Design for Six Sigma 

 July 8, 2022

By  Dave Litten

Design for Six Sigma 

Before we get into Design for Six Sigma (which is for new processes, products, or services), first let’s check out how Lean Six Sigma deals with existing processes:

Improving Existing Processes Dmaic - Design For Six Sigma

Improving Existing Processes – DMAIC

DEFINE: Identify the problem and describe what needs to be improved
MEASURE: Describe the problem more effectively by gathering facts and data to understand how the processes work and perform
ANALYZE: Use the facts and data to determine the problem’s root causes
IMPROVE: Identify potential solutions to the problem, select the most suitable, and test or pilot it to validate your choice
CONTROL: Implement the solution and use data to check the level of performance improvement,

Design for Six Sigma (DfSS)

Design For Six Sigma Dfss
Design For Six Sigma

These modules are a companion to our Projex Academy Lean Six Sigma Masterclass:

Lean Six Sigma used the DMAIC Method, whereas DfSS uses the DMADV Method – Define, Measure, Analyse, Design, and Verify

DMAIC Is used to improve existing processes, but sometimes the process does not exist and you need to create one, perhaps for new services or products

It may be that your current process is so poor that scrapping it and starting again makes more sense! DMADV Will help introduce new products and services quickly and to a consistently high standard

Design For Six Sigma Dfss
Design For Six Sigma

When redesigning a process we focus on the customer, when designing a new service or product there may not be a customer yet, so we focus on the demands of the (potential) marketplace

Where the customer is involved, this means both end-user customers and internal business stakeholders and users
Customer requirements and the resulting CTQ’ s are established early on and the DMADV framework ensures that these requirements are satisfied in the final product, service, or process

As with DMAIC, we managed by facts and not speculation to ensure that new designs reflect customer CTQs and provide real value to the customers

Design For Six Sigma Dfss
Design For Six Sigma

DMADV projects are more resource-hungry than DMAIC projects in terms of people, IT involvement, and cost
Despite these higher costs and risks, DMADV projects bring higher rewards

In large organizations using Lean Six Sigma, 20 DMAIC projects will likely be carried out for every one DMADV project

There are likely to be many lean projects, a moderate number of DMAIC projects, and a few DfSS projects as the organization gradually reduces rework and waste, improving process flow and reducing cycle times. DfSS often concentrates on the ‘delighter’ curve in the Kano model

Design For Six Sigma Dfss
Design For Six Sigma

DMADV Projects are normally used to introduce radical change within an organization.

It is important that you have a change management framework in place to address and manage the various people issues that are so important for the successful completion of the project

The Dmadv Define Phase
Design For Six Sigma

The DMADV Define Phase

This is where scoping, organizing, and planning the journey for your project occurs

It’s where you understand the purpose, rationale, and business case, as well as define what resources you need to help you and how you will manage the project

Here you will understand the boundaries of the project, including the processes, markets, customers, and stakeholders. It is vital that everyone understand why the project is being undertaken.

Using the affinity and inter-relationship diagrams can be helpful when starting a DfSS project

These also identify the key areas for the project’s success, bringing the team together, while creating involvement and ownership for the team members within the project

The Dmadv Measure Phase
Design For Six Sigma

The DMADV Measure Phase

The measure phase provides the framework around which the design can be built and is used to make design decisions needed in further phases

Here, the focus is on defining and understanding customer needs, and the different customer segments

DfSS projects seek to optimize the design of products or processes across multiple customer requirements, so a detailed understanding of these is essential

Project needs are translated into measurable characteristics (CTQ’s) that provides the requirements for the product, define the measures, and set targets, and CTQ specification limits service or process

For new products or services, you must make sure the design can be produced with existing processes or designing new processes to accommodate such new designs

Process capability is considered in the measure phase rather than after the design is complete (as in DMAIC)

The DMADV Analyse Phase

Here is where the functional specification and high-level designs are developed, going on to create and test the detailed design

The Analysis phase moves from the ‘what the customer needs’ to the ‘how we might achieve it’

Mapping the CTQs onto internal functions and looking at alternative design concepts

