Sometimes they are the stakeholders, in which
case they have a terrific interest in the operation and success of the business. In the case of
shareholders, they are interested in the value and worth of the business and the return on
21
investment. Internal customers do not have too much influence over the control of a company but
they help to achieve its objectives.
Six Sigma quality is aimed tat satisfying the expectations of the customer and may include the
profitability and corporate safety for the shareholder. In terms of processes, Six Sigma is related to
defect reduction and this not only influences customer satisfaction but also improves the internal
processes. As far as employees are concerned, Six Sigma is aimed at increasing the value of the
employees because they become more engaged with the operation of a successful business.
22
a need, determining a specification for solution, searching for information, evaluating alternatives,
making decisions and finally resulting in outcomes. Some studies based on the customer are
extremely complex and it may take several months to understand.
There are two definitions of core processes. The customer definition is that it is all those which add
value, affect satisfaction and quality of the product. The business definition is all these which add
value to the customer. The only way to identify core and enabling processes is to ask and study the
costumer. The result of this is a quantifiable ranking of such processes against customer judgement.
Six Sigma initiative consists of quantifying customer satisfaction. In most cases the customer does
not realise that he/she is the most important part of a Six Sigma initiative and the customer
information is the basis of Six Sigma. The important customer information needs to be processed
and analysed.
Customer satisfaction means that all the needs and requirements of the customer are achieved. It is
necessary to realise what the characteristics of the products or service are. The team needs to
improve customer satisfaction and list these needs and requirements for each process. This step is
quite difficult because the own customers sometimes do not know what really expect of the product
or service. For example, safety and accuracy are always vital to good service and a failure in any of
these is a dramatic defect for the customer. For this reason it is necessary to determine the things
and the process which create unsatisfied customers.
23
CTQs (Critical to Quality) as the key measurable characteristics of a product or process whose
performance standards or specification limits must be met in order to satisfy the customer. They
align improvement or design efforts with customer requirements.
CTQs represent the product or service characteristics that are defined by the customer (internal or
external). They may include the upper and lower specification limits or any other factors related to
the product or service. A CTQ usually must be interpreted from a qualitative customer statement to
an actionable, quantitative business specification.
It is necessary to determine a list of CTQ characteristics for processes that are not only measurable
but also actionable. Customer needs must be converted into CTQs. The success of Six Sigma is
based on a good conversion between CTQs and requirements. Only when this is achieved, the
process of change is possible and the organisation is closer to its goals. Customer research is often
expensive and all too often ongoing research is required to find a solution to reduce costs as well as
to provide continuous information, which helps the business to act.
At this stage, it is necessary to determine what the customer requirements are for each individual
process and to set specific target limits. The team needs to analyse customer requirements and set
targets. In order to do this the team should put their shelves in the position of an external customer.
Some of the steps that need analysing to determine critical to quality characteristics are the voice of
the customer and customer requirements and needs analysis.
Customers often have difficulty in expressing what they want or expect from a service or product
and forget the main things. However, this raw information can be quite useful even though it may
not be complete. The methods for obtaining information about the voice of the customer include
focus groups, interviews, market research, and becoming a customer and observing customers in
24
action. The most important for achieving the company goals is the ability to listen to customers
requirements.
In this section, the key issues will be determined. In order to identify a key issue it is necessary to
list one or more requirements, which would resolve the issue of the customer satisfaction. In some
cases, requirements are difficult to be expressed and sometimes they can be analysed from different
points of view. A good requirement should be measurable, concise and complete, but not a solution.
When new processes or products are designed, it is necessary to work down from high level
concepts of general customer needs. One of the tools that can help to the team in this regard is
Mallows hierarchy of needs which is a list of customer requirements from the basic to the abstract
(see Figure 5)
1. Physiological Needs
Physiological needs are the very basic ones such as air, water, food, sleep, etc. When these are
not satisfied a person may feel sick, irritate, in pain, discomfort, etc. Once they are satisfied,
he/she may think about other things.
2. Safety Needs
These needs are mostly psychological in nature. Security is essential in any purchase.
3. Social Needs
Humans have a desire to belong to groups: clubs, work groups, religious groups, family, etc.
There is a need to be accepted by others.
25
4. Esteem Needs
There are two types of esteem needs. First is self-esteem, which results from competence or the
mastery of a task. Second, there is the attention and recognition that comes from others.
People, who have all of their lower needs satisfied, often drive very expensive cars because
doing so raises their level of esteem.
5. Self-Actuation
The need for self-actuation is the desire to become more and more what one is, to become
everything that one is capable of becoming. People who have everything can maximise their
potential.
SELF-ACTUATION
SELF-ESTEEM
SOCIAL
SAFETY
PHYSIOLOGICAL
Figure 5. Mallows Hierarchy.
Kano Analysis is another tool used to determine customer requirements and needs. It is a quality
measurement tool used to prioritise customer requirements based on customer satisfaction and to
determine which requirements are important. Not all identified requirements may be of equal
importance to all customers. Kano analysis can help organisations to determine which requirements
have the highest priority. This is useful because customer needs are not all of the same kind and
they do not all have the same importance. The results can be used to achieve customer satisfaction.
Note that the Kano model can be used to help identify customer segments, based on the relative
priority of each segment's requirements. Once segments have been defined, the Kano model, which
26
is shown in Figure 6, can be applied once again to each segment to further defined the segment's
priorities.
There are four types of customer needs:
1.
2.
3.
The must have things. Without this, the product is not sold.
4.
The dissatisfies, the things that cause customers not to like a product.
Satisfaction
One dimensional
Delighters
Performance
Must-haves
Lack of performance
Dissatisfaction
Figure 6. Kano diagram.
Another powerful tool used to design new products is Quality Function Deployment (QFD)
which is used to define customer needs or requirements and convert them into specific plans to
produce what the customer wishes. The voice of the customer, as has been stated, is a necessary
27
tool to reflect the requirements of the customer. The next step, once that the requirements and
needs has been collected, is to build a product planning matrix with all the information about the
customer requirements. This matrix is used to translate higher-level needs into lower level ones
and determines how the needs are transformed into a satisfying product. The basic QFD
methodology has four phases. During each phase, one or more matrices can be built to realise
critical product and process planning and design information.
Once the majority characteristics of the project have been defined, the team can begin to
establish what they are going to measure and to establish the procedure to achieve a good
measure of the process.
28
29
There are more potential useful forms of analysis with continuous data than with
discrete.
Six Sigma measurements are about understanding the relationship between the X factors and the
impact that they have on customer satisfaction and profitability, Y. There are two types of
measures, output and input. Figure 7 shows a simplified process.
(X)
(X)
(Y)
PROCESS
INPUT
OUTPUT
Figure 7. Simplified process
The goal of the team is to find the Xs that are the indicators of the critical output Y. For this
purpose, it is necessary to use a SIPOC diagram and sub process map. The basic steps for
implementing any measurement system in a project are select what to measure, develop
operational definitions, identify data sources, prepare a collection and sampling plan,
implement, and refine the measurement.
measure of the product. The team needs to evaluate what is more valuable for the customer and
to determine the possible opportunity area. The team has to be trained to select the right
measurements and needs to bear in mind, when selecting the items to be measured, which are
the available data, the complexity, the cost and the lead-time required of getting the data.
31
Sometimes it is a good idea to give data collectors the opportunity to provide inputs on
the process.
