Why Six Sigma fails in the real world


This may ruffle feathers of the Black Belts among you but I’m going to highlight an example to explain precisely why the majority of pragmatic clients aren’t interested in the upper levels that Six Sigma goes to. I’ve found through experience over the years that a client is more interested in building a continuous improvement culture involving everyone and through understandable means than knowing the equation to calculate the trajectory of Apollo13.

Call Center Astrophysics

Here’s a classic example, word for word, of a simple Call Center process improvement case through Six Sigma:

I need a help to solve this Six Sigma Case, for many it might be easy. Case: There is a Call Center process in which CTQ is Customer Response Time. If Customer Response Time is more than >60sec then its a defect. In this Case opportunity will be 1. Suppose there are 300 Calls and out of that 123 call Response time is >60 sec (This is defect) and opportunity is 1 then my Sigma value will be 1.727. This value came by this Formulae process Sigma = NORMSINV(1-((Total Defects) / (Total Opportunities))) + 1.5 and DPMO is 410000, this value came by this formulae Defects Per Million Opportunities (DPMO) = ((Total Defects) / (Total Opportunities)) * 1,000,000

Question is, Can anybody tell me from this case how can i get bell curve. How I should calculate USL and LSL value. How I will draw this Bell Curve in a excel. Waiting for all your master response.

A pragmatic response

I asked another BPM consultant on their take on the above, this was their response:

“…at the end of the day if you produce a bell curve telling me the USL and LSL for my call centre, along with the number of defects per million and a sigma value of 1.727, is this really a useful measure ?

More to the point, what can I – as a business person – do with it ?”

That last line is telling. Presented with information of this nature I would find it hard to glean anything useful in terms of process improvement to action. What’s more, turning a process into a set of calculations to eke the last drop of efficiency from it appears counterproductive.

The law of diminishing returns

There is an unwritten limit to continuous improvement that deems a point where the Law of Diminishing Returns starts to apply and no amount of incremental measurement to improve a process will yield significant return. Sadly where Six Sigma fails from a real world and pragmatic standpoint is that it doesn’t recognise that limit and will forever try to improve a process ad infinitum.

At no point in these improvement cycles and methods is there a step that says, “stop hacking the process to death and just start over from scratch”. Six Sigma becomes harder to sell and to explain the ROI when you reduce the cost efficiencies with each project to the point they become insignificant and the project more costly than the return. It becomes more about justifying the cost of running a team of Black Belts through cost savings than the efficiency of process because it improves the organization.

Starting afresh isn’t a bad thing

We need to teach organisations and Six Sigma practitioners that it’s not bad practice to throw something away entirely in order to achieve the greatest gains. We need to educate the leadership that wringing the last drop from a process out of some fierce Black Belt desire is wrong and that it’s ok to pick up the blank sheet of paper.

And we as a discipline and industry need to adopt the same approach for ourselves, that it’s ok to start over and create methods and tools that are context rich for today’s adaptive enterprise, not keep hold of the old ways because they form part of decades old tradition.

Continuous improvement is about adaptation. Given the rapid changes in business and technology around us Six Sigma is about the only thing that isn’t.


Tags: , , ,

Categories: BPM, Continuous Improvement

Author:Theo Priestley

"I had more creative ideas from Theo in 6 months than I have had in 6 years from most people." Theo Priestley is one of the most recognised independent technology industry influencers and evangelists, ranking in the Top 100 thought leaders across Virtual/ Augmented Reality, FinTech, Artificial Intelligence, Big Data, Internet of Things and future trends. Theo has written insights for Forbes, Wired, The European Magazine, Venturebeat to name a few, and has been interviewed for many online publications including the BBC on his thoughts on technology and the future. A regular paid keynote speaker and panelist at conferences and events, Theo is engaged for his forthright views and isn't afraid to challenge conventional thinking and the marketing hype surrounding the industry when presenting, never pulling punches to get the message across on how technology can be applied to improve business and the customer experience. He has also successfully organised and run TEDx and Ignite events. Highly active across social networks, he sits in the Top 1% for social media engagement on Kred and Klout and is constantly sharing articles and his analysis that he feels his audience would be interested in. Theo is also active in the startup community, mentoring within UK and US accelerators and sits on a number of advisory boards. Former VP and Chief Technology Evangelist at a Top 25 European enterprise software company with a career spanning both innovation strategy and delivery of software and business change in Financial Services, and as an independent technology industry analyst. Follow Theo on Twitter @tprstly or connect here directly for constant insights on tech and marketing trends. • Top 1% Influencer on Kred (915) • Top 1% Influencer on Klout (70+) • 12,000+ Followers on LinkedIn • 13,000+ Followers on Twitter • Recognised Top Influencer in AI, Virtual/ Augmented Reality, Fintech, IOT and Wearable Tech, Big Data and Analytics.