The Dmadv Analyze Phase
Design For Six Sigma

“Some people think design means how it looks, but if you dig deeper, it’s really how it works”
Steve Jobs

“Design means how something works, not how it looks – the design should evolve from the function”
James Dyson

For a service, the analysis identifies the key functions; for a more tangible product, the analysis identifies its key part characteristics. The sub-system characteristics are developed next followed by the components (parts) of the sub-system

The Dmadv Analyze Phase
Design For Six Sigma

Functions are what the product, service or process has to do to meet the CTQs identified and specified in the design process

For a service, functions are key high-level processes to be considered

You will carry out an analysis of the functions to understand their performance capability to ensure they are fit for the purpose

The DfSS project emphasis is on alternative design concepts and the generation, analysis, and assessment of a high-level design. Analysis analyses and selects the best design concept and starts to add more detail to it

Each design element is considered in turn with high-level design requirements for each. Consideration needs to be given to how each component fits and interacts with each other

The Dmadv Analyze Phase
Design For Six Sigma

The analysis phase usually creates several high-level designs, assessing their suitability and selecting the best fit
You need to assess performance and develop design scorecards to help analyze how capable the design is to deliver the CTQs. Such assessments may be carried out through simulations, field tests, or pilots involving the customer to capture their feedback

The Design Scorecard method deploys the requirements into the design and predicts the capability of the new design. The Design Scorecard captures the critical performance measures at each level and visibly tracks the measured performance as the design evolves

The Balance of Measures

It is important to get the balance of measures and understand how they interrelate. The different variables will all need corresponding measures to help you assess performance. To fully understand the performance of your process, you need a balance of input, in-process, and output measures with one to three measures for each. Your input and in-process variables will influence the results in your output variables, so your measures should clearly link together

The Balance Of Measures

Failure Mode Effects Analysis (FMEA)

Based on the outputs of the review, the high-level design requirements can be finalized and a thorough risk assessment undertaken using FMEA. For each of these potential events (failure modes), you assign a value, usually on a scale of 1 to 10, to reflect the risk

A failure mode is a mechanism by which the system fails, or the process fails. A failure mode usually impacts the customer or business

Failure Mode Effects Analysis Fmea

The DMADV Design Phase
Developing the Design

The Dmadv Design Phase

The design phase starts by developing the ‘how’ thinking in more detail Adding this to the various elements of the high-level design. The emphasis is on developing designs that will satisfy the CTQ requirements of the process outputs. The lowest level specification limits, control points, and measures are determined to form the basis of the control plan that must be in place following the implementation

The design is specified at a detailed enough level to develop a pilot and test it because, before implementation, you need to pilot the design. Enough detail should now be available to test and evaluate the capability of the design by preparing an effective and realistic pilot in this design phase

Learning Today Leading Tomorrow!

All Courses for just $15/month with a simple annual payment

The DMADV Verify Phase
Verifying the design works

The Dmadv Verify Phase
Design For Six Sigma

The design is piloted and assessed in the verify phase, and apart if any adjustments following the pilot, implementation, and deployment. The final step in the cycle is to assess the achievements made and lessons learned
The results are verified against the original CTQs, specifications and targets. Once the solution has been standardized and transferred to operations and process management, the project is closed.

During the handover to the process owner or operational manager, you must work closely with your team to ensure a well-planned and well-documented transition

Choosing between DMAIC and DMADV

Choosing Between Dmaic And Dmadv

You may start your project using the DMAIC method yet find yourself at some point changing to DMADV
If so, here are the likely decision points in transitioning from one method to the other:

Quality Function Deployment (QFD)

Quality Function Deployment Qfd

QFD is also known as the House of Quality

Customer research is used to determine the voice of the customer. Competition research helps us hear the voice of the market. Quality Function Deployment helps translate both voices into high-level requirements (the CTQs) and to place measurable definitions, specifications, and targets on these requirements.