Another option is to use existing data; this has some advantages such as achieving data, learning
about the input, process, or output and being quicker and easier than gathering new data. There
are however cautions to remember:
-
The team needs to know how the data were initially collected; the procedure may not
be consistent with the operational definition.
The data should be representative of the process and the measurement system.
32
Decide on the accuracy of the measuring system and indicate it on the form.
Another tool which can help the team is the utilisation of check sheets which are employed
for distinguishing between fact and opinion and also for gathering data about how often a
problem is occurring and the type of problem occurring. Some of the most common check
sheets are:
1.Defect or cause Checksheet.
2.Data sheet. Captures readings, measures, or counts quantities.
33
3.Concentration Diagram Checksheet. The item is observed and the data collectors
mark where the defects are produced. The data collection sheet is based on the
physical representation of a product, workplace, or form.
4.Traveller Checksheet. It is a check sheet that travels with the product through the
process. It is one good way of analysing data at each step of the process.
Once the collection forms have been developed, it is necessary to establish stratification plans.
The purpose of stratification technique is to obtain a baseline measure of performance against
customer requirements. Sometimes the team wants to know more about the data and
stratification helps to clarify what is really happening. The team can identify the strata by
making a list of factors, which is related with the problem. Some of these include:
1. Identifying output measure.
2. Introducing the questions that the team has about the output measure.
3. Identifying characteristics which define different subgroups.
4. Creating specific measures for each subgroup.
5. Discussing if the specific measures help to analyse the output Y. If not it will need
to apply the measurement system to other items.
After stratifications plans haves been designed, the team can begin the sampling procedure.
Sampling means using some of the data in a group to represent the whole process. Statistics is
based on sampling and this discipline has the ability to draw conclusions about all items based
on a part of them. Getting a valid sample is in some cases hard work and presents a significant
challenge. Different factors affect the number of samples and size of them that the company
needs to collect. For this reason it is necessary to establish if the items change with time or not
34
(process or population), the data type (Continuous or discrete), the finality of this data and the
confidence that the company put in the results.
The most difficult aspect of sampling is to ensure that bias is kept occurs to a minimum. Bias
occurs when the sample does not represent the whole. There are others concepts that it is
important to define to understand the importance of collecting data:
- Confidence level: this concept is related to how certain that the company wants to be.
- Precision: The quality of being exact.
The team, after developing the stratification plans and selecting the sampling population, needs
to achieve accurate data, which will be analysed in the next stage.
35
Some disadvantages or problems make defect measures less attractive. One is that they only
look to see if one process is good or bad and in the case of continuous data the measure can hide
information. Some key concepts of defect measurement that need to be signed are:
Unit. An item or the final product or service that has been delivered to the customer.
Defective. An entire unit that fails to meet acceptance criteria, regardless of the number of
defects within the unit. A unit may be defective because of one or more defects.
Defects per unit. It is the number of defects over the number of units sampled.
Defect opportunity. One product can have multiple customer requirements and for this
reason there can be several opportunities for defects. Sometimes it can be difficult to
determine the number of opportunities that one product can have.
The defects per million of opportunities indicates that how many defects would arise if there
were a million of opportunities.
All the defect measures that have been described are based on results or measures at the end of a
process. Sometimes when the purpose is to determine what the customer requirements are or to
evaluate the efficiency of the process, these measure are not useful and give no real indication
of how the whole process is working. Internal yield measures are based on data collected from
the process operation. The team needs to measure internal defects to know how well the process
is working otherwise a lot of information is hidden.
Once the team has a clear definition and a guideline of what to measure and the procedure to do
this, it is necessary to analyse the collected data, which it is described in the next section.
37
38
The analysis phase can be represented as a cycle as shown in Figure 8. The cycle diagram
begins at point 1, if the objective is to identify defects, or at point 2, if a hypothesis is to be
validated. If the hypothesis that the team has chosen is not correct it is necessary to go back to
the beginning of the process and redefine the problem.
1 Analyse
Data/proces
s
Refine or
Reject
Hypothesis
Confirm and
select vital
causes
2 Develop
causal
hypothesis
Analyse
Data/proce
ss
Figure 8. Analysis phase diagram
There are two key sources to determine the true cause of the problem.
1. Data Analysis. Use data from the measurement phase to create thinking about the cause
of the problem.
2. Process Analysis. Investigate how the work is being done to identify problems during
operation.
The tools that are going to be described in this Chapter can help the team to identify the root
causes of problems and find new solutions. They can be divided into two groups, those for
identifying and those for verifying potential causes.
39
Data displays
These tools are focus on creating ideas about potentials causes. There are many which the team
could use in their investigative efforts. Some of these are:
Time series plot. It is a tool that shows if a process is changing over time and depicts the
data in the order in which they occur. It is easy to construct manually and requires less data
than other tools.
Pareto chart. A Pareto charts is a type of bar diagram in which the x axis represent the
categories of the problems, defects or error and the vertical axis y represents a count or
percent of defects. To create a Pareto chart it is necessary to follow the next steps:
40
1. Determine the total number of problems and the counts or impact in each category if there
are categories with very small frequency they can be combined.
2. Group the problems by frequency or by level of impact.
3. Draw a vertical axis and divide into increments.
4. Draw the bars of each category.
5. Represent a curve showing the cumulative percentage.
Pareto analysis gives the team a visual interpretation of the data. If there are one or more
categories with a high percentage, this means that these categories of the problem account for
the most occurrences or impact. A possible example of Pareto chart representation is shown in
Figure 9.
Pr
ob
le
m
1
Pr
ob
le
m
2
Pr
ob
le
m
3
Pr
ob
le
m
4
Pr
ob
le
m
5
O
th
er
s
120%
100%
80%
60%
40%
20%
0%
5 whys. This method is used to help to the team to think about root causes and not to be
satisfied with supercilious solutions which do not resolve the problem in the long run.
41
The fishbone diagram is used to identify causes of problems or prevent future problems.
The procedure to create and use a cause and effect diagram is as follows:
1. Name/define the problem or the effect, which is going to be analysed and write it in
a rectangle placed on the right hand side of a piece of paper (at the head of the
fishbone).
2. Decide the major categories for causes and create the basic diagram (Figure 10).
3. Brainstorm for more detailed causes and create the diagram( for this step the team
can use the 5 whys method).
4. The team need to review the diagram and eliminate causes that do not apply.
5. Based on the final diagram the team need to identify the causes that are most critical.
42
cause
cause
cause
Effect
cause
cause
cause
Failure Mode and Effect Analysis (FMEA). This technique is used to identify the ways in
which a product or a process can fail and to prioritise the actions to reduce risk.
The team can perform an FMEA in the following way:
1. The first step is to review the process or product.
2. Brainstorm potential failure modes.
3. The team needs to list the possible consequences if a failure occurs.
4. Ranking the severity of failure: 1-10, 10 means most severe impact on customer and
assign rating to likehood that a failure will occur: 1-10, 10 represented most likely
occur.
5. Assign a detection rating to each failure: 1-10, with 10 representing least likely to be
detected using current control.
6. Calculate
risk
priority
number
(RPN)
for
each
effect
multiplying
(severity*occurrence*detection).
7. Use the RPNs to prioritise actions with highs RPNs or those that have a very high
severity rating.
