Understanding Kano Analysis


Products have to be consumer centric. The manufacturers of any product must have a clear idea of what purpose their product is promising to serve. In other words, a clear-cut definition of the functionality of the product is very important. A pudding is good only if the person who is eating approves it (in other words, the test of a pudding lies in its taste!)

Kano Analysis

Kano Analysis, as proposed by Noriaki Kano, a noted Japanese Researcher focuses on how to improve quality by addressing needs of a consumer. A lot of emphasis is placed on Quality assurance, Statistical Quality Control, Total Quality Management and a host of other production-improvement techniques. Sometimes, in the mechanics of production feasibility, the needs of the consumer are sidelined. Ideally, the producers make products to be sold to the final consumer as shown in the below ‘gear’ figure. All the steps are interlinked. The Kano analysis works backward, and suggests that the consumer’s needs should be assessed first and then the product should be fine tuned for improving performance.


In this context, it is important to think on the following lines:
1. What are the needs of your consumers?
2. How is your product’s functionality serving that need?
3. Are the consumers satisfied with the product that you offer? What is the level of satisfaction and dissatisfaction, if any, and what are the associated causes?
4. Are there ways by which you can increase consumer satisfaction?
5. Is it possible to classify product features and how each, or a group of them, serve consumer needs?
6. Is it possible to eliminate any unnecessary features (and also reduce cost) without compromising the functionality of the product?

How do consumers evaluate a product/service?

According to the Kano Analysis, it is important to understand what your consumer wants and then relate his wants to what your product can offer. This can help you work out a feasible strategy for segmenting, targeting and positioning of your products. Ideally, consumer preferences can be divided into many categories, but for practical purposes we will take a look at the three basic sub-sets.

1. Basic Needs (Also referred as ‘Must Haves’): A product must serve its basic purpose. A blender must blend well. A muffin must taste good. A high definition TV must have excellent picture quality. A consumer expects that the product will serve its basic function; he will mention that his car is great, but may not always mention that the choke plug, tires, battery and spark plugs are good too. He would notice that his refrigerator is very nice, but he may not always mention the thermostat, the door seal, the crisper cover, the door switch or the light. According to Kano’s analysis, there are product features which are absolutely essential for overall product performance, but looking from the consumer’s perspective they should just work well. There is no need to add extra cost to enhance these features, as a consumer would not be able to tell the difference. His level of satisfaction will not increase/decrease if you increase the cost of such features (For him ‘they should just work!!’).
2. Performance Needs (Also referred as ‘More the Better’): A consumer expects his product to perform well and this is imperative for satisfying his need. If he is unhappy with the performance of the product, he may never purchase it again.
3. Premium Needs (Also referred as the ‘excitement features’): Often we hear people saying,” Wow! I did not know that this device has such cool features’ Features that totally excite a consumer and give them an additional value for the price that they have paid are addressed by Kano’s Model as ‘Excitement features/Attractive features’. It is observed that ‘buy one get free offers’ are very popular in consumer products and excite buyers a lot. Many hotels offer coupons for free rides and other incentives which attract the potential tourists.

Conclusion: It is clear that Kano’s Analysis offers concrete pointers about how the product is perceived by a consumer, what he values most and what he does not care about so much. All that is needed is to ‘explode’ the product and relate it to how best the consumers needs are being met, how best can the consumer needs to be met potentially and how best can cost be reduced without compromising the satisfaction level of the consumer.

Leadership Failure: Are You Stumbling?

Understanding the Trap

Leadership failure can happen more often than you think. According to a recent survey, over 40% of new Chief Executive Officers (CEO) fail within their first 18 months. That is a very powerful statement! Newly hired CEO’s enter the arena with confidence and attitude that is success focused. They wouldn’t have gotten past the minefield of the interview process without the mindset and attitude required. However, for some, it’s not about having the corner office in the C-Suite or having the control to make crucial decisions that direct a company. Some CEO’s thrive on these trappings of success. Others understand that the buzzwords like “integrity” and “collaboration” are crucial to their success. But, before long, what they professed as their core values quickly change, to the disappoint of the board and their subordinates.

Preventing Leadership Failure

Unfortunately, most organizations don’t do a great job supporting their leaders. They don’t set or manage the behaviors or accountabilities in a manner that is clear from the start. Too many times a CEO finds out they aren’t meetings expectations when they are let go. The system is broken and too many people are failing and taking down other good people around them, for no reason. Analysis tells us there are some major reasons why new leaders are unable to successfully perform:

  • Confidence – They are filled with confidence and bravado, talking the talk. But, when it comes down to delivering on all that bravado and bluster, they stumble and fail.
  • Leading with Wrong Expectations – They don’t have a clue what they are getting into. They are unprepared to deal with the people challenges. They become slow to react or hesitant to act at all.
  • Lack the Skill Sets – They lack the ability to lead and encourage other’s success. Skills required to lead and motivate are opposite of what it takes to individually perform.
  • Ignoring Relationships – Being a leader is all about relationships. New leaders pay a high price, ignoring the importance of trust and relationship building.
  • They Don’t Listen – Too many leaders believe that because of their position, they naturally have all the answers to every situation. They just simply don’t! They jump into decisions and situations that create havoc, without listening to the voices around them.