Subscribe to the blog

Subscribe and receive an email when new articles are published

21 Comments on “Why Six Sigma fails in the real world”

  1. March 6, 2013 at 8:15 am #


    Great blog. My own experience in Six Sigma/Lean is rather limited. It seems there is another fly in the ointment. Being able to make the improvements to a business process requires that the process can change. This process is likely locked up in an enterprise software system that was not not meant to change. These systems, including the most current version of the market leading ERP packages, simply cannot move the needle fast enough or without staggering risk and resources. So it seems that the anchor to any Six Sigma/Lean exercise is the underlying systems that trap the process.


    • DaveG
      June 2, 2013 at 7:21 am #

      Steve, constraining one’s business to what the ERP system “allows” is the height of stupidity, and I wouldn’t be surprised to hear that it happens alot.

  2. March 7, 2013 at 6:26 am #

    C’mon Theo, what is so hard about y= Fx + Fx + Fx + Fx…? 😉

    Having lived on both sides of this formula (going through Green Belt training, running a project, and being a business owner needing improvement), I feel your pain.

    I can’t wait to hear with the real Six Sigma (and even Lean) practitioners have to say about this. Here is my take.

    I would suggest the Six Sigma crowd will say you, as the business owner, don’t need to worry about the tactics of the methodology. Leave the stats work to the grown them, you focus on your business measures. First contact resolution, costs per contact, volume per rep, etc. Whatever those things are. They will do there magic and you should see improvements in your business measures. The “in process” measures they use are for them to control the project and the process. When you state a goal for your business measures (i.e., I want a target measure of “x”), the improvement effort should stop and the focus should move to maintaining that level of performance.

    But, I agree, I’ve met more than my share of Six Sigma projects that didn’t know when to stop. It seems like they want to keep going just because they can. Kind of like developing an app to help you know the best time to pee during a movie, we’ve done it, but should why? (https://itunes.apple.com/us/app/runpee.com/id450326239?mt=8).

    As for when to scrap a process and start over vs. build a new one, I have seen that evaluation occur. It is usually a business decision based on the initial process review. The process in question then moves into a DFSS (designed for six sigma) approach. But, again, I’d love to hear our Six Sigma brethren chime in here.

    • Jewell
      March 8, 2013 at 6:45 am #

      agreed! Sigma is a methodology that comes with a set of tools and the application of it should be driven by business rules NOT love for the methodology. Applied correctly, it can yield meaningful results.

  3. March 7, 2013 at 1:51 pm #

    There are a number of issues here. One is the inane “Process Sigma Metric,” fatally flawed from a statistical viewpoint and just silly from a management viewpoint. It drives no end of useless activity; data transformations, tests for normality, etc.

    I tell my Six Sigma clients to use DPMO and yield instead. It’s actually intuitive. A DPMO of 410,000 quickly translates to a 41% defect rate, or a 59% yield. For manufacturing measurements, I have them go back to a capability index like Ppk or Cpk. Using DMAIC and the quality tools in a Black Belt’s toolkit, improving these metrics is a rational exercise.

    More importantly to our call center, this response time sounds like a convenience-based metric, not a customer-based one. What customer told anyone that they didn’t mind waiting 59 seconds, but they’ll hang up after 61 seconds? What they should be concentrating on is reducing response time and ensuring that they are able to give great service once engaged with the customer. An arbitrary 60-second specification won’t help anyone, and probably does more harm than good.