43
8. The team need to establish a plan to eliminate or reduce the risk associated with
high-priority failure modes. This plan begins identifying the potential causes of the
failures modes and then develop recommended actions (with timeplan, etc.).
9. Develop the plan and recalculate RPN.
t-test. The t-test is statistical tests that permit to the team if the null hypothesis is likely to be
true. It is necessary to calculate a t-statistic to determine probabilities comparing the
statistics with a t-distribution. This test is used in two situations:
Regression Analysis. It is a tool that uses data of important variables to develop an equation
Y=f(x). The team needs to evaluate which input variable have the biggest effect in the
response variable and delete unimportant factors from the model.
44
ANOVA (Analysis Of Variance). This tool is used to analyse the relationship between
categorical inputs and one continuous output.
Chi-square test. The Chi-square test is a hypothesis testing method used when the
contributing factor and the result are attribute data. The steps that the team need to follow to
calculate the chi-square:
1. Collect the data from the measure phase and summarise the results in an
observations table.
2. Based on the observation table develop an expected frequency table.
3. Compute the relative squared differences.
4. Add together all the relative squared differences to get the chi-square.
5. Determine and interpret the p-value.
In this section have been described the tools and methods used in order to analyse and
investigate the potential cause of the problems. Now, the team, based on the analysis of the data,
has to begin to construct and develop solution for the problems. This is developed in the next
section improvement phase.
45
46
Three tools, which would help the team in doing this, are brainstorming, affinity diagrams, and
benchmarking.
3.4.1 Brainstorming
Some of the keys to successful Brainstorming are:
1. Clarify the objective of the brainstorming. Everyone should agree on the purpose and
work in the same direction to get good results. The brainstorming is necessary once the
goal has been clarified. It can provide important information for eradicating defects. In
order to generate good ideas, the team needs a period of silence to think about how the
process can be improved and to write down some ideas.
2. Listen to and build on the ideas of others. It is necessary to pay attention to other
peoples ideas and develop them.
3. Abandon assumptions. Past assumptions sometimes prevent the team from developing
good ideas.
4. Consolidate similar ideas. Discuss certain ideas with a view to consolidating them.
5. Understanding of the idea deployment process.
47
3.4.3 Benchmarking
Benchmarks are measures which a company compares its performance against other companies
practice. This tool determines how to use this information to improve companys performance
and can help the team to focus their objectives to get good solutions to problems.
The potential solutions that have been developed in the last step, need to be evaluated, selected,
and optimised. Two tools use for this purpose are the Solution Selection Matrix and the Pugh
Matrix.
The purpose of the solution selection matrix is to make sure that the solution selected for
implementation is the best to achieve the objectives of the team and the organisation. It is used
when it is necessary to choose one out of two or more solutions.
The Pugh matrix is a tool that uses decision-making that compares concepts, based on the
requirements of the customer and functional criteria. The objective is to work with potential
solutions to converge to an optimal solution.
The team has selected the solution and now needs to implement it. One of the tools that the team
can use is Pilot testing which evaluates the solution and its implementation. Pilot test determine
if the process is agree with the customer requirements and the design specifications. To pilot a
solution it is necessary to follow the next phases:
Phase 1: Plan. The team need to determine what needs to be piloted, what will be
involved in the process, how long will the pilots run and how will the pilots be
conducted.
48
Phase 2: Review design. It is necessary to review the plans to ensure that all the
elements of the design are completed. The team needs to find possible defects that can
be tested at the pilot phase and review both the pilot and the implementation plans. At
the end of this phase the team is able to list the key issues raised, the possible actions to
apply in the pilot and changes in documentation.
Phase 3: Finalise design and implement. In this phase, the team will implement design
changes that have been found in phase 2 and will start the implementation.
Phase 4: Evaluate the test and verify results. This is the last phase of pilot testing. The
team should use statistical tools to evaluate design predictions and improve on the pilot
if there are any weaknesses.
During the improvement phase, the team has selected and finally implemented potential
solutions. Once these solutions have become effective, the team needs to begin the next step,
which it is the control phase.
49
50
The team needs to make a good documentation and for this reason they can complete their
written documentation with pictures and flowcharts for getting a clearer and more accessible
documentation. One way to show the efficiency of the new process is to document the possible
behaviour and some procedures to be applied under various conditions. Finally the
documentation must be handy and update. The team needs to review and follow the procedures
that there are in the documentation in order to achieve good results.
51
mind to measure the next areas: innovation, process customer and financial. Therefore, the team
is able to establish common performance measures.
Response plan
It is necessary for the team to have guidelines, which indicate what to do and when do it in some
circumstances. The team needs to know what actions have to be taken in each situation and
when those actions are necessary. It is necessary to establish clear standards in input, process
and output of the process because if the key points are standardised and fully identified con help
to the team to discover which are the actions to apply in each case. A good response plan must
to be a continuous improvement plan. It should identify and prioritise ongoing or serious
problems, which can be improved. The response plan must anticipate possible problems. Two of
the tools, which can help the team to anticipate the problems, are FMEA and potential problem
analysis.
52
3. The process owner should be the first person who discovers the problems before they arise
and defines guidelines for long and short term solutions.
4.
When projects to improve are identified, the process owner needs to select a team, which
carries on the project, and they have the responsibility to maintain the gains.
5. Process owner has to work with customer and supplier to reach the goal of the company at
the highest level of performance. Thus, the company could achieve high Six Sigma levels of
quality, efficiency, and flexibility.
The process owner has to focus on gains and the customer and should be present where the core
process are running for trying to identify weakness zones. However, the principal purpose of the
process owner is to lead the team in order to achieve a high sigma level.
53
All the efforts made for the team in the process improvement become the key elements that can
help to the team to reach a high Six Sigma level and respond to the customer requirements and
demands for new products, services, or capabilities.
Once the team has a clear definition of the tools and the procedure to achieve the required sigma
level, it is necessary to find out how this methodology is applied in some real case studies and to
know the difficulties, which can reach in the implementation process.
54
BACKGROUND
This published case study of Gack [2002] centers on the use of the Six Sigma DMAIC process
in an organisation which develops software packages. It achieved substantial reductions in costs
within 6 to 12 months of the beginning of the implementation of the project and reached a 15 to
25% reduction in software development cost in two years.
1. DEFINE PHASE
The main purpose of the define phase was to determine the critical-to-quality characteristics,
create the project charter and develop a high level process map. The principal goal, Y(s), of the
organisation was to improve the capability of its software which meant ensuring that the project
was delivered on time, with predictable effort, and with an acceptable number of released
defects. In order to achieve the teams goal, they identified the following characteristics:
Time to market.
Then, the team would focus on the percentage of project commitments that were delivered on
time. It was assumed that the organisation was decentralised with various divisions in different
parts of the country. The team considered data from previous projects separately and so selected
55
the project that had more precedents, in the hope that the process could be replicated in other
divisions. The fruits and nuts division was selected by the first team.
Once the specifications had been identified, the team began to build the project charter which
should include the expected pay off if the team was able to improve the critical to quality
characteristics. It was also necessary to provide a good definition of the problem:
"Conversions that miss planned dates are causing unexpected budget increases that impact our
profitability."
The next step, the development of the project charter, was to define the goal statement. One of
the principal was to improve the on-time project percentage from the baseline value of 62% to
90% within the next year. Black Belt was selected as project leader. The champion was the Vice
President of Marketing, with the Director of the Project Office, three software Project
Managers, and the Director of Quality Assurance as team members.