Avoid the 40 Percent

It is a sad state of affairs when we have a 40% leadership failure! Fortunately, it is a trend that can be changed. As leaders, we must first be aware of our own strengths and weaknesses. Otherwise, it is difficult to manage our own behaviors, the behaviors of others or to develop the proper relationships that build success. We must understand that focus and confidence are great attributes, but they must be tempered  with a touch of realism. Leaders should continually work on their leadership skills and sharpen them to a fine point. It is a cycle of continuous learning and building that keeps us on target and successful.

Efficient Use of Poka Yoke

With billions invested in manufacturing across the world and a lot of talk about Quality Assurance, there is immense focus on what goes on at the production floor space. ‘Statistical Quality Control (SQC)’, ‘total Quality Management’, ‘Zero Defects’ and techniques for Inspection focus on similar issues: How to reduce the number of errors? How to prevent defects? How to aim for Zero Defects, where the number of defects is equal to zero? How not to make mistakes?
‘Poka Yoke’, a Japanese term that means ‘mistake proofing’ was developed by Shigeo Shingo, a Japanese engineer, who developed the concept of Zero Quality Control (ZQC), as a part of his ‘Total Quality Management’ efforts.

His approach relied on the:
1. use of ‘Poka-Yoke’ (mistake proofing) devices that helped to prevent mistakes in manufacturing or
2. to inspect all the products using simple and affordable methods.

Although ‘Poka Yoke’ is equally applicable in non-product based manufacturing, but its most significant contribution is in manufacturing industries.
We must understand that when any company manufactures a product, be it dolls, shoes, sauce, electronics or computers, it has avoid defective products from reaching the final consumer. In this era of information, technology and competition, a manufacturer can no longer take his consumer for granted. He has to give the best quality and keep up to the standards that he promises. If a consumer is unhappy, it is likely that he will tell 10 people about his bad experience. But, if he is happy, he will be loyal to the products that serve him well. He will give a positive response to his friends as well.

So, how does a manufacturer apply common sense to manufacturing to make the number of defectives/errors less than before? Or, rather how can he delight his consumer by making a product that serves him well. By applying simple ‘Poka Yoke’ (mistake proofing) techniques and devices to your product, this problem can be solved if not eliminated completely. Over a period of time, many products have evolved to serve the same function more efficiently.

Following is a list of mistake proofing as applied to our everyday lives to avoid mistakes, reduce cost and live efficiently. This has been a result of products which have been changed a little for a lot more benefit.


  • Sensors in urinals and wash sinks detect when water is to be turned off. This eliminates the problem of wastage of water at public restrooms.
  • Timer-induced Lighting systems that turn off automatically after the duration set on the timer reaches its mark. This saves electricity in public washrooms, laundry rooms, etc.
  • While smoke alarms inform about a potential hazard, doors near a library/departmental stores can beep off if the materials are not checked out in the correct order. This reduces risks and makes sure that people check out borrowed/purchased materials properly.
  • The car does not start unless the ignition is turned on. The key does not turn off if the gear is in ‘reverse’ mode and the seat belt sign turns on if you forget your seatbelt. The gas level starts indicating if you need to refuel. These small indicators help us to drive safely and avoid accidents.
  • The word processor or the web browser prompts us to save current work before closing the session. When we fill online questionnaires and personal information, there is an alert message which helps us not to miss out any details.


Conclusion: These changes have been made as a result of product research, product development, quality assurance and total quality management. ‘Poka Yoke’ deals with how simple techniques can be applied to ensure that all products are of good quality at the ‘producer’s end’ and how simple product modifications can be made to increase satisfaction at the ‘consumer’s end’.

How Will Leaving The EU Impact HR & Employment Law?

As with any legal system we would all no doubt welcome the opportunity to cherry pick a legal framework that is easy to navigate. This of course must be a system that provides the necessary protection for people and organisations alike.

There is no doubt that the EU has provided us with a legal framework that we have now become accustomed to and have proactively adapted to work with, this will not change overnight. Let’s be clear, the EU has without doubt provided us positive aspects of law (as well as with some onerous processes).

We must remind any scaremongers that no employment law will automatically disappear once the UK actually leaves the EU. There will firstly need to be a formal repealing of any legislation that is to be removed. The UK Government will need to clearly bring forward a sensible case to challenge any laws it wishes to. We currently have no idea of the agenda to do so or indeed if an agenda to change any laws whatsoever even exists at this stage. This certainly applies to employment law. When and if article 50 is invoked we may have a clearer view of priorities and/or requirements, but it will certainly be a useful series of events for those of us with HR and legal responsibilities to follow.

There is also a huge amount of law in the UK that will be totally unaffected by BREXIT. We need to bear in mind that we will be required to be nimble to adapt to any variations and iterations of the legal framework, but that notwithstanding, the overall landscape should have a sense of familiarity to us all. Something we should all take responsibility for here is to ensure that we remain in control and project a message of calm; this will demonstrate to the global economy that we (the European HR and Legal community) are able and well positioned to guide and advise a safe route back to normality or stability.

The outlook for now has to be business as usual for us all. We know with certainty that aspects of employment law in the UK will change but we have no idea which aspects or when. This certainly does not scare me and I can’t help but think – “WHAT’S NEW?”!

Keep The Faith!  The UK/European HR and Employment Law communities are without question nimble and will continue to demonstrate this.

For those of you in HR there is a lot of useful information available about sensible steps to take post BREXIT and for what it’s worth I wanted to share the points that I have found useful.

Sensible steps to reassure the fantastic EU National Employees:

Email from CEO/MD to reassure the entire workforce that it is business as usual.

  • Contact all of your EU national employees – listen to any concerns and then obviously reassure that nothing whatsoever has changed to impact their legal right to live and work in the UK.
  • Contact/Register with the Home Office and DWP to ensure that you are receiving any relevant employer updates.
  • For any employees who need further reassurance, encourage them to contact the UK embassy for their respective nations, many of which already have working groups set up and mailing distribution lists that will duly notify people of any noteworthy changes or updates.  (this has proved a usual and helpful step to take)



QFD Can Help to Create a Robust Request for Proposals

When it comes to managing supplier quality, many people immediately think of controlling the quality of incoming raw materials as they enter a process. However, as companies look to shed their non-core business processes by relying on other specialized firms to perform the work, suppler quality takes on a much broader meaning. Supplier quality incorporates all the activities, materials, products and services provided to the business to ensure that quality is built into the product or service and that customer requirements are ultimately met.

Vendor management, the management of the relationship between the company and a third-party provider, is a challenge for most companies and is an important aspect of maintaining supplier quality.

Consequences of a Weak Process

A company that recently decided to outsource one of its functions, a backroom process that was not considered a core competency, makes an excellent example. After an extensive vendor selection process, the company chose a third-party provider, contracts were signed and the work was transferred. The entire process took months and management breathed a collective sigh when the new vendor was handling the task.

Then, things started to go wrong. The outsource provider was not living up to the requirements of the business. In its defense, the vendor claimed the company had expectations that were never covered in the service level agreement. The relationship between both parties quickly soured. The vendor became unwilling to handle ad hoc requests and was unresponsive to process improvement suggestions. Instead, the vendor followed the minimum contractual obligations.

When the company’s legal team became involved it found that there was no cost effective solution to terminating the relationship. One of the senior managers in the company asked in frustration, “Where was the quality rigor in this process? Wasn’t there a Black Belt assigned to this project? What could we have done to prevent this?”

Vendor management is a challenging task for most companies. At best, it is mutually beneficial and may offer process improvement opportunities for both parties through performance monitoring and mutual problem identification. At worst, the relationship can damage profits, reputations or customer relationships. To increase the likelihood of a positive outcome, it is crucial that the company select the right vendor – a provider that has the capabilities to meet the customers’ critical-to-customer elements (CTQs) and satisfy internal company requirements.

Many companies follow the request for information (RFI) and request for proposal (RFP) process to select a vendor. Initially, a pool of potential vendor candidates is identified. The company sourcing departments may use market analysis, benchmarking or use other techniques to help identify possible vendors. After the pool is identified, an RFI is sent to solicit high-level information from potential vendors in order to further narrow the field. High-level process maps along with general specifications are included in the RFI to determine if the potential vendor is able or interested in performing the work required.

Following the RFI, an RFP is sent to the final candidates. The RFP is much more detailed than the RFI and explicitly describes the performance expectations. Suppliers return their response and, when a vendor is selected, both parties sign a master services agreement (MSA). The MSA contains the legally binding service level agreement and statement of work, among other documents.

The RFP is the vital document in the vendor selection process, therefore a properly written and inclusive document is critical to the success of the future implementation. To the left is a generic table of contents for potential topics included in a typical RFP.

The House of Quality

An effective way to help ensure success in the RFP process is to employ quality functional deployment (QFD) during the early phases of the project, specifically during the vendor selection process. QFD was first developed in the 1960s by Japanese engineer Yoji Akao to improve manufacturing processes. Like most Six Sigma tools, QFD is applicable in many processes and environments.

QFD offers a structured approach for translating customer needsinto specific product or process attributes and allows the project team to systematically evaluate or prioritize each attribute in relation to meeting customer needs. Typically, the project team will fill out several iterative matrices, known as a “house of quality,” to drill down from the customer needs to specific attributes. The first house or matrix in the QFD translates high-level voice of the customer elements into CTQs. The second matrix translates the CTQs into functional requirements, while the third transforms functional requirements into design requirements. The last house converts design requirements into critical-to-process variables.

After establishing the initial house of quality the project team can weight each customer need. Within each matrix, the team determines the strength of the relationship between each element. Usually, a scale of high-medium-low is used to measure the correlation, which is later converted to a numerical value. A quick summation of customer needs multiplied by the correlation value will create the priority values at the bottom of the matrix. These weighted priority values become the relative weights used in subsequent matrices. The figure below demonstrates the relationship between successive matrices.

Relationship Between Successive Matrices

Relationship Between Successive Matrices

There are many software programs available that can assist in creating a house of quality. A simplified version of a worksheet in MS Excel is attached.

Each successive matrix in the QFD is used to help create the RFP and is completed during each phase of the DMADV (Define, Measure, Analyze, Design, Validate) Design for Six Sigma (DFSS) roadmap. This list highlights how the QFD is used in each phase and translated to create the RFP and evaluate the resulting vendor responses:


  • A cross-functional project team is formed.
  • Customers (internal or external) and stakeholders are identified.
  • Team gathers the needs/requirements of the customers.
  • High-level “as-is” process maps (used later in the RFI) are created.


  • The first house of quality is used to convert the generic needs into specific, tangible CTQs.
    • Identified needs, with their definitions, are listed in Section 1.0 of the RFP and help create the framework of the proposal.
    • The CTQs are inserted into Section 4.2, “Measurable Requirements” (Refer to Table 1).
  • The project team, or the sourcing department, identifies potential vendors and sends out RFIs as necessary to gather more information.
  • Second matrix of the QFD is completed.
    • Process functional requirements are identified and prioritized.
    • The description of these functional requirements is laid out in Section 4.1 of the RFP.
  • Detailed “as-is” and high-level functional mapping are completed. These are added to the appendix to assist the recipients of the RFP in understanding how the business specifically handles the process.
  • Benchmarking studies are performed to understand potential vendor capabilities.


  • Third matrix of the QFD is completed.
    • Potential design or technical requirements are identified.
    • These elements are added to Section 4.3, “Technical Specifications,” of the RFP.
  • Service performance matrix (SPMX) is completed. An SPMX contains detailed information regarding process performance and sets expectations for the vendor. Typically, the SPMX is broken down by critical Y, or high-level CTQs, and contains the description, unit of measure, target performance, specification limit, and how the data is collected.
  • The project team should have enough information to complete the RFP and submit it to potential vendors. Vendor responses can take days to weeks depending on the complexity of the process.


  •  Fourth matrix of the QFD is completed.
    • Technical or design requirements are translated into critical-to-process variables.
    • These variables will help the project team evaluate the responses received from the vendors. The results of this last house of the QFD can serve as a prioritized checklist and will help explain to stakeholders why a vendor was chosen.


  • Validate the vendor prior to signing any service level agreements or contracts. The project can establish a proof of concept, or pilot, in order to ensure the vendor can meet the CTQs identified in the first house of quality.

Leadership: Developing the Essentials

Are You Preparing?

Leadership is often misunderstood. Often times, people define leaders as those appointed to positions of authority above them. It is expected that if you hold a position of authority, then you are a leader. Those same people then follow those in positions above them, with sometimes reluctant obedience. They see that their ‘leader’ isn’t making decisions that make sense, or they are reckless in their behaviors, or they simply bully their workforce forward towards the stated goal. Being a leader requires more than education or holding a stated position. It is an art that is formed through character, experience and understanding. Leadership requires vision and understanding. It demands focus and a will to succeed. Leaders show the way, doing exactly what they are asking their followers to do. Too often in business, leaders are in their role simply because they have acquired a position of authority. They fall short in all of the other factors that make a truly great and inspirational leader.

Defining Leadership

The fact is that effective leaders are made, not born!  They learn from trial and error, and from experience. When something fails, a true leader learns from the experience and puts it behind them. Great leaders must know how to reward those who succeed and know when to retrain, move or fire ineffective staff.  A great leader knows that when they are successful, people follow you out of natural curiosity and trust.

 “A true leader has the confidence to stand alone, the courage to make tough decisions, and the compassion to listen to the needs of others. He does not set out to be a leader, but becomes one by the equality of his actions and the integrity of his intent.”

Douglas MacArthur

These traits are learned and earned. Some develop these traits early in their career, in a time where it appears they aren’t ready or it seems they don’t have the years of experience some may require. But one of the most significant traits of a great leader is that they are a great follower! Business environments and situations are fluid. That requires that leaders are assigned by a “best fit” for the situation. Identifying a leader by position is not always the right choice. There maybe another that possesses the right skills, at the right time, for the right situation. Established leaders must know when to transition from leading to following, in a smooth and seamless manner. It cannot be viewed as a demotion or professional transgression. Organizational needs dictate that the right person must lead to a achieve the desired results.

Being a True Leader

Being able to truly understand the role of a leader is the key to professional success. Positions do not define a leader and leaders are not defined by positions. A true leader knows when to be a strong and loyal follower, and a successful follower will just as easily rise to the occasion and lead! The real key to success as a leader is to master their craft, develop true compassion and understanding of others and have a focused intent for success.

Using Censored Data in Transactional Processes

Censored data is commonly used in reliability studies to determine the mean time to failure in order to establish warranty and maintenance periods for products. A large number of samples are subjected to either normal-use or accelerated-use conditions. Failure modes and occurrences are logged.

Plotting the distribution of the sample failures over time allows the prediction of the mean time to failure for the product being tested. If the warranty period is fairly short, or the manufacturer can wait until all the products fail before establishing the warranty period, then the testing is run until all products fail. If the failures require a long running time, and the product goes to market before all the products fail, then censored data analysis is used to predict the mean time to failure. The data is tagged to indicate whether the data point represents a sample that has failed or whether on the date the data is collected the sample is still operating. If the sample is still operating, the data point is considered to be censored.

Based on the data that has failed, the model establishes the likely failure points for the population of samples – predicting how much longer any censored samples will continue to operate. In Table 1, windings at 80 degrees Celcius failed at an average of 48 months, yet there is one sample still running at 99 months. Based on the remainder of the data (not shown in this table), 15 percent of the windings survive at least 99 months. Looking only at failed data points can skew the understanding of the actual performance of the population.

Table 1: Example of Censored Data
Winding at 80 C Time (Months) Censor
1 50 1
2 60 1
3 53 1
4 40 1
5 51 1
6 99 1
7 35 1

Figure 1: Survival Plot for 80 C


Transactional processes behave in a similar fashion, if the terminology is translated. Consider booking sales orders. In the process, a salesperson visits a potential customer and attempts to match the customer with a product on offer. In some situations, booking the order is rapid; the customer knows what they want, the salesperson can offer it, the price is accepted and the order is taken. In other situations, there may be negotiations or difficulties in matching the right product with the customer. It may take several contacts between the salesperson and the customer before the order can be booked. All of these sales visits that are pending a received order are censored – the end date is still unknown. If the booking has already become an order, then the event is no longer censored. Sales visits that do not result in an order have failed.

For example, Table 2 shows a number of customer contacts. Six of these contacts have resulted in an order in hand. These orders took an average of 1.3 days to close. There are three customer contacts still pending – no order received. These are censored data points. Sales still expects to receive an order from these customers, although the data in this table suggests that the pending sales contacts may never result in an order. Should the sales team move on to other contacts?

Table 2: Censored Data Points in Example of Sales Process
Sales Contact Date Time Elapsed (Days) Censor
1/8/2016 4.0 Order received
1/12/2016 0.5 Order received
12/4/2015 38.9 Pending
1/12/2016 0.3 Order received
1/9/2016 3.4 Pending
1/9/2016 2.9 Order received
12/24/2015 18.7 pending
1/9/2016 2.9 Order received
1/12/2016 0.1 Order received

Companies frequently predict bookings to establish production plans and sales visit schedules. Being able to predict how long to continue to work a potential customer based on the likelihood of achieving a sale can improve the efficiency of the salesforce. Capturing failure modes related to no receipt of an order can help salespeople adjust their approach to potential customers. Understanding the order projection can also help to adjust Sales’ efforts over the course of the reporting period. For example, if the mean time to close an order is two weeks (or two visits) and the end of the reporting period is approaching, it will be better for Sales to focus efforts on bookings that are in progress over efforts to land new customers.

How Censored Data Analysis Works

Censored data analysis requires that a mix of censored and uncensored data be available. In reliability studies, the start time is known for all the samples and at any point in time after that start time, each sample can be assessed as having reached its failure point or as still running. The distribution of all the data points, censored and uncensored, is understood through the science of the product performance, distribution analysis techniques, or using the flexible Weibull distribution.

The bathtub curve represents the typical life of a population of a product (Figure 2). Some products fail early due to flaws in manufacture or assembly. The surviving product samples enter a long period of useful life. Finally, the samples begin to fail and when the failure frequency accelerates, the end of useful life has been achieved. Business decisions determine how to set break-in, maintenance and warranty periods for the product.

Figure 2: Example of Bathtub Curve

Figure 2: Example of Bathtub Curve

Survival plots, cumulative failure plots and hazard plots can all be generated from this analysis to select the accepted level of risk for warranty coverage, preventive maintenance cycles and the like.

Distribution Probability Plot

The distribution probability plot establishes the distribution of the censored and uncensored data, as shown in Figure 3.

Figure 3: Probability Plot for 80 C

Figure 3: Probability Plot for 80 C

The motor winding data fits the lognormal distribution fairly well – a “fat” pen would cover all the data points. Note that many distributions cannot handle negative or zero value data points.

Survival Plot

The survival plot in Figure 4 shows how many samples are still running over the duration of the testing period.

Figure 4: Survival Plot for 80 C

Figure 4: Survival Plot for 80 C

Cumulative Failure Plot

Conversely, the cumulative failure plot in Figure 5 shows how many samples have failed over the duration of the test.

Figure 5: Cumulative Figure Plot for 80 C


Hazard Plot

The hazard plot in Figure 6 shows how the rate of failures changes over time. In this case, the rate of failures continues to increase from about 20 months up to 90 months. Between 90 and 120 months, the failure rate is stable. After 120 months, the failure rate begins to decline.

Censored data analysis predicts when failure will occur for the population based on the non-censored and censored results to date. If many censored data points have exceeded the time on the non-censored results, the cumulative failure plot will shift to the right predicting longer life for the samples that have not yet stopped. When we consider only the non-censored data points, the prediction will be a much shorter lifetime, in this scenario.

For transactional processes, such as the sales booking example, the same math works and the same interpretation applies. Today, look at the sales orders and bookings for the month and there are some orders already booked. There are a number of other bookings that have not yet turned into orders. In addition, there is a gap between orders, pending orders and the goal for the month. What needs to happen during the remainder of the month to be certain that the number of orders meet the goal?

The model created from prior analysis of the transformation of bookings to orders can provide confidence in the month’s performance and direct the attentions of sales people to the bookings most likely to become orders. Expanding the model can assure that time is properly spent between this month’s orders and future orders.

Elapsed Time and Censored Data

Survival Plot

A survival plot shows the probability that a sample will survive to a particular age. Depending on the nature of the science behind the life of the sample, the shape of the curve will vary. A survival plot on bookings would show the likelihood that a booking will result in an order at any given point in time. The data shown in Figure 7 that half of sales contacts are still pending an order after nearly 5 days, and 5 percent are still waiting to become an order at 155 days.

Figure 7: Survival Plot for Time Elapsed

Figure 7: Survival Plot for Time Elapsed

Cumulative Failure Plot

In comparison, the cumulative failure plot shows the probability that failure has occurred at a particular point in time. A cumulative failure plot on bookings would show the likelihood that a booking has become an order at any point in time. Figure 8 shows that 50 percent of contacts booked orders in about 5 days and 95 percent booked orders in 155 days. Given that the failure modes for transactional processes are events such as receiving an order or making a delivery, it is common for the cumulative failure plot to be easier and more meaningful to discuss.

Figure 8: Cumulative Failure Plot for Time Elapsed

Figure 8: Cumulative Failure Plot for Time Elapsed

Hazard Plot

Perhaps the hazard plot is the most useful in this case. The hazard to be looking for is when bookings turn into a big waiting game. At what point should a sales contact ever turning into a booked order be given up? At about 95 days, there is a point of diminishing returns. The last 8 percent of unresolved sales contacts trickle in over 608 days, as shown in Figure 9.

Figure 9: Hazard Plot for Time Elapsed

Figure 9: Hazard Plot for Time Elapsed

Why Include Censored Data Points?

Analyzing with – or without – the censored data points is dependent on a number of factors. First and foremost is the mix of censored and non-censored data points at any given point in time. As the data available on any given day approaches mostly non-censored, then the influence of the censored data is minimal. The mix of factors that we want to analyze, however, must also be proportionately represented in the non-censored data. Other factors to consider include the significance of the differences in the analysis based on the shape of the distribution. If looking at a comparison of statistics associated with the non-censored and total data set for the above sales bookings data, misdirection can occur when ignoring the censored data.

Table 3: Comparison of Data at Different Time Points
  Order Received (Days) All Data (Days)
85 Percent Booked Orders 23 49
50 Percent Booked Orders 6 5

Here is another example to demonstrate the difference between capability including, and excluding, censored data. The bane of many meeting facilitators is getting people to return from break on time. Consider participants receiving a 10-minute break. Track the time when each participant returns and at the end of the break time calculate the average break time for the group. There will be a different break performance than if those who return late from break are included. In Table 4, if looking only at uncensored data, it is possible to believe that the average break time for participants is 8.4 minutes and that the participants are 100 percent compliant. However, after 10 minutes, only 60 percent of the participants are back for the next session. Looking at the full data set, the average break time per participant is 10.1 minutes.

Table 4: Example of Including, and Excluding, Censored Data
Participant Break Duration (Minutes) Average Break Duration for Group (Minutes) Percentage Break Less Than or Equal to 10 Minutes
1 7 =7/1=7 100%
2 8 =(7+8)/2=7.5 100%
3 8 =(7+8+8)/3=7.7 100%
4 9 =32/4=8 100%
5 10 =42/5=8.4 100%
6 10 =52/6=8.6 100%
7 11 =63/7=9 =6/7=86%
8 11 =74/8=9.3 =6/8=75%
9 13 =87/9=9.7 =6/9=67%
10 14 =101/10=10.1 =6/10=60%

Example – Expense Report Accounting

An accounts payable department experienced a lot of rework to properly allocate expenses to the ledgers when books closed without expense reports in hand. Department managers found their actual expenses did not match their expectations and later changed as accounts payable chased down late expense reports and reallocated expenses to the correct accounts. The baseline data for the project showed 68 percent of expense reports closed within the required 34 days of the charge. Thirty-two percent of the expenses reports had expenses that were subject to reallocation once the reports were finally successfully submitted.

Figure 10: Completed Expense Reports – Percent Completed by Number of Days

Figure 10: Completed Expense Reports – Percent Completed by Number of Days

The table of statistics shows that 2,885 of the expense reports achieved “failure,” meaning that they were successfully closed and accounted properly. Another 203 expense reports are tagged as “censor,” which refers to expense reports that have not yet closed. These censored data points are distributed along the entire curve. Some have been open for a day or two while others have been pending for weeks.

When the expense report team investigated the various stages of the reporting process (purchase, create and submit the report, report approval, and final accounting audit) the stage with the longest cycle time was creating and submitting the report. An investigation into the demographics of the report creators (departments, positions/roles, types of expenses, etc.) showed that there were no particular culprits. Many individuals were rare submitters. An investigation into the follow-up on expense reports that had to be revised showed that the real root cause of the long submittal time was that the rules for allocating the charges and completing the expense report correctly were confusing for the purchasers – experienced and inexperienced alike. A spaghetti diagram showed that getting questions answered was a long, convoluted process.

Figure 11: Process for Submitting Expense Reports

Figure 11: Process for Submitting Expense Reports

This process was simplified by putting all of the information in the same place, accessible to all. Individuals with questions, or individuals who submitted faulty expense reports, were directed to that source.

Figure 12: Revised Process for Submitting Expense Reports

Figure 12: Revised Process for Submitting Expense Reports

The final results of this action in Figure 13 show the improvement in expense report cycle times.

Figure 13: Cumulative Failure Plot for Turnaround Time for Expense Reports

Figure 13: Cumulative Failure Plot for Turnaround Time for Expense Reports

When the cumulative failure curve shifts to the left, the speed of the operation is faster. In this case, the number in compliance with the 34-day target climbed to 81 percent, which was deemed “good enough.” Perhaps the best outcome was when a key individual in accounting complained that the managers were taking too long to approve the reports and another project was needed; the manager approval cycle time was less than 2 days, including weekends.

If the censored data points had not been considered, the accounts payable department would never have worked to solve the problem. A comparative group of only closed reports shows a mean of 18 days and 85 percent of the submitted reports complying with the 34-day specification. Including the censored data points in the analysis properly weights the tail of the distribution. In this case there was a bias associated with early failures (that is, successfully submitted reports) that pulled the data that excludes censored data into shorter times, as shown in Figure 14.

Figure 14: Summary for Submitted Reports

Figure 14: Summary for Submitted Reports

Why Lean Thinking and Shop Problem-Solving are Critical for Truck Equipment Distributors

“When you look at this word, it has come to mean many different things to people, and it also means not much to anyone,” said Sievert, senior VP of operations for Henderson Products. “Most people think of it as a manufacturing-focused set of tools. In actuality, lean is about problem-solving. Everything with lean is problem-solving. Those tools were all created to solve some sort of problem that someone has at some point along the way.”

In his presentation, “Why Lean Thinking and Shop Problem-Solving are Critical for Truck Equipment Distributors,” he said problem is often viewed as a dirty word.

“Someone comes up to you in the office and says, ‘I have a problem,’ and the first thing you think is, ‘Oh, no, now what?’ ” he said. “One of the first questions you need to ask on the lean journey is, ‘How does the word problemmake you feel?’ Problems are our chances to improve our business. Problems are the way to find out what is truly going on. I once had someone tell me that someone who is experienced has made an awful lot of mistakes in his life. If you went through your entire life without making any mistakes, you’ve have lived in a bubble in a corner and didn’t do anything. As soon as you try something, somewhere along the line you are going to make a mistake. And have a problem. You have to figure out, ‘How am I going to attack that problem?’

“The true definition of a problem is knowing where you want to be and know where we are today. The difference between those two is a gap. Any time you have a gap, you have a problem. So typically problems are reactive. We want to be here, and our quality or our delivery, or whatever it is, is down here. You have this gap.”

Jon Sievert, Henderson Products

He gave an example of a company that has a lead time of 120 days. Here are some things that go into it: sales and marketing activity; take the order; writing order, design and ordering material; fabricate the job; wait for paint; paint the job; receive accessories; final assembly; customer sign; deliver the job; fix any defects or make repairs; invoice; chase payments; and get paid.

The value items are: taking the order; fabricating the job; painting and final assembly; delivering the job; invoicing; and getting paid.

“Value is anything the customer is willing to pay for and also anything that positively affects the customer’s present situation,” he said. “Are you willing to pay for something that has a negative effect on your life? Do we find value in that? No, we don’t.”

Sievert said every single person in the audience has the same person working in their organization. His name is T.I.M. I. W.O.O.D.

“He’s just a normal-looking guy, but he’s in every single organization in the entire world,” he said. “T.I.M. is actually wanted in every one of our organization for distributing waste. We actually want to fire T.I.M.”

T.I.M. is actually an acronym for the eight deadly wastes:


“Any excess or unnecessary movement of materials or information. Examples: chasing parts; walking to the parts room; moving vehicles in and out of bays; picking up a body and moving it slowly through factory. We can never fully eliminate transportation. But that person who’s transporting things could be adding value to the process. To figure out how much transportation waste you have, go to Wal-Mart and buy a $10 pedometer, put it on your employees, and every night break down how many miles they walk. We did this in our facility in Iowa. We had the actual shop techs wearing them. One individual walked 15 miles in one day. And he’s one of our most trusted employees.”


“It can be in the form of piles of files, information, raw materials, work in process, and finished goods. It can be all the parts we have. It can be people. Inventory is an asset we need in order to run the business, but we often get too much of that asset. Inventory is like a big, cozy blanket. When we see the chassis lined up with daisies popping up in the fields, we feel nice and cozy. ‘Man, look at that backlog. Look at all that great work we’ve got to do.’ But who’s waiting on all that waste? The customer. Inventory makes us feel comfortable, but it’s a very expensive way to feel comfortable. Why is it waste? It ties up cash and space. It covers up problems.”


“Any movement or motion that does not add value. This one is tougher to see in most cases. It’s doing things I don’t necessarily need to do. I think about the surgeon. Does he get his own gauze pads? He puts his hand out and someone sticks it in his hand. If we need information, we have someone go find a computer and look it up. Every time we do that, we have motion waste. And we’re not adding value to the product.”

Information wastes.

“It’s information that is unclear, incomplete, redundant, unavailable, unnecessary, or missing. We use lots of words. We have a document of specs that’s full of words. When someone wrote those words, they had a picture in their head of what those words would mean. Someone started grabbing it and reading them and building a picture in their head through words. But is the picture in my head the same as the person who wrote it? More than likely not. Information can actually get in our way, especially when it’s unclear or missing.”


“One of the most frustrating forms of waste for employees is waiting. That involves waiting for: the copier; the scanner; a document to be retrieved; a reply to emails; information; the right tool; and equipment to become available. Look around, how many people at any one time are waiting for something? What are we not doing? Working. We take our minds off what we were working on and lose focus. Then we have to jump back to it. Every time you open up another thing to work on, the amount of work you have to manage every day continues to grow—all because I had to wait. Waiting becomes frustrating. Most employees get paid to do what they do, so nobody wants to stand around and wait.”


“Doing more than the customer wants, needs, or is willing to pay for. This one becomes part of waiting. Because we’re waiting, we over-process and open up another job. You’ve got all these half-built tasks sitting in the yard. ‘Oh. Bill’s on vacation, he started that job.’ Lots of times when we’re waiting, we want to add more value to the process. So we slow down what we do, we take a little more time to straighten up and to shrink-wrap everything real nice to make it look good for the customer. Typically, it’s because we have the time to do it—maybe more than the customer is willing to pay for. You’ve got to get back to that definition of value.”


“Doing more than is necessary to meet customer demand. This can also be described as ‘make work.’ The industry is taking off right now and people are finding a boom in what’s going on around them. In order to get more done, we feel like we need to put more into the system. You end up putting significantly more in than what comes out. We put more work in. ‘Let’s see if we can find another spot to put in another chassis.’ We keep jamming more in and no one can figure out why we aren’t getting more out at the other end. It’s because we didn’t put the processes in place. We didn’t actually solve any problems that we needed to. We just shoved more into the system. Now we’ve got more costs and people and time tied up all because we wanted to get more out of the system.”


“Any mistakes or errors in the business. Defects are around us every day. When a defect happens, what’s the first question we ask? ‘Who did that?’ As soon as we ask the who question, boom, we just shoved everybody down. Who’s going to want to stand up and say, ‘Oh, I did that’? The challenge with defects is not to run people over when you find the defects. Ask, ‘What happened? How did that happen? What are we going to do to fix it?’ ”

He said a company has variability in its system—“strategic/competitive variability”—that helps it to continue to grow the business.

“Being able to do what the customer is looking for—this kind of variability is what we want,” he said. “We want to be able to create product lines that fit together, to have innovative solutions for the customer. We have to take that picture the customer has in his head and build it for him. The #1 question I get when it comes to lean and upfitting facilities or distribution centers is, ‘What about the customer piece? How do I deal with that?’ The first question I ask is, ‘OK, Customer A wants to build it this way and Customer B wants it this way. These are two things that are very different.’

“If you break it down to simple elements, what we do every day is consistent: hydraulics, turning wrenches, running torches. So what we want to try to do with variability is to actually increase strategic variability and decrease dysfunctional variability. Dysfunctional variability is waste. If we have all that waste in our systems, how can we increase that strategic variability?”

FMEA: Failure Mode and Effect Analysis – Risk Assessment

Need for Failure Assessment

Traditionally, a lot of emphasis was placed on designing a product. Those who planned ahead of time were given a pat on the back. Then, most of the effort was positioned towards manufacturing the product. Lastly, the product was checked for any defects or areas where quality was compromised. This cycle continued for ages, until researchers came up with better ideas of Quality Management. The prophets of Quality Management advocated that there was no sense in testing for quality assurance of a product at the end of the production cycle. They believed that if the promise to maintain quality was imbibed within the planning stage and the production stage, the results would be a lot better. This school of thought has gifted the concepts of ‘Ishikawa Analysis’, ‘Kano Analysis’, ‘Total Quality management’, ‘Zero Defects’ and ‘FMEA Analysis’ to name a few.

Failure Mode and Effect Analysis

Failure Mode and Effect Analysis (FMEA), formerly known as Failure Modes and Effects Criticality Analysis (FMECA) is a methodology which suggests the meaning of the popular saying “look before you leap!” The FMEA methodology, which has a lot of application in the manufacturing sector, encourages product engineers and project managers to calculate potential functionality problems that a product may exhibit in the ‘product/process development stage’ and the associated effects of such problems. This avoids issues at later stages. We must understand that a lot of energy, time and money are associated with new product developments and improvement projects. Early involvement with the ‘cause and effect’ of all possibilities helps early recovery. It is always better that the warning bell rings early than in the later stages!!

The FMEA explores the following questions:

1. What are the chances of failure? In other words, what can go wrong?
2. What is the probability of failure or something going wrong?
3. What will be the effect of something going wrong?
4. How can the chances of failure be avoided? How early can the chances of failure be detected?

While the first four issues focus on solving the ‘problem’ part, the last issue focuses on how ‘not to have a problem’. This proactive approach to product engineering is the crux of Failure Mode and Effect Analysis.

How to detect early signals?

There is a lot of literature available on the process of Failure Mode and Effect Analysis:
• Reliability of products processes and services cannot be compromised and this is the first step in the FMEA model: to assess the reliability. It is important to identify failure (non performance) modes and its effects. The more pessimistic approach in this stage, the better it is for the design of the product. Let us consider the example of a blender. Reliability of a blender gives rise to the question of failure modes: the blender stops working, it gets short-circuited, gives out a bad odor, the functions do not work, a spare part needs replacement, or it can hurt the user! (these are just a few scenarios)
• After identification of failures is complete, it is important to point out the causes. Is it an engineering problem or a design problem? Is the problem due to poor materials and spare parts used or poor process application? Who is to be blamed and what is to be blamed for failure?
• The entire process described above must be associated with severity of the problem (very severe or not that severe!), the likelihood of occurrence of failure and stage when failure can be detected (early detection or late detection).

The above is a very brief description of how FMEA works and gives a conceptual overview. The model involves use of Work Sheets, flow diagrams and algorithms to facilitate the process of predicting failures, identifying causes of failure, reducing the severity of effects and early detection of the failure itself. All in all, FMEA is a logical culmination to the thought process of being careful and proactive in the product designing stage and in knowing what to expect!