In the analysis and design phases, QFD is further used to develop specifications for the lower-level details of the designed product, process, or service so that high-level customer CTQs can be satisfied

The House of Quality

The House Of Quality

QFD Is a graphical representation of the logic flow, from identifying customer requirements to the detailed development of actions to ensure those requirements are met. A series of interconnected matrices are developed, moving from the requirements, through to design, implementation, and deployment

The House Of Quality

The House of Quality “room numbers “, represent the logic flow order for completing the QFD
I will now take you through each “room” in more detail …

QFD – Customer Needs ROOM 1

The first thing you need to do is identify and segment your customers, build a data collection plan, and conduct your customer research. You now have a good amount of voice of customer information. Here you are focusing on organizing the information that you have collected and then interpreting and translating that information into a set of CTQ statements

Qfd Customer Needs

QFD – Competitive Information ROOM 2

After collecting and completing the CTQ’s their relative importance is established where the QFD approach initiates a trade-off analysis. Here you will determine which requirements are the most important so that when you enter the design phase you must understand the priorities from the customers’ viewpoint. The customers should be asked to rank you against the competition as part of organizing and prioritizing customer needs.

This comparison can help your team determine the strengths and weaknesses of their current product, service, or process with the resulting information being input to the target setting process for CTQs

Qfd Competitive Information

You will want to assign a symbol for your own organization and one for each of your competitors who should be market leaders or the most distinguished organizations that you compete with. For each need in Room 1, ask your customer to assign a rating between 1 (lowest) and 5 (highest) for both your company and your competitors

Qfd Competitive Information

The highest rating on the scale is reserved for how the perfect service performs. Clearly, ‘Our Company’ is not doing very well!

QFD – Characteristics and Measures ROOM 3

Qfd Characteristics And Measures

Here, we start moving from the ‘what‘ to the ‘how’ of the customer requirements. So far, we have only understood what the customer requires – now we need to understand the characteristics and measures needed to ensure the end design meets those requirements.

For each customer attribute, you must determine what characteristics and measures will indicate how well you are meeting their needs. You will want to develop measures for which targets and specification limits can be established. Do remember that measurements can produce two types of data – continuous data, such as time or temperature, and discrete data, such as accuracy

QFD – Relationships ROOM 4

Qfd Relationships

Here we carry out some analysis using the CTQs derived from Room 1 and the characteristics and measures described in Room 3 and start to draw relationships between them. Our purpose is to ensure the requirements for each characteristic and measure are considered. Remember that Room 1 concerns the ‘whats’ and Room 3 concerns the ‘hows’

Here in Room 4, we are asking: “Can this ‘how’ achieve that ‘what”

Qfd Relationships

For each relationship (Room 1) and characteristics (Room 3), perform the following steps and complete the relationship matrix below:

  • Rate each relationship using a scale of Strong (9), Medium (3), Weak (1), or none (0)
  • Calculate the score for each cell by multiplying the priority for the customer need in Room 2 by the (9, 3, 1, 0) value of each related cell
  • Once the relationships have been rated, add up the individual scores for each measure to determine their importance

QFD – Competitive Benchmarking ROOM 5

Qfd Competitive Benchmarking

Benchmarking refers to looking both inside and outside of your own organization to see how well others are doing at providing products and services like yours (performance benchmarking), and how ‘best practice’ organizations provide their products and services (process benchmarking). Get competitive data for every key measure and analyze it using a 1-5 scale (1 = poor performance, 5 = best-in-class) with a separate row for each 1-5 grading

Qfd Competitive Benchmarking

For each measure in Room 3, assign the symbols to the appropriate grading. Qualitative benchmarking was done in Room 2 where customer perceptions of your qualitative performance in relation to the CTQ’s were established.

Quantitative benchmarking performance was done in relation to the measures established in Room 3 – Here, we are looking at actual performance as opposed to perceptions about performance

QFD – Targets and Limits ROOM 6

Qfd Targets And Limits

Using the results from the Room 5 benchmarking, we now set the goals and targets against the measures and characteristics defined in Room 3. Setting targets and specifications is a matter of judgment using business know-how and technical expertise along with the use of tools.

Two tools worthy of mention are the analysis of benchmarking data and a good understanding of customer requirements using the Kano model.

Kano’s Model – evaluating requirements

Kano'S Model Evaluating Requirements

Delighters are expectations the customer was not expecting, and delivery quickly satisfies the customer, plus non-delivery has no downside. Wants are the stated requirements from the customer who will be satisfied or dissatisfied in proportion to your achievement. Identifying and delivering delighters are business success critical, they innovate and differentiate you from the competition. Needs are expected requirements they are a ‘given’ and satisfaction is neutral if achieved, if not, the customer is dissatisfied

QFD – Correlation ROOM 7

The roof looks at the impact of each of the measures on the CTQs and how the measures affect each other. You first need to examine each measure and assess the likely impact of increasing, reducing, or hitting the target for that measure on customer CTQs.

Next, you examine the relationship between each pair of measures to understand the impact and effects of any relationship on the final design. Carry this out by assigning one of these four symbols to represent that relationship:
++ strong positive, positive, negative, and — strong negative

Qfd Correlation

This shows how the various characteristics and measures impact the end design – the objective is to resolve the conflict in situations before you build the design. Innovation may be required to achieve two apparently conflicting measures. A trade-off between measures may be required.

Good solutions and designs will not require trade-offs but satisfying negatively correlated measures requires time and forward-thinking. Conflict resolution should focus on meeting the customer’s needs not yours and measures with strong positive correlations can become part of the overall design strategy

Qfd Correlation

Developing more Houses of Quality
QFD Drill-down

The aim here is to develop further houses of quality slowly refining the level of detail until the design is specified at an implementable level. Moving from one house to the next you carry over the corresponding targets and importance measures where the number of houses used in drill-down may vary – increasing with the level of complexity.

In the second House of Quality (developed in the analysis phase), the CTQ measures and the targets/importance ratings are taken from the first House of Quality. You find the correlations to enter in the cells by asking “if I design this particular function correctly, what impact does it have on my ability to meet the CTQ’s measures/targets?”
This helps identify where the design effort must be concentrated to satisfy the CTQs and thereby the customers

Qfd Drill-Down

The Pugh Matrix

The Pugh Matrix

As the QFD house constructions evolve decisions need to be made about the various design concepts and ideas throughout the DMADV phases, and The Pugh Matrix is often used to carry this out.

The Pugh Matrix

The Pugh Matrix (controlled convergence) is a framework for comparing solutions/concepts against a set of predetermined criteria. The Pugh Matrix helps refine competing designs by improving the S and – rankings to + and combining the + attributes into a super alternative

The Pugh Matrix is also used to aid the selection of the best design. It is most often used in ‘design projects’. The tool provides a structured way to evaluate alternative or competing concepts and benefits from being both non-numeric and iterative.

The list selection criteria against which the concepts will be evaluated is crucial and you should already have a detailed understanding of customer needs, Business, and internal stakeholders. The final list of criteria should be unambiguous and must be agreed upon by the full team.

Although not necessary, weighted criteria are helpful in helping you to focus on the key concepts. Choose one of the concepts as a reference concept ideally something that already exists and represents best in class. Then compare each concept with the reference scoring if it is better or easier, or worse or harder – If the same then market as ‘S’

The Pugh Matrix

For each concept add up the plus and minuses and rank them in preferential order


Projex Academy’s Lean Six Sigma Masters Series has been designed to fully prepare you for the Green Belt (Practitioner) Exam.

Lean Six Sigma Masters Series

Lean Six Sigma Masterclass – Everything you ever wanted to know about DMAIC


Boosting the core fundamentals of L6S 

  • Lean Principles
  • Six Sigma Process
  • Kanban
  • Just In Time (JIT)
  • Heijunka Leveling, Sequencing and Standardizing
  • Takt Time
  • … and much much more

Lean Six Sigma takes the features of Lean ( speed), and combines them with Six Sigma (stability and accuracy). This Lean Six Sigma Green Belt Masters Series will teach you how to streamline processes, improve business performance and supercharge your career

Dave Litten

David spent 25 years as a senior project manager for USA multinationals, and has deep experience in project management. He now develops a wide range of Project Management Masterclasses, under the Projex Academy brand name. In addition, David runs project management training seminars across the world, and is a prolific writer on the many topics of project management.

The Projex Academy

related posts:

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}

Project Management Masterclasses