The financial opportunity was based on 10% of the potential service contract revenue which was
lost due to late software deliveries which amounted to approximately $800,000 per year.
2. MEASUREMENT PHASE
The first thing to be done was to determine the current state of the process and to collect reliable
data. It was decided to investigate four controllable factors:
"Levelled" resources (ensuring an 80-hour week has been scheduled for the team)
56
It was also necessary to evaluate the measurement system to be used. The team needed to be
able to understand the quality of the data before making any conclusions and to convert attribute
data into quantitative information. The four controllable factors selected were called potential
Xs, which influence the behaviour of the outcome.
Scoring system, which had been used to rate the X(s) in previous projects was presented. This
would show if the high score of an X(s) high score was related to the high performance of Y(s).
Table 1 shows the score system used in some historical projects.
Using the score system the team collected the Y(s), schedule performance defined as percentage
of plan, and X(s) for each project. The results are listed in Table 2.
57
In Figure 1, the straight line indicates that as the score increases, Y is improved ( percentage of
plan) suggesting
that the higher the attributes of a professional plan the higher is the
improvement in Y. However, there are some points which are marked in Figure 1 that do not
correspond to the general pattern and so perhaps it is necessary to consider more factors which
influence Y.
58
3. ANALYSIS PHASE
From the measurement phase it was observed that only 20% or 4 projects were achieved on
time, not the 62% as required by the software team. If the companys goal was to achieve 90% a
450% improvement would be needed. The Six Sigma team constructed an estimation model and
a new column "Plan %" which is defined as: (actual planned duration) / (duration indicated by
the estimating model), could be generated and added to Table 2. If the Plan percentage was
less than 100 %, the planned schedule would be too optimistic. Table 3 shows the
implementation of this in the case study. As can be observed the planned schedule was too
optimistic because there were 15 projects which heir Plan percentage less than 100%.
During the measurement phase, it is also necessary to identify which of the critical factors Xs
have more influence in the determination of the schedule performance of the Y(s). In order to
achieve this, the team applied the multiple regression analysis and concluded that 78% in the
variability of this sample was due to task duration, predecessors and plan percentage.
59
Once the critical X(s) were determined, it was necessary to redefine the financial benefit
forecast. The average expected profit for the 20 projects was $850,000 and the projects were
scheduled to take 15 months. The projects required 134% of the planned duration and they
needed 5 months more than expected. The opportunity cost was 5/12 of $850,000 with an
expected time of money of 15%. The team estimated that the opportunity cost associated with
the 20 projects obtaining a profit of $1,065,500 but the company had a target of 90% and a final
profit of $956,250.
4. IMPROVEMENT PHASE
The principal purpose of the improvement phase is to generate, select and finally implement
solutions. The following possible solutions were identified:
1. Train personnel responsible for project planning on best practices.
2. Assign mentors or coaches from the Project Office to review the draft plans and help project
managers bring them up to best practice standard.
3. Use a combination of these options.
In order to determine the best option in terms of cost and effectiveness, it was necessary to pilot
two or more options and try to discover the best solution. The team decided to implement the
alternative of trying each one with different teams and comparing them after two or three
months.
5. CONTROL PHASE
It was necessary to develop a control plan and define how the X(s) and Y(s) will be monitored.
There was established that if any score was below 5 for any of the X(s), the team would try to
60
reach that level and if the target for the Y(s) was not met, they would review the project.
Finally, it was necessary to define responsibilities for the Six Sigma team project and to fully
document of all the process records and data.
61
BACKGROUND
This published case study of Goyal [2004] is based on a Medical Transcription company
established in India with customers in the U.S.A. Medical Transcription involves writing the
notes from doctors, transcribing them according to appropriate rules, protecting them from
errors, and recording the final results. This is an important process because it avoids
mistreatment of patients. The principal purpose of the business is to increase of their revenues,
control the absenteeism and patients load.
Internal problems that were causes of customer problems, rather than basic problems.
The first category focused on customer satisfaction and the team, using a prioritisation system,
put the fundamental CTQs as:
1. Consistency of quality
2. Timeless, which was allocated the higher priority
62
In order to achieve a precise definition of the problem, the team interviewed different customers
and the following conclusions were made:
The customer wanted to be dispatched before 5.30 PM every day. This means that a customer
who was received between 7.30 AM on day X and 7.30AM on day X+1 should be dispatched
by 5.30 PM on day X+1.
The turnround of a file was defined as the time taken from receipt to dispatch it. This meant that
a file received at 7.30 AM on Monday should have a turnaround of 24 hours and a file received
at 7.30 AM on Tuesday. Following on analysis of customer requirements, the team realised that
the problem was "meet the dispatch deadline of 24 hours" rather than "reduce turnaround of
data". The team suggested measuring the discrepancy between the time of dispatch and the
customer's target. The team designed a check sheet in order to collect the actual time of dispatch
during two weeks including:
Variation: 82 minutes
Thus:
Average + 3= 335
2. ANALYSIS PHASE
In order to ascertain the vital causes of the problems, the team developed a check sheet to
collect the following data: starting time, loads, past and actual transcriptions, capacity per
63
transcription and finishing time. These data were collected over several days; however, the
results from the data collected between 7.30 AM on day X and 7.30 AM on day X+1 began to
be processed at 8.00 AM on day X+1. The principal reason for this was that the data came from
doctors of U.S.A. who normally dedicate the end of the day to the dictation of the data and this
time corresponds to the morning in India.
A group was selected to collect the data from the United States every hour for a week. Based on
this data, the team started the improvement phase in which ideas were generated, selected and
finally implemented.
3. IMPROVEMENT PHASE
There were two fundamental ideas:
Use two teams during the night shift, one team during the day shift.
Use one team at night, one team for an early morning shift, and one team for the day shift.
Eventually the first idea was selected and put to the test. The initial expectation was that lower
efficiencies would provide required results, which was supported by the outcome of most of the
cases. In the first two days of testing, the dispatches were achieved two hours earlier than
expected.
Table 1 compares the results from the third week after the changes and the week before the
project began.
64
Load
Delay time
Average
plus 3
Week
Average Maximum Average
Sigma
sigma
-1
285
397
89
82
335
3
348
531
-134
43
-4
Table 1. Project results.
The results showed 22% higher load with an average dispatch time of 134 minutes
early compared to 89 minutes late and compared to 335 minutes late at the same reliance level.
65
1. INTRODUCTION
This published case study of Hallowell [2004] is based on an information technology (IT). IT is
a service used by companies as a means of providing call centre support to a variety of
customers. IT services can be really important in a business because when prices are similar, the
customer chooses the company which provides the best service. In this published case study,
benchmarking was used to show that the company was not competitive enough and needed a
change. Six Sigma was the initiative adopted by company to improve their service.
A Champion was chosen for the new project, and then he/she chose the leader of the project.
The objective of the team was to reduce support costs while improving new account growth.
The champion, with the help of a Black Belt and after verifying the benchmarks and other data,
was able to select the DMAIC project.
The team compared customer satisfaction against average and best class companies and found
that their rate was a little below average (73 percent for the studied company against an average
of 76 percent).
The company was aware that customer satisfaction could be the driver of new account growth.
The team studied this and realised that the range of customer satisfaction ratings for the best
competitors produced about 76 percent of the changes in new account growth. The team also
noticed that customer satisfaction and business growth do not have a direct relationship to
support costs per call because those companies with the best customer satisfaction and business
growth spend less money on support costs per call.
66
2. DEFINE PHASE
The team needed to evaluate the elements of the project charter. The following situation was
described:
The competitors of the study call centre are increasing their customer satisfaction levels by
reducing support costs per call. The team needs to increase the companys growth from 1
percent to 4 percent, without increasing the support costs per call, which would mean an
augmentation of the companys growth revenue of $3 million.
The teams goal was to increase their customer satisfaction level from 75% (90th percentile) to
a target level of 85 percent (90th percentile). The team discussed with the team leader the
objectives of the project charter and then collected the customer satisfaction data.
The case study included some individual customer responses, which were representative of the
data gathered and are summarised in Table1.
Requirements
Connect with the right person.
Get the required information.
Personal help.
How to prevent problems.
67
3. MEASUREMENT PHASE
Once the customer requirements had been identified, the team could focus on the measurement
phase. The team prepared the data collection plan and identified all the parts of the project and
the possible measurement areas.
Y(s)
Measurement
Customer
Satisfaction
Support Cost
Case research
Call back time.
Days to close
Wait time
Calls in queue.
Transfer
Service Time
The team then needs documented the current baseline and the best estimate target as is shown in
Table 2.
68
MEASURE
CURRENT BASELINE
TARGET
Customer Satisfaction
90th percentile/70-
85%satisfied
Support cost
90th percentile/$40
Days to close
data
less
Wait time
data
Transfers
90th percentile/2
data
Wait time
Mean< 8 Minutes
data
It was also necessary to identify the factors which drive the Y(s) and their behaviour. When the
principal objective is to discover the Y behaviour, the team needs to focus on the Xs behaviour;
this offers the possibility of addressing the root cause.
At the end of the measurement phase the team had began to forecast what could be revealed in
the analysis phase using the data collected to produce an Xbar-R control chart, which represents
the variation in customer wait times.
4. ANALYSIS PHASE
The team graphed and studied each Y data set. Table 3 summarises and compares the capability
checks to establish targets.
69
Measure
Capability
Customer satisfaction
Target
90th Percentile = 85%
satisfaction
Days to Close
Wait Time
Transfers
90th Percentile = 2
Service Time
Minutes
0.5Min
During the measurement phase the team tried to identify some important Xs factors. At this
point, the team had to verify the previous Xs factors and discover others in order to make
comparisons and an analysis based on the Xs identified in the last study. The team observed that
on Mondays and Fridays, the company was often understaffed whilst on Sundays, it was
overstaffed. They developed an analysis of variance for wait time. Based on the results of the
ANOVA, the team also observde the influence of callbacks on the call wait time. (P<0.05)
5. IMPROVEMENT PHASE
At this stage, the team needed to identify and select potential solutions in order to reduce the
support cost (controlling staffing ratios, transfer and callbacks). During the analysis phase, the
team identified some improvement areas, but now alternative solutions are sought for some of
the Xs identified in the last phase:
70
Solution alternatives
* Additional staff on Monday and Friday and
reduced on Sunday
*Staffing model
The team had to consider how these solutions could improve the process compared with the
current and the possible benefits. This was the objective of the team in the following steps of the
improvement phase.
In this step, it was necessary to evaluate and quantify the key relationship between the Xs and
the Ys and the possible benefit. If the benefit appeared to be positive, the team had to determine
how to implement that improvement. To evaluate the staffing option, the team had to bear the
following points in mind:
The variables that affect the process wait time, support cost and volume/staff (v/s) ratio.
Measure the benefit value of account growth minus the cost of additional staff.
Based on the collected data, a regression analysis was performed which, showed the wait time
was related to the v/s ratio and so the same technique was used to determine if the wait time
71
affected to new accounts. In this analysis 61 percent of the variation in account growth could be
attributed to wait time. Before the team modified or improved any process, it needed to forecast
what would happen theoretically.
With reference to increasing the staff numbers on Mondays and Fridays, the addition of 14
people on Mondays would result in the following:
After performing the same analysis for web service, transfers and callbacks, the team
recommended that:
Changing staff levels should be the first step but in the future discover another solution
Reducing the number of transfers and callbacks in order to increase customers satisfaction.
Y(s) to measure:
- Wait time, v/s ratio, customer satisfaction, transfers, callbacks and service time.
- Carry out a hypothesis test to compare new staff against old staff.
72
In order to perform the pilot test, the team needed to increase the current level of customer
satisfaction sampling, which only gave 1 set of data per month. The team arranged to have 5
sets of data per day over 90 days using 80 percent of the current staff and 20 percent new staff.
During the first month, the new staff took 6 minutes more than the old ones; this difference was
reduced during the second month to 2.5 and then to 1 minute in the third month.
To determine if the wait time had decreased with the staff increment, a two-sample t-test was
performed; this resulted in a pvalue less than 0.05, which means there was a difference. At the
end of the improvement phase, the team evaluated the goals that had been achieved. They were:
Initially new staff needed more time than the old but after three months both groups required
The overall conclusion was that the Monday staffing pilot programme has been a success and
the team recommended full implementation.
73
modify old ones for determining the right solution to implement and try to eliminate defects in
order to know if the variable was out of limits or not.
74
75
fact and data, and not merely relying on peoples opinions and experience which can be fallible.
Furthermore, Six Sigma is focused on issues and problems important to both the business and
the customer, and is delivered through manageable sized projects prioritised in line with the
businesss strategy.
Six Sigma is far more powerful in service operations than in manufacturing. Not only is the
average project value higher, savings or earnings are often replicable across a large volume of
users (millions of banking customers, insurance subscribers, etc).
Approximately 30-50% of the cost in a service organisation is caused by costs related to slow
speed or performing rework to satisfy customer needs. Six Sigma in the service industry attacks
these costs by eliminating waste, speeding up processes, and delivering customer wants and
needs. Some of the results, which can be achieved:
Advertising
Banking
Insurance
Document management
Professional services
76
Traditionally Six Sigma has been used in the manufacturing industry. Now, in the new
competitive world, where the financial industries need a distinction factor to increase their
revenues, Six Sigma is starting to play an important role as distinctive quality methodology. Six
Sigma is not only merely a quality initiative but it is also a business initiative. The use of the Six
Sigma methodology is more valuable in financial institutions now than it has ever been and
some companies are achieving true savings and revenue growth. Six Sigma is being introduced
77
through the financial services, where performance management is becoming critical for firms
throughout the sector. From investment banking to credit cards, the need to respond to customer
demands is increasing daily. Customers expect faster and easier service and if they do not get it,
they will go elsewhere.
Six Sigma is a methodology that helps to satisfy customers demands, find out how to improve
the cycle times of all their processes, eliminate bottlenecks, minimise errors, cut cost-sand and
increase capacity. It is necessary to provide more value per customer transaction, rather than
trying to acquire more transactions and keeping existing customers can be more important than
finding new ones. By using Six Sigmas define-measure-analyse-improve-control process,
leading financial services organisations have worked to reach the methodologys aim of near
error-free performance. Moreover, this goal is relevant to all processes, from handling
customers money, to processing payments, to sending out bills, to closing a loan, or answering
the phone.