    What I don’t necessarily agree with is the idea of “the law of diminishing returns.” Genichi Taguchi re-defined world class quality in 1960 when he wrote about the loss function. The new definition was “on target with reduced variation.” What he showed was that, in many cases, loss grows as you move away from the nominal value, and the only way to minimize the loss is to get on target and constantly strive to reduce variation. It can be shown mathematically that this is true wherever there is an actual nominal value and rational specifications. So there IS a point of diminishing returns, but it’s much different than most people think, and most processes I’ve seen are miles from it.

  4. March 8, 2013 at 6:41 am #

    This example is pretty simple and applies to all service industries that make customers wait (e.g., healthcare, call centers, repair, etc.).

    Use an Excel SPC product like the QI Macros to draw a control chart and a histogram of wait times with an upper spec limit (USL) of, in this case, 60 seconds. If the control chart shows the process is stable, then do root cause analysis on why wait times are longer than 60 seconds.

    The same analysis can be applied to wait times and length of stay in hospital emergency rooms, restaurants, or any service industry.

  5. March 8, 2013 at 8:43 am #

    I agree totally with the intent of the original post. Six Sigma folks create waste! What is useful is to simply say a) this is not acceptable performance and it must be improved; b) we now have baseline data, lets make a daily or weekly graph so we can watch it as we make efforts to improve; c) let’s brainstorm causes and solutions; d) lets experiment and watch the data to learn what happens; e) standardize what works; and f) experiment again! The insistence on using complex statistics when simple statistics get you to improvement quicker, easier and in a more understandable manner, is a form of waste. The people who do the work are the people who matter. They can understand a simple run chart. They can understand that it is improving or getting worse. That is all that really matters.

  6. March 8, 2013 at 7:04 pm #

    The built in response that arises whenever continuous improvement gets to the stage where it starts becoming obsessed with minutiae in order to justify its worth is the old ‘paralysis through analysis’ argument.

    Much as I wholeheartedly endorse the critical importance of developing a culture based on continuous improvement, it has to be driven by creative leadership and you don’t always find the answers to problems by looking at bell curves.

    So at some point creative leaders simply need to turn to their favourite Six Sigma black belt and say – enough – go find something useful to do – or go somewhere else.

  7. March 11, 2013 at 8:41 pm #

    Mostly agree with the views and this is currently a large part of my research it always returns to one central point of departure: Management understanding of how Six Sigma improves a business in staged deployment”

    It does not make your company a market leader, it is a tool in conjunction with others i.e. Lean management principles asking questions in terms of process optimisation and removing and reducing all forms of wastes economically and for the more seasoned business seeking ultimate ROAM incorporating DFSS and lateral thinking asking questions such as: Is the process we are seeking to improve actually the desired process?

    Great topic with many opinions as we can see….

  8. March 12, 2013 at 5:40 am #

    Theo, you may recognize my name as the Co-Creator of Six Sigma and National Best Selling Author on the subject. I mention this not to “toot my own horn,” but rather to point out I’ve been around the block a few times when it comes to business improvement through Six Sigma. I’ve been doing this for about 30 years with the likes of Jack Welch, Larry Bossidy and Bob Galving, just to mention a few. I’ve also successfully installed Six Sigma in over 50 companies with demonstrated success, as evidenced by Wall Street’s “First Call” reports — the results are real and verifiable. Simply stated, they are a matter of public record.

    With this as a back drop, my comment (and message) is quite simple. You make use of an example that was seemingly selected to “fit your conclusion;” therefore, would any other conclusion be possible. in our line of work, your use of a single case study to reason forward is quite flawed. In my opinion, your approach to “research” is like trying to make a universe from a sample size of one. How can you possibly draw any reasonable or valid conclusions from one case study?

    Owing to this, it would appear that your need to create a controversial headline to attract readers takes precedence over your need to present sound advice. Perhaps unwittingly, you inflict a small but rather grave injustice to our great nation through the presentation of such suspect conclusions, especially at a time when companies are desperately looking for ways to get to the next level. As a remedy, if I were your doctor, I would strongly recommend you take about 10mg of humility chased by 20mg of scientific reasoning; only to be followed by a 50 mg does of reality before bedtime.

    Following this regiment of prescriptions for a short while will definitely help, especially in cases such as yours. The effect of this medicine is amazing. Within a few days, you will wanting to roll-up your shirt-sleeves and show us how its done.

    Our hopes and best wishes are with you.

