Featured Video
This Week in Quality Digest Live
Six Sigma Features
Donald J. Wheeler
You can still put an upper bound on measurement error
Kyle Pheland
Solving problems and forging better partnerships
Belinda Jones
Solving problems and forging better partnerships
Donald J. Wheeler
The art of using a consistency chart
Davis Balestracci
Answer: none of them

More Features

Six Sigma News
Nov. 30, 2016, in Copenhagen
A story about how organizations rise and fall—and can rise again
Quality Essentials includes downloadable tools and resources, videos of Juran, and his Quality Handbook
Company headquarters and 30 jobs in Dayton, operations in Europe, stay in place
Same great content in mobile-friendly format
Enhanced to dramatically improve workflow with unprecedented speed and ease
New e-modules and blended e-learning curriculum help academic institutions reduce waste, improve processes

More News








Davis Balestracci

Six Sigma

Can We Please Stop the Guru Wars?

The universal road map for improvement

Published: Tuesday, February 11, 2014 - 09:19

The various improvement approaches are, in essence, all pretty much the same. Any competent practitioner would neither want to be called a “guru” nor have any problems dealing with another competent practitioner of another improvement philosophy.

In my opinion, any approach should also involve the use of data in some way, shape, or form. I once had a lean sensei (local “guru”?) vehemently make the point that lean does not involve data at all. I'll let you decide. Would you rather be effective or right?

Unfortunately, data conjure up the dreaded word “statistics”... and a huge misconception. The influence of W. Edwards Deming has created the mistaken belief that mass training in statistics will cure all quality ills. We all know the monster that has created.

Not ‘statistics’ but ‘variation’

I believe that one of Deming’s most profound statements was: “If I had to reduce my message to management to just a few words, I’d say it all has to do with reducing variation.” Reduced variation ultimately yields a more predictable process. But let me first expand your conception of “variation.”

Taking the “respect the process” theme of my last column, let’s look at it from an even larger perspective: How deeply has the concept of “process” permeated the very fabric of your organization?

In any organization, there is no lack of raw material needed for improvement: enthusiasm, focused attention on key issues, and good, hard-working people with good ideas. They may be necessary, but they are not sufficient. They need the catalyst of critical thinking inherent in process-oriented thinking.

The universal road map for improvement

The original version of The TEAM Handbook (Oriel, 2003) was published in 1988. Its process orientation provides the most robust framework within which to consider any improvement process—summarized beautifully through the elegant and logical road map provided by its “Sources of Problems with a Process” (now updated from six to seven). Here are the first three:
1. Inadequate knowledge of customer needs [added in the 2003 edition]
2. Inadequate knowledge of how a process does work:
• Variation in people’s perceptions of how things currently work
• Variation in training delivery or comprehension
• Individual preferences (variation) evolving from undocumented reactions to unforeseen inputs or situations
3. Inadequate knowledge of how a process should work:
• Variation in people’s perceptions of how things should work
• Given the current process state or objective: poor process design
• Naive or obsolete design
• Poor training
• Expansion beyond initial needs and capacity

Key question: What is actually happening vs. what should be happening?

Let’s consider this “gap” as variation: What we know vs. what we don’t know.

Here is a brilliant observation by Henry Neave in The Deming Dimension (SPC Press, 1990):

“[S]ystems are unlikely to be well-defined in practice unless they are both suitable and adequate for the jobs for which they are intended, and are written down in a way comprehensible to all involved.... [T]here can be big differences between what is written down—the way the system is intended, or thought, to operate—and what actually happens. Incidentally, the acquisition of this latter information can be difficult if not impossible in an unfriendly work environment. But, even in a constructive environment, it may still turn out to be almost impossibly difficult to complete a flowchart or other description because what is going on is so unsystematic. To be blunt, if a system cannot be written down, it probably doesn't exist: That is to say it probably functions more on the basis of whim and ‘gut feel’ rather than on any definable procedure. This surely implies that the variation being generated is some scale of magnitude higher than really necessary, with the resulting (by now) well-known effect on quality.”

Human variation is a huge factor in these first three steps. Documentation—i.e., flowcharting—reduces variation in perception of the situation. But even with this variation reduced, as the project moves forward, the additional problems of human variation still remain in both the improvement process itself and any subsequent data collection.

So, one could say:

Reduced human variation = higher quality improvement process = higher quality data collection.

It has been my experience that, the closer one gets to the frontline, the more difficult it is for those workers to see the bigger organizational picture. They are working very hard—and make sure you know it. And, to them, that’s quality. I’ve seen the pride as many tell me how they, in essence, treat every interaction in their job (whether customer, patient, department, or work colleague) as a special cause. They have minimal awareness that interacting with an internal or external customer is a process yielding “predictable”—and to them, many times frustratingly chaotic—results. No problem: Many have also proudly developed their own, individualized workarounds (read: special cause reactions) when these “unexpected” things occur. These add compexity in the form of human variation but no value.

Questions to ask about a process

In almost any project, as you inevitably deal with frontline processes, you will either have to get workers to see their job(s) as a process or talk to them so that you can then understand their jobs as processes. If the latter, you can educate them to that fact as a project progresses and solutions are implemented. I might suggest using the following script (developed by Joiner Associates and contained in The TEAM Handbook) at the beginning of a project, which would yield mutual benefit:

The process being studied is (fill in the blank).
1. Who are its external customers? What individuals, groups, or systems outside our organization rely on, or could benefit from, this process? Who has (or could have) expectations about this process?
2. How do we know what the external customers like or don’t like about this process? What satisfies or dissatisfies them?
3. Who are its internal customers? Describe those within the organization who do (or could) rely on the successful operation of this process or the resulting product or service.
4. How do we determine what the internal customers like or don’t like about this process? What satisfies or dissatisfies them?
5. What are the operational definitions of quality in this process? What specifically determines whether the process is working well or poorly?
6. What records are kept regarding quality? Who uses this information? How do they use it? Are these record formats suited to how they are used?
7. What are the most common mistakes or defects that occur? What is the operational definition for each mistake or defect? What proportion of these is commonly assumed to be a worker’s fault? What proportion do we usually attribute to the system? How do people arrive at these conclusions?
8. By what process do we inspect, evaluate, and report problems regarding:
• Planning required for this process
• Incoming materials, supplies, and information critical to this process
• The process itself
• The final product or service received by the external or internal customer
9. List the critical elements of this process: materials, components, parts, information, etc.
10. List the suppliers or vendors of each critical element.
11. Describe the company’s procedures for purchasing materials or services brought in from outside the facility. To what extent is “low bid” a governing factor in our purchasing decisions?
12. Describe the impact of the most common mistakes or defects in this process. What do they cost in time, money, customer loyalty, or worker pride?
13. Who is responsible for quality in this process? Who is responsible for detecting mistakes or defects? Who is responsible for identifying and correcting the causes of mistakes or defects?

I can’t think of a better experience for educating a culture in probably the most important concept to grasp in improvement: process-oriented thinking.

By the way, the above is the “plan” of any initial PDSA.

Challenge: How is human variation compromising your improvement efforts? Have you ever thought about your organizational quality education and training as a process? How about your overall “improvement” process? Do Henry Neave’s comments apply?

If you have any “aha!”s, please share them with me.


About The Author

Davis Balestracci’s picture

Davis Balestracci

Davis Balestracci is a past chair of ASQ’s statistics division. He has synthesized W. Edwards Deming’s philosophy as Deming intended—as an approach to leadership—in the second edition of Data Sanity (Medical Group Management Association, 2015), with a foreword by Donald Berwick, M.D. Shipped free or as an ebook, Data Sanity offers a new way of thinking using a common organizational language based in process and understanding variation (data sanity), applied to everyday data and management. It also integrates Balestracci’s 20 years of studying organizational psychology into an “improvement as built in” approach as opposed to most current “quality as bolt-on” programs. Balestracci would love to wake up your conferences with his dynamic style and entertaining insights into the places where process, statistics, organizational culture, and quality meet.


Toyota and measurement

At the 2014 Lean Transformation Summit, there’s a presentation about Toyota (the TSSC) helping the Food Bank for NYC.

They all made it clear that data was important… looking at measures was a key way of figuring out if you have improved.

But, of course, it wasn’t all about measures… it was also about provide the right service in the most dignified way.

The guy who said Toyota/Lean isn't about data is completely wrong and uninformed.

let's cross our fingers ...

... and only hope that all this guruing doesn't turn into gu-ruining.

Reducing variation goes against human nature...

A couple of thoughts struck me while reading your article. First is a quote by Eli Goldratt that said, "Without Data, I'm just another opinion"...

The next thing that came to mind is that it is funny that as quality improvement professionals we talk about identifying the sources of variability and the reduction and elimination of unwanted variation and so on... But we don't live what we preach duing working hours.

Because who of us has the same thing for lunch every day? Who wants to watch the same movie every night and so on. As a culture we thrive on variation. If there was no variation then sporting events would not be as popular, we would not try new restaurants, we would not seek out and try new recipes. Think about it. In almost everything else in our lives we not only desire variaion, we thrive on it. And if no one wanted variation then many of the worlds leading industries would be out of business...

It's almost an oximoranic life we lead, how do you provide consistent and repeatable variety??? Chew on this... In order to provide a living for us and our families so we are able to enjoy the varities and spice of life, we must work to identify the sources of unwanted variation and eliminate it....

Its a mad, mad world...

thenson the quality geek. I even straighten the rugs in the hallway of our offices when they are not centered and parallel to the wall... the saga of the unwanted rug placement variation...

Enjoy this crazy profession.

Another quote from Henry Neave

Thank you for commenting. Your note reminded me of another quote from Henry Neave:

"Note the difference between 'variation' and 'variety'. Variety is the 'spice of life'; variation is not. Variety in available products and services can of course enrich life. In contrast, variation is nasty: it makes life difficult, unpredictable, untrustworthy. It is 'bad quality'. 'Good quality' implies reliability, trustworthiness, no nasty surprises: i.e. little variation."

This is taken from his outstanding article "Understanding Variation," which can be downloaded free at:



Not all...

Not all variation is bad!

Variation in what we eat every day as individuals... if that's what we want to do... probably doesn't affect a customer.

A hospital cafeteria's customers WANT different food each day (specials), but also sometimes want standard items that are always available. It's the customer need that drives this. Lean and Six Sigma would both tell us to listen to the voice of the customer.

Variation is only bad if it hurts quality or safety, increases cost, hurts financial performance, etc.

Straightening the rugs is just borderline OCD :-)

Variation is too benign a word

I call it "deviation" because it deviates from the customer's desired target value.

Nobody likes "deviant" behavior, but we all seem to tolerate variation as an acceptable norm.

In healthcare, I often hear clinicians argue that there's too much variation among patients to follow a medical cookbook.

But as the IHI has found, checklists save lives and lots of them.

In Don Berwick's new book, Promising Care, he argues that there is too much variation in standards of care. One hospital is good at preventing surgical complications and another is good at preventing ventilator-assisted pneumonia.

But according to many estimates, hospitals still kill 400,000 people a year unnecessarily making it the third leading cause of death in the U.S.

If there's any place that we need to eliminate deviation it is in healthcare.

Maybe if we stop calling it "variation" and start calling "deviation" it would get more traction.

How to decide what's best?

To make it more confusing, some in healthcare promote an approach called "positive deviance" -- good in principle, but a TERRIBLE name. Who wants to be a "positive deviant"?

It's worse branding than the term "Lean."


One challenge in healthare is deciding WHAT is the best treatment. Dr. Brent James from Intermountain Healthcare has said that there's really only best evidence about WHAT treatment protocol to follow for about 30% of medicine.

If we agree that "evidence-based best practices" are a good thing, it's hard to decide "best" without evidence.

And then there is the challenge of variation in patients... do we want the doctors to slavishly and thoughtlessly follow a protocol that works "on average" or "for most patients"??

There are cases, thankfully, where more standardization (without being completely identical) is helping improve clinical decision making (what do we do?) AND medical processes (how do we do it?).

I don't want overly rigid "cookbook medicine" if it's bad for me as a patient. But that's not excuse to not make things more standardized where it's helpful and beneficial.

Lean and Data

I once had a lean sensei (local “guru”?) vehemently make the point that lean does not involve data at all. 

Wow, if I ever heard somebody say that, my reaction would be to think, "That person got really bad Lean training." The person you're writing about either isn't any sort of sensei or guru, yet alone "competent," unless there's some context missing or some misunderstanding.

Toyota has long placed an emphasis on "facts" (things that are observable) and data and measures... a problem is a measurable gap in performance (not just a feeling or an opinion)... and we use data to see if we've closed the gap.

Now that I think about it, are you thinking of the old Taaichi Ohno quote about data and facts?

"Data is, of course, important in manufacturing," Ohno often remarked, "but I place greatest emphasis on facts.".

The person you're thinking of might have misunderstood Ohno. He says pretty clearly that data is important, not irrelevant. What Ohno was warning against was relying ONLY on data (such as reports and spreadsheets) instead of confirming things with your own eyes and your own questions.


Nice comments, Mark...

...and I agree with you. Maybe through the critical thinking of good design, clear objectives, and solid operational definitions, "facts" can now be converted to meaningful data...provided the "human variation" of the collection process doesn't contaminate things!

As to your opening comment, there was neither missing context, nor misunderstanding, and she is not unique in my experience. I addressed a Lean conference where my BASIC message about variation, data collection, and merely plotting data over time was met with blank stares and defensiveness.

I'm afraid we're seeing another manifestation of the "human variation" inherent in Deming funnel rule #4 (make each one like the last one) on the wonderful original theory as developed by Ohno and others. In fact, as suggested by Mark Hamel, many Lean processes have become muda because of misunderstandings of its basic theory and the ineffective ways in which it is implemented.

My guess is that you, me, and Mark Hamel could sit down, have a meaningful dialogue, and become good buddies -- and I would welcome the opportunity if it presented itself!


I'm still confused


It would be good to chat. I'm not taking issue what you're saying. 

I'm still just dumbfounded about how a "lean guru" could say Lean isn't about data... how would they ever evaluate if anything is better? Based on feelings? I don't understand where that person was coming from. Did you have more dialogue with them that you can use to elaborate on their comment?