The principal goal of the Six Sigmas team should be improving service delivery and customer
loyalty. This can be achieved reengineering processes for speed and flexibility, retaining good
employees and boost shareholder value. Six Sigma provides the connection that allows the
company to improve and sustain performance. Whether the company goal is to improve
customer experience, comply with the latest regulatory requirement, or simply improve the way
that the service is delivered, the organisation can rely on Six Sigma as the quality differentiation
factor. The Six Sigmas team needs to understand and to be centred on the next points:
Knowledge and experience in financial services industry
Change management expertise
78
79
Six Sigma methods sometimes cannot work really well. Nevertheless, what Six Sigma does give
the company is a very neat and coherent Management Pack for conducting the business
activities in an improved and more sophisticated manner. After all, business is actually very
simple, and if companies do the basics well and maintain a strong focus on their customers and
core services, they will always be successful. Six Sigma provides the organisation with a means
of applying great management practice in a structured approach, but the company should never
overlook the most powerful tools, good leadership, good management and of course, common
sense.
Since its emergence in the mid 80s, Six Sigma has been used primarily in manufacturing, but
actually, it is being used in all sectors for different reasons. Six Sigma was developed to provide
a means of reducing common cause variation in business processes. There is hardly a Financial
Services business out there in the market today that is not suffering the effects of uncontrolled
variation in one or more of their client facing processes. In many cases, companies need a new
approach to service and operational improvement that will enable them better to serve their
customers.
A core theme of the Six Sigma concept is centred on the idea that all activities conducted in
business are processes, and every process has inherent variation. Whilst it is easy to view a
production line in a factory as a process, it is harder to identify discrete processes within a
financial services business. Yet the Financial Services industry is built on them and beneath all
behaviour of the front offices in these organisations lays a maze of (often-manual) operational
processes. The insurance sector is a good example. For years, its members have led everyone to
believe that the industry is highly complex and shrouded in mystery, but when one breaks it
down the underlying business is built on three core processes risk placement, claims and
80
Front office
Bussiness
Risk
Customer
acquisition
Identification
&Verification
Credit
Operations
Operations
Customer
Service
Know your
customer
&Money req.
Service
Fulfilment
Account
Management
This is an area of business thinking where the Six Sigma approach is offering Financial Services
companies an opportunity to rethink their operating model to improve their capability in
delivering end-to-end services to their customers. Some companies have already moved to a
model broken down into core processes, aligned to the customer experience, such as Enquiry to
Order, Order to Remittance and Billing. If this thinking were applied to the business
81
described above, it would look considerably different from the status quo flutter. The key point
is that the business would be aligned to the decisive moments, those interactions with the
customer when they experience the service and make a judgement upon it.
Most importantly, the management structure of the business should be realigned also. Figure 2
represents the new possible alignment of the process, which is centred on the customer
experience.
Front office
Customer
acquisition
New
Customer
Activation
Service
Delivery
Account
Management
Customer
Departures
One of the most important and high profile management positions in any services company
should be the end-to-end owner of the new customer activation process. In order to understand
better this point the following should be consider. Whether a bank, building society, mortgage
lender, insurer or investment manager, each new customers first experience with a financial
services provider is the account opening process. If the process goes well and the human
interaction with the provider is impressive, the customer will come on board with greater
reliability and commitment. They will accept more service failures later on before they depart to
another supplier. They will tell their friends and colleagues about the experience. On this firm
foundation, the provider is more likely to achieve cross selling of additional financial products,
and it goes without saying that if the experience is poor, the impact is far greater, but in reverse.
This position would have a significant impact on a companys growth and profitability.
82
Adopting Six Sigma methods within a Financial Services business, it is never going to be
simple. Financial Services companies may be built around processes, but in the majority of
cases these processes rely on people. People provide the inputs, people conduct the processing
and people produce the outputs. This has a number of implications. Firstly, the inputs to the
process are often incomplete and inaccurate. Secondly, there is a high likelihood of human error
and customs within the transaction processing. Lastly, and most importantly, if the organisation
wants to change the process it is necessary to persuade the people.
It is rarely to hear of manufacturing companies who have failed spectacularly in their Six
Sigma programmes. Yet it is common to hear of service businesses where the initiative has
failed to get off the ground and achieve the commitment it requires to succeed. This has caused
a fair level of scepticism with senior executives in the Financial Services market and has
deterred many companies from adopting the method. The principal Commandments for Six
Sigma in Financial Services are:
83
appropriate means, then use it as a great approach. However, a clear and meaningful operational
strategy is the starting point and that drives everything that follows.
2. All projects must satisfy the strategy and priorities of the business
In defining the operations strategy, the business should conduct a rigorous operations review,
identifying risks, problems, opportunities and recommendations, and assigning real cash values
to them. This drives priority in the selection of projects, and in the selection of appropriate
methodologies and practices for each project. At this stage, the appropriate tools and training
can then be properly deployed. Experience has shown that few companies have been successful
when they simply launch into DMAIC projects as soon as a corporate decision has been made to
implement Six Sigma. This is because the head first approach into DMAIC does not allow for
appropriate and deliberate project selection. The net result is that companies invest fortunes in
training for, and working on, the wrong projects.
84
company. A greater chance of success and cultural acceptance is achieved when launching by
stealth. Companies should select a few high profile and high impact projects, train only the
required number of resources and focus on achieving tangible results and stakeholder support.
Then it is time to consider a wider role for Six Sigma across the business.
4. Training Programme
If Service companies do choose to implement Six Sigma, they should be very selective about
any element included in their programmes and should not accept the conventional view that
every aspect of Six Sigma is applicable or necessary in every circumstance. Training staff in
only those specific elements of Six Sigma, which are relevant to their industry, service
businesses, can avoid the lengthy training programmes on the market, which are sometimes
more academic than practical, and which have been derived predominantly from non-service
environments.
In Financial Services, initial Six Sigma training to get a Black Belt up and running can be
successfully and effectively achieved in days, rather than weeks. This can be further supported
as necessary with specific coaching during actual projects, so that Black Belts learn from real
experience. A primary reason for this is that not every Black Belt needs to be an expert in
Statistical programmes. Whilst this statistical software plays an important role in data analysis
and the production of high quality dashboards, it can be a supporting function provided by a
team of business analysts, rather than expecting every project leader to be an advanced
statistician. In this industry, the soft skills make the difference.
Finally, experienced and highly qualified service industry practitioners, whose business, should
deliver training and coaching and industry sector expertise is both personal and current.
85
Consequently, the case studies they present during training will be up-to-date, directly relevant
and culturally appropriate. Training which is delivered around industry specific process
simulations, such as customer account opening, has proven to be an extremely effective way of
transferring appropriate levels of Six Sigma expertise in this sector.
86
customer experience. A first step should be to define and monitor relatively simple indicators
for the central tendency and the variation of the process. With non-normal data sets, it is
necessary be concentrating on the median rather than the mean. If the mean is significantly
different from the median, this is a clear indication that there are big outliers within the data
and/or that the distribution of the data values is skewed. To measure and manage the variation in
a service process, it is advisable to look at percentiles.
In a manufacturing process, it is possible to hire a post-graduate with a technical background,
give them 5 weeks of Six Sigma training. Even Manufacturing companies which are
experienced in the successful deployment of Six Sigma in their production and design
operations are not getting comparable results when they use the same approach to tackle their
non-manufacturing transactional and administration areas.
It is not really possible to train and prepare someone for the organisational, cultural and political
challenges they will face as a Black Belt in a service business. Black Belts in this industry need
a manager or coach with meaningful commercial experience, who has fought the cultural and
political battles involved in implementing Six Sigma projects in a people, based business. There
is only one way to gain that knowledge.
In a Financial Services environment, there are often opportunities to make immediate common
sense changes when the solution seems obvious. Alternatively a team may have spent months
on a project and developed a process and organisational solution that will transform process
capability and performance against customer expectations. However obvious that change may
be, it is worth remembering that the implementation will involve new staff working practices
and possibly the removal of age old habits.
87
Over the last three years or so, with the economy and markets in a state of depression, many
financial businesses have focussed actively on cost reduction programmes, but with the recent
uplift in economic conditions, there has been a shift in focus from cost cutting towards service
enhancement and improving the customer experience. There is likely to be a continuing change
of mindset in the services industry towards process orientation and understanding processes
end-to-end from the customers perspective. Consequently, Six Sigma is destined to gain further
traction in the Financial Services industry in Europe, but in the context of a growing realisation
that in service and transaction businesses, it is truly a different ball game. Companies deploying
Six Sigma methods are becoming increasingly knowledgeable and thoughtful of how they
should launch their improvement programmes, supported by the growing number of successful
case studies in this sector that have steered a different path to the traditional company-wide Six
Sigma roll-out. As was so often said at GE, these are great management practices, which should
be integrated into business as usual. In Financial Services, Six Sigma should simply be the way
of working.
The next stage is centred on the process of implementing Six Sigma in a bank, how to
implement it and the benefits that different banks have obtained by applying this methodology.
88
A new competitive scenario has become visible in banking. Liberation and the globalisation of
financial services have produced a rapid expansion of new financial products and services.
During the last few years, banks have implemented numerous new technologies and marketing
programs to improve service delivery and revenues. Banks offer an enormous range of financial
new products through constantly expanding branch networks. All these new products, such as
investments and insurance, need to be introduced into the new competitive market. For this
reason and in order to meet with customer requirements, it is necessary to implement Six Sigma
as an idiosyncratic factor.
The tools and techniques that have worked in manufacturing industries cannot be simply
adapted and applied to Financial Services businesses. Even those manufacturing companies
which are experienced in the successful deployment of Six Sigma in their production and design
operations often do not achieve comparable results when they use the same approach to begin
their non-manufacturing service, finance and administration areas.
The phases of implementing Six Sigma, in order to improve customer satisfaction in bank
branches, are almost the same as described in the other chapters of this project. The first step is
to define the factors that result in customer dissatisfaction such as wait times, errors and
inaccurate customer information are some of the factors that need to be analysed. Nevertheless,
not all factors can be measured in the same quality scale. For example, customers wait time
normally, although not always, have the most impact on customer satisfaction. In order to apply
Six Sigma to customers wait time it is necessary to determine the activities that affect wait time,
89
such as the number of tellers and their activity. Once the factors that affect wait time have been
determined, it is necessary to minimise their impact, find out potential solutions and implement
them.
Six Sigma can be used for much more than reducing customer-waiting time. It can be used to
enable managers to make relevant decisions. For example, banks have reduced fraud using Six
Sigma.
Successfully implementing Six Sigma in banking requires a relentless focus on customers and
on meeting their needs as efficiently as possible. First, it is necessary to define what is critical to
customers and confirm which core processes are based on those requirements. This is probably
the most unpredictable step because it is necessary to understand the full process from the
beginning to the end. Sometimes a process depends on different departments and each
department probably has a clear understanding of the process in its area but a full map of the
process does not exist. For this reason, the Six Sigma team needs to understand the full process
in order to achieve the required goals. There are many ways to find out the customer
requirements and needs such as surveys, call center results, focus groups and whatever means
allows the voice of the customer to be heard loudly and clearly.
The next stage is to translate customer requirements into measurable characteristics of the
process. Once the team has a clear understanding of the customer requirements, it is able to
measure process effectiveness and efficiency based on those requirements. The effectiveness of
the process determines the problem or defects that the process can produce. For example, if
delivery of the product or on time service is really relevant for the customer, then the metric
90
would typically be on-time delivery as measured by the time from the promised date to the date
of actual delivery.
Once the team has defined the possible measurements, it is necessary to determine the cost of
poor quality. This means determining the cost of not reaching the required quality levels and
quantifying the impact of particular gaps in the processes. For example, a mortgage lender
whose customers want prompt action on their applications might find that the process includes
a high number of abandoned customer calls by , long delays in producing a quote that cause
prospects to drop out, and numerous inaccurate credit reports. The Six Sigma methodology
includes powerful tools for analysing each of the gaps and quantifying what the poor quality is
costing the company in each case.
Having clearly understood what each process gap cost the organisation, the team can begin to
prioritise the improvement efforts according to what is most critical to the customer and less
expensive for the organisation. The team needs to ensure that they are investing in the right
project and in the right order because the prioritisation of some projects against others could
produce a different and unexpected output.
In order to understand the process of implementing Six Sigma better and having good approach
to the real process, it is necessary to look at some Six Sigma banks such as Citi Bank and Bank
of America.
91
5.3 Citibank
Citibanks objective is to be the most important international financial company. In order to
achieve this ambitious goal, the company had to acquire quality initiatives that satisfied
customers quickly and differentiated them from other banks. In recent years Six Sigma quality
has always been applied in the manufacturing area but Citibank selected this methodology to
meet customer satisfaction. This bank has achieved substantial performance in process
timelines, cash management, and customer satisfaction, applying methodologies such as cycle
times helped with the detection of defects and organised teams of Six Sigma.
Citibank has discovered the benefits of cycle time reduction in financial areas such as customer
banking and emerging banking. The Citibank Six Sigma team developed the process map and
looked for defects in each step of the processes, in order to achieve a cycle time reduction. Once
the defects were located, the team began to eliminate wasteful steps. For example, when
Citibanks customers wanted to transfer money from their accounts, they had to call their banker
and then phone, fax or mail in the requests of the transaction process. Customers used to
complain because they always had to wait to the confirmation of their request.
When the opportunity areas have been identified, the Citibank team was able to correct the
problems in the Six Sigma program. Citibank's principal purpose is to reach a cycle time
reduction by 10 times by the end of 2000 and another progressive defects reduction and cycle
time by 10 times every two years from there on. Not all the Six Sigma tools, which have been
described in the previous chapters, can be applied here because a bank is a special case. In the
case of banks, the Six Sigmas team usually uses easy tools such as Pareto charts. These charts
92
show the problems, which occur more often, and the team; based on this data, decide what
needs to be corrected first.
By using Pareto chart, Citibank realised that one of the principal problems was the internal callback procedure. This required staff to return to call the person who had ordered the transfer of
funds to make sure that all the data was correct. In this case, the reduction in return calls to
people who had ordered fund transfers of the cut in-coming transactions by 73%.
Citibank Six Sigma teams are usually made up of 30 to 50 people in each functional
departments. Then the Citibanks team, in the implementation process, follows five phases. First
of all, it defines the problem. Next, the team tries to map all current steps and identify what is
not working. The team then confirms the accuracy of the map with all the departments involve
in the process and it can be modified if they think that are necessary or they find another new
problem. After finishing the process map, the team removes useless process and non-value steps
and creates new ones. The last phase is the implementation of the change of items to reach the
required goals.
Some of the goals, which have been satisfied in two of the Citibank divisions, are:
The Private Bank has achieved a reduction of between 80 and 85 percent in internal and
Global Equipment Finance: This subdivision has improved all processes from the
reception of customer orders to the delivery of the product. They have also reached a 67 %
reduction in the credit decision, from three to two days.
93
Copeland Companies. They have improved the accuracy and timeliness of statements.
They have also reduced the cycle time of processing statements from 28 to 15 days.
Citibank has achieved great results applying Six Sigma in all its divisions. In the next step, how
the bank of America applies Six Sigma is described and the results which they have obtained
implementing this methodology.
94
The Bank of America is one of the largest financial organisations, which after trying to
establish other quality strategies and efforts without any improvements, introduced the Six
Sigma philosophy in 2001. In the beginning, some people were sceptical about it but now Six
Sigma has become an integral part of the culture of the bank. This discipline was implemented
as a core process performance metric, business approach and a leadership philosophy. Some of
the results that the Bank of America has achieved applying Six Sigma are:
The benefits of the implementation started rising quickly and the bank achieved an important
cost reduction in several areas. With the enthusiasm and knowledge of the BBs and MBBs,
quality training was extended throughout the organisation. Then Six Sigma tools were
implemented and utilised across the Bank of America. Businesses and engineering teams were
collaborating on projects to eliminate variation and errors in key processes. Based on the
principie of spreading the discipline across all Bank of America businesses, the Six Sigma team
95
demonstrated that this methodology could work in financial services by focusing on customer
issues and performance gaps.
For example, Six Sigma helped teams reduce system problems that occur in hardware and
software systems. Thirty Six Sigma projects were concentrated on the root causes of these
defects, reducing defects by 88% through the end of 2002. The customer delight was increased
by 25% across the company; in some operational areas of the company, they achieved even
greater gains. These good results convinced leaders that Six Sigma works and they then desired
to reach a better quality level.
Another important aspect to achieve the required Six Sigma level is that not only the
organisation needs to establish Six Sigma methodology but also all the vendors and suppliers
have to be involved in this quality program. For this reason, during the last two years, the Bank
of America has established the objective to become a real Six Sigma organisation and this
requires that all their processes reach a high Sigma level including the processes which involve
suppliers, vendors, etc.
This project has been developed in order to understand the Six Sigma implementation process.
With the intention of approaching to this methodology and showing the possible difficulties,
which can arise during the implementation process, and the benefits of Six Sigma, some
published case studies have been analysed. The reason to study the impact of Six Sigma in
Banking is a way to show that Six Sigma is emerging and influencing in all the areas not only in
manufacturing but also in the service industry and evidently in banking. Nevertheless, this effort
96
has not reached its leitmotif yet and needs to continue growing and improving in all aspects
from the employed tools to the team training.
97
7.0 REFERENCES
Tennant, G. [2001]. Six Sigma: SPC and TQM in manufacturing and services. England: Gower
Publishing Limited.
Eckes, G. [2003]. Six Sigma para todos. Espaa: Ediciones Granica S.A.
Rico, M. [2002].Seis Sigma para directivos. McGraw-Hill. Madrid.
Pande, P., Neuman, R. and Cavanagh, R.N. [2000]. Six Sigma Way. McGraw-Hill.
George, M., Rowlands, D., Price, M. and Maxey J. [2002]. Lean Six Sigma Pocket toolbook.
McGraw-Hill.
Catherwood, P. [Aug, 2002]. Whats different about Six Sigma? Manufacturing engineering.
pp. 186-189.
Antony, J., Banuelas, R., Knowles, G. [Aug, 2002]. Implementing Six Sigma. Manufacturing
engineering. pp.181-185
Reilly, J. [Aug, 2002]. Recession proofing Six Sigma. Manufacturing engineering.
pp.178-180
Morgan, J. [Jan, 2001]. Six Sigma meets ISO 9000. Quality world, Vol 2, Issue 1, pp. 22-24.
Hutchins, D. [May, 2001]. The sigma enigma. Quality world, Vol 2, Issue 5, pp. 40-42.
Cleary, M. [Dec, 2002]. Just another statistic? Quality world, Vol 28, Issue 12, pp. 44-45.
Caplan, F. [Nov 2002.]Six Sigma success must begin at the top. ASQ Six Sigma Forum
Magazine. Milwaukee: Vol. 2, Issue. 1; p. 8.
Lynch, D. and Cloutier E. [Feb, 2003]. 5 steps to success. ASQ Six Sigma Forum Magazine.
Milwaukee: Vol. 2, Issue. 2; pp. 27-34.
Hopen, D. [ Feb, 2003]. The softer side of Six Sigma. ASQ Six Sigma Forum Magazine.
Milwaukee: Vol. 2, Issue. 2; p. 8
98
Lewis, K.[ May,2003]. Featured company: Bank of America. ASQ Six Sigma Forum Magazine.
Milwaukee: Vol. 2, Issue. 3; p. 52.
Fundin, A., Cronemyr P. [Nov 2003]. Use Customer Feedback To Choose Six Sigma Projects.
ASQ Six Sigma Forum Magazine. Milwaukee: Vol. 3, Issue. 1; p. 17.
Godfrey, B. [May, 2004]. Is Six Sigma Here To Stay? ASQ Six Sigma Forum Magazine.
Milwaukee: May 2004. Vol. 3, Issue. 3; p. 6.
Faltin, F., Hahn, G. and Redman, T. [May, 2004]. ASQ Six Sigma Forum Magazine.
Milwaukee: Vol. 3, Issue. 3; p. 31-36.
Websites
Why Six Sigma is not TQM (online) [November, 2004]. Available at:
www.pyzdek.com/six_sigma_vs_tqm.htm
Gack, G. [2002]. Six Sigma Software Development Case Study (online) [January, 2005].
Available at: www.software.isixsigma.com/library/content/c030528a.asp
Key Ingredients for a Successful Six Sigma Program (online) [October, 2004]. Available at:
www.onesixsigma.com/experience/white_papers/whitpaper_paes/antony_successful_ss.php
Quality Deployment Function (online) [November, 2004]. Available at:
www.isixsigma.com/tt/qfd/
Six Sigma at Citibank (online) [March, 2005]. Available at:
www.qualitydigest.com/dec99/html/citibank.html
Six Sigma at Bank of America (online) [March, 2005]. Available at:
www.asq.org/economic_case/markets/pdf/bank_of_america.pdf
Goyal, N. [2004]. Medical Transcription Case Study (online) [January, 2005]. Available at:
www.isixsigma.com/library/content/c030203a.asp
99
Hallowell, D. [2004].A Six Sigma Case Study for an Information technology (online) [January,
2005]. Available at: www.software.isixsigma.com/library/content/c030528a.asp
Selecting Six Sigma Projects (online) [November, 2004]. Available at:
www.onesixsigma.com/_lit/white_paper/palomaconsulting_wp1.pdf
Adopting Six Sigma (online) [June, 2001]. Available at:
www.qualitydigest.com/currentmag/html/asixsigma.html
Integrating Innovation into Design for Six Sigma (online) [February, 2002]. Available at:
www.triz-journal.com/archives/2002/02/a/index.htm
100