    Mikel J. Harry, Ph.D.

    • March 12, 2013 at 5:56 am #

      Hi Mikel

      Thanks for your comment. No
      medicine needed, I took the red pill a long time ago.

      As for case studies, research, conclusions, grave injustice ? It’s a blog post, citing an example, quoting a response and a dollop of 15 years of observations doing change projects in the UK.

      Strangely enough this is an amalgamation of a couple of posts from as far back as 2009/2010, I don’t think I heard your voice back then to counter.

      I think you’ve taken this just a little too seriously.

      Take the red pill my friend and enjoy life a little more.


  9. March 13, 2013 at 5:41 am #

    Dr. Harry, we can hear your horn all over the world. You’re an opportunist and self-promoter sui generis, considering that Six Sigma is just TPS dumbed down for American executives, who set the bar for evil incompetence.

    To all the above, why the bleep didn’t anyone ask why the call center exists?

  10. Donald Smith
    March 13, 2013 at 6:36 am #

    Great article Theo. I would like take your argument a little further by saying most of our “continuous improvement” efforts, with or without Six Sigma, follow the same relevance path you described and we would be better off if we pursued what I would like to call “continuous relevance”.

  11. Jim G. George
    March 13, 2013 at 11:47 am #

    You can hack a process to death. Too many projects are only targeting small improvements and failing to ask more basic questions. Too often the Six-Sigma groups are just routinely walking through the Six-Sigma process. I worked with one group that was so formal in DMAIC that when they couldn’t get measures for a badly broken process they just got stuck unable to go the the Analysis step. Zero ever got done because executives refused to let them go on without numbers and Finance was told not to give them any.

  12. Tony Tomeski
    March 14, 2013 at 5:37 am #

    Your scenario rings true particularly in organizations that are just beginning their Sigma jouney. I found that the process takes on a life of its own until the bell of reality strikes and the the sigma teams realize they are in place to solve business problems. We combatted this issue head on by placing a review committee in place to determine the viability of the improvement project and to specifically determine is it better to start all over. We also force each lead to articulate the value of the endevour specific to the business problem expressed in a way that both the sponsor and clients will understand. In this way we minimize the projects launched simply for the sake of improvement and force the team to understand the real reason why we are here – to increase business value.

  13. March 14, 2013 at 6:26 am #

    Theo’s points are very valid. We tend to assume that every six sigma project must target a sigma level of six. The coaches need to guide the Black Belts there.

    My very first project was similar – over a decade ago. My Order-To-Delivery cycle time for putting up cryogenic installations was 6 months; our aim was to reduce it to 3 months. I was coached and guided by Professor A.K.Chowdhury of Indian Statistical Institute, Bangalore. I understood that the way we define defects can actually change the sigma levels. We just simplified the defect definition and stuck to it. My guide set the target of only 3 sigma for me, beyond which we did not expect to get any returns. We had no LSL and we set the USL at 3 months. We did a bell curve though we were aware that cycle times normally follow a log normal distribution. Later we did away with doing any curve at all. We just tracked defect counts and monitored cycle times. The topline and bottomline benefits started flowing in as we speeded up the process. Primary tools were Fishbone, FMEA, control chart, visual management. We did use ANOVA for understanding baseline and process dynamics and we utilised hypothesis testing for proving the effectiveness of our improvements early in the control phase. Later on, I took up a dozen projects but only 5 – 6 were good. I think we need not sack six sigma. Learn it and apply it in situations that have lot of input factors, data is available at a little cost or effort, quality is the primary metric, etc. You must permit guesstimates where there is no hard date – normally six sigma does not allow that.

    On Continuous Improvement (CI) programs which primarily depend on lean, the plus point is we take one small step at a time. It drives consensus, builds culture and manages change slowly but surely. Cycle time reduction projects should ideally use lean rather than six sigma. We cannot get quarter on quarter $ savings with lean. Six sigma will give that quick return but might allow a fall back as quickly to the old regime if we do not put effort to sustain it. When the technical environment or business situation changes, we have to again redo the six sigma project. However, CI programs focus on culture and thinking more than method and tools; it can help manage long term changes more easily. There are pros and cons. We must learn to make the most of six sigma, CI (Lean), TOC, TRIZ and every other method available in the right manner for a given situation. Do not reject anything in total and do not be attached to any one thing as well.

    Thanks and good luck.

  14. fallatah
    March 15, 2013 at 4:25 am #

    six sigma have tools help to defined the critical few from trivial meany, so it will guide you where to improve where you need to concentrate your effort, so it will help in saving and prevent from consuming energy and money. six sigma practitioner should be certified other wise not reliable.

  15. March 21, 2013 at 9:37 am #

    This example and most of the concurring posts highlight a failure in leadership literacy when it comes to applying Six Sigma, Lean, or any other continuous improvement effort. The key point is, before you begin any action, you need to determine if there is a business case to be made for the goal you intend to achieve. Every project or action that is undertaken needs to be linked to a business benefit, and ultimately to a customer benefit as they are the ones who pay. This will serve as a self correcting mechanism so you don’t spend more on the effort than you get back in return. There is a shared responsibility between the business leader and the Six Sigma practitioner to communicate clearly with each other around this cost – benefit evaluation before significant resources are committed. In the first example – the six sigma practitioner was articulating the problem in terms that made it hard to relate to the business. That didn’t change the underlying problem of slow response time or dropped calls which could lead to lost customers, lost revenue, and lost profit. If the business leader doesn’t understand how the data being presented to him relates to the business problem – then he shouldn’t be making decisions to spend resources based on that information. The solution lies in the six sigma practitioner becoming better at translating the measures into meaningful business terms, and the leaders becoming more knowledgeable in basic Six Sigma and Lean concepts and questioning the data if it doesn’t relate to the problem at hand, or just doesn’t make sense. In my experience with Lean Six Sigma, The business literacy of the practitioners – and the basic Lean Six Sigma literacy of the leaders, have been key leverage points to improving the outcomes of continuous improvement efforts.

    An example of this cost-benefit self regulation can be seen in the common experience of air travel. When we get on an airplane, we expect to be able to take off and land safely at our destination every single time. In fact in Six Sigma terms, we expect this to happen at a higher than 6 sigma rate, and thankfully it does. We are willing to pay for this level of safety and if an airline performed at a significantly lower level they would lose customers and business. On the other hand, when we check our bags, we expect that they will arrive with us at our destination MOST of the time. if they arrive with us at least 95% of the time(about 3 sigma) – we will probably accept that rather than have to pay significantly more in ticket price to achieve the same performance level as safe take off and landings. It doesn’t mean that we don’t want to improve the rate to higher than 95% – it just means that what the airline is willing to spend on the effort will be constrained by what the customer is willing to pay for the result. In the end there is a balance between what the customer will pay for and the level of service he expects. If you choose as a business leader to ignore this point – you probably won’t be happy with the results of the improvement effort as it will only produce losses rather than profit.

    While I stated earlier that their is a shared responsibility between the practitioner and the business leader to communicate clearly on cost benefit with each other – I believe the Leader owns the larger share of this responsibility as the decision maker. The leader needs to demand that practitioners present data and information in a way that clearly relates to the business issue at hand, and provide for the necessary training for those who have a deficiency in this area. He then needs to demonstrate that he makes decisions based on facts and data. By doing this, the six sigma efforts in the company will consistently contribute to improving the business performance of the firm.

  16. August 12, 2013 at 7:23 am #

    As co-creator of Six Sigma and national best selling author on the subject, let me assure you there is no point of diminishing return on quality improvement — continuous or otherwise. This idea is based on a white paper Sperry Corp. published in the 1950’s. Essentially, it shows the math that as the defect rate declines, the cost of those defects increases. One problem with this logic — its based on the assumption that lowering the defect rate relies on the use of detect-and-fix methods. However, if you run the numbers assuming that the defects are preventable through such devices as robust product design, the saddle point of optimality also declines. Thus, there is no point of diminishing returns. This point was well studied at Motorola in the early to mid 80’s.

    Mikel J. Harry, Ph.D.


  1. Why Six Sigma fails in the real world | fred zimny's serve4impact - March 17, 2013

    […] See on successfulworkplace.com […]

  2. Six Sigma ‘killed’ innovation in 3M | World Of Innovations - March 31, 2013

    […] Why Six Sigma fails in the real world (successfulworkplace.com) […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: