iSixSigma

Every six sigma project MUST reach 3.4 DPMO?

Six Sigma – iSixSigma Forums Old Forums General Every six sigma project MUST reach 3.4 DPMO?

Viewing 22 posts - 1 through 22 (of 22 total)
  • Author
    Posts
  • #28049

    Saul Garcia
    Member

    I have been discussed with some partners about the goal of a six sigma project. Some of them says that the goal of every succesfully six sigma project in a process MUST BE reach the six sigma level (3.4 DPMO at least). It´s that correct?
    Thanks for your opinions
    Saul Garcia P.

    0
    #69472

    Mike F
    Participant

    It may be technically correct to reach 3.4 DPMO in order to qualify as a successful six sigma project, but at what cost? At some point, you have to inject business requirements (customer needs) and reasonable resources (money and manpower) into the equation to see if 3.4 makes business sense (as opposed to technical sense) and that’ll determine if it’s worth it.

    0
    #69474

    Murray
    Participant

    Nothing like an answer from a guy who does not know.The answer is the six sigma process is a learning process. We seek to close the gap between our reality and what is possible by at least 50% every time we initiate a project. The only ones who will contend that every project has to reach 6 sigma to be successful are ones who obviously have never tried this.As far as the contention of business reality and all of the other nonsense written — this is a learning process, we learn our way to six sigma over time. Your alledged reality of not being worth it is only todays reality. As we reduce defects and time in ways that are important to the customer, the sigma level that is a achievable and of good business benefit increases as well. There is no known answer to how good of quality and responsiveness a company can afford over time, there is only what is our next best opportunity to improve.

    0
    #69475

    Kim Niles
    Participant

    Dear Allen:
    I like your answer and since you do seem “experienced”, I’m wondering if you care to extend the confinds of your definition a bit to include what “we” really include in our measurement.  For example, our project DPMO measurement really consists of some CTQ/CTC/CTD scorecard of sorts which seems to vary from company to company and project to project.  Some include ROI/COQ or other Hidden Factory measurements/Gains in Market Share, etc. 
    What have you or anyone else reading this seen that can help me with implementation at my company in this regard?
    KN 

    0
    #69477

    Murray
    Participant

    Kim,Tie projects to measureable company objectives. Every project chartered is a 50% reduction of the specific problem of interest. Finish one project and go get the next best opportunity to the best of your organizations ability to define such things.

    0
    #69479

    Jonas Williamsson
    Participant

    With the BB projects you make sure that you fulfill the short term goals for the organisation. The short term goals make sure that you reach the long term goal of having Six Sigma in your processes and on your products…
    Jonas

    0
    #69488

    john beaudoin
    Participant

    The answer to the original question posed would require a rather lengthy response, which the sum of the smaller responses may do well in answering the question.  (With the exception of the 1st response).
    Everyone has touched on some good points and objectives.  I also assume that everyone is using the same definition of what the big picture of “6-Sigma” is.  For example, it is very obvious that for a successful 6-Sigma project that is at a 6-Sigma level, you do need to be not only at 3.4DPMO, you need to sustain the 3.4DPMO limit over time.  However, you can also have a successful 6-Sigma project that takes you to a 4.2-Sigma level.  Depending on your original COPQ and Sigma Level (Could be 2 even), you can have very successful projects that are a 3-Sigma level.
    No project can be called a success unless whatever gains achieved are only temporary.  Many companies are using the sustainability and the $ amount saved to determine truely successful projects.  Because companies can exist from very large to very small, $ level may not be good by itself to use.  For example, a good project for us is saving a company over $100K in a year.  The figure could be scaled as a % of all operating expenses.
    The point I am trying to make is that when you implement, your largest savings will come first from your worst processes.  Each year as you improve, your savings per project may shrink.  You will get much more savings going from 3 Sigma to 4 Sigma than you will going from 5 Sigma to 6.  Which of these two projects would you define as more successful?

    0
    #69494

    walden
    Participant

    I have found that whilst aiming to reach 3.4 DPMO against customer CTQ’s, the actual CTQ effectively changes over time as customers become more sophisticated in their requirements. So for example you could be hitting 4.5 sigma., but when the CTQ changes, you could go back down to 3 sigma.
    6 Sigma really can be a moving target depending on the CTQ. Also it is said that to hit 6 sigma you need to remove humans from the process as most people based process have process entitlement of around 4.5 sigma.
     

    0
    #69495

    Marcel Power
    Participant

    Saul,
    Everything’s relative; if your baseline is at 2-sigma and you reach 4-sigma, that’s a success. I’ll leave you with 2 axioms:
    1. The important thing to remember is that you sustain the gains and standardize them, the C in DMAIC, the A in PDCA, etc.
    2. Never stop improving; do not let-up in the (your) relentless pursuit of eradicating (to tear up by the roots) Muda, variation, etc, etc, etc…
    In the words of George Eckes “6-sigma is a management philosophy”; ie, how you embrace 6-sigma will determine your outcome.
    Regards, Marcel

    0
    #69500

    Stephen Curtis
    Member

    In general there is no need to set an absolute target of 3.4dpmo which in many cases is an aspirational goal.
    Each six sigma project should have an improvement target as a measure of its success.  We may set a project goal of 80% reduction in defects, a 1 sigma shift, or 4.5 sigma (LT).  
    In manufacturing, a process yield target doesn’t change too often (ie: the customer requirements may be static) however in the service and transactional environments customer needs change quite often and the payback in achieving what may be a costly effort to sustain may be low.

    0
    #69512

    I Allen
    Participant

    You said –The point I am trying to make is that when you implement, your largest savings will come first from your worst processes. Each year as you improve, your savings per project may shrink. You will get much more savings going from 3 Sigma to 4 Sigma than you will going from 5 Sigma to 6. Which of these two projects would you define as more successful?I think this is just an opinion, show me your data if you even have experience in the 5+ sigma range. I do.

    0
    #69513

    john beaudoin
    Participant

    Because Sigma Level is not a linear function of errors this is mathmatical fact.  If each defect has a fixed COPQ, say $10 for this example, the following holds true mathmatically:
    At 2 Sigma, you have 308,537 errors per million.  At 3 Sigma, you have 66,807 errors per million.  In moving from 2 to 3 Sigma, you have saved 241,730 errors.  At $10/error, you would have $2,417,300/million units in savings.
    If you look at the same example, but different sigma levels you get this: At 4 Sigma, you have 6,210 errors per million.  At 5 Sigma, you have 233 errors per million.  This means your project will save 5,977 errors or $59,770 per million units in savings.
    Again, I ask….Which project is more successful?

    0
    #69514

    Wäsbom
    Participant

    The last one, because relevant savings for this comparison are CUMULATIVE savings. If you are standing at 4 Sigma level you just saved $ in the passing from 2 to 3 or 3 to 4. Do not forget they. Take its in account. Follow the road of the continuos improvement…Good luck,
    Diego

    0
    #69518

    Murray
    Participant

    You know what it is like to work in a 5+ sigma environment? If you don’t, you have now idea of the value. What is the value, for example, of Toyota having the most popular car in the US with huge margins?Defects are not your only cost. Warranty, repeat customers, market domination, higher margins — where do they fit into your equation?Again I ask — where is your data? In other words, have you actually ever experienced a 5+ sigma environment? You are making statements from your imagination, not from your experience.

    0
    #69519

    Murray
    Participant

    Who told you that about people? Totally untrue. You think Toyota build cars with robotics? They are in very limited use (mainly welding of the body)

    0
    #69526

    Ricardo Banuelas
    Member

    The answer to this question in most of the cases is no.
    The reason behind this is that we need to take into account the current sigma level of the organisation and the current level of critical to quality characteristic that the project is trying to improve. Bearing this in mind a successful project can reach levels lower than 6 sigma. It is important to select projects and goals based on the impact that they have in the organisation and customers.
    Theory of Constrains (TOC) could help to organisations when project targets are selected. Applying TOC projects can be improved until they will not constrain to the organisation in the improvement process. It is important to stress that the ultimate goal is to achieve Six Sigma.

    0
    #69527

    john beaudoin
    Participant

    I work for Sony.  In our manufacturing environment, I would assure you that we are 5+ sigma.  Are products have a strong reputation for quality and our name is recognized more world wide than any other brand name.  The fact that your company is also producing quality does not aid the original person asking the question, as their company may not be in the position to discuss 5 sigma yet.  I would also bet that as a Transactional Greenbelt, that your company is not anywhere near 5+ sigma level at the transactional level.  If it is, show me your Data.  GE pioneered transactional six sigma, and like our company, found that there were not many measurements in place to even know what sigma level you were at in terms of keying data, photocopying paperwork, phone calls, filing, etc.
    I am not arguing the point that there can be substantial savings going from 5.1 Sigma to 5.5 Sigma, because you can save there.  Also, if you are a 5.1 Sigma company, that is the type of project you are going to look at for continuous improvement.  However, I do not think it makes sense to tell a person who is in a 3 Sigma company (near the national average) that the only good project is one that takes them all the way to 5+ Sigma, because that simply isn’t true.
    In evaluating projects, you also have to determine the length of time and amount of effort and resources it will take to complete the project.  You must weigh the potential savings vs. investment.  If you can knock out a $20,000 project in 1 month, this may be a better project than a $100,000 project that will take a year and a $30,000 investment in software or equipment.  From a continuous improvement standpoint, you will want to eventually do both projects, but it probably makes more sense to do the project with the quickest gain first.
    Someone also made the good point earlier about a company having to evaluate what Sigma level is worth to them based on customer expectations and resources.  For example, if you would need to invest $100,000 to go from 5 errors per million to 3 errors per million, would that project be worth more than a project you could invest $100,000 into and go from 2,000 errors per million to 500 errors per million.  FMEA has a role here, but assuming risk is the same, your best project will be in not achieving 6 sigma with one process, but achieving 5+ sigma in another.

    0
    #69543

    Mike Carnell
    Participant

    Just as most others have said – the answer is no.
    If yo look at the original goals we had at Motorola the goal was a 10 fold improvement evert 2 years and 100 fold every 4 years. This is a 68% reduction in defects each year.
    Common sense should prevail in all situations. The people telling you they all need to reach 3.4 ppm are either completely inexperienced in deploying anything particularly SS or else they are just stupid. If you are working on a 1 sigma process, driving it to six sigma probably does not make sense unless the rest of your processes are at that level. The other side of the argument is that we frequently see (as consultants) a person doing a DOE taking the best treatment combination as the best solution. A simple EVOP or RSM would get them optimized.
    You should expect to plateau around 5.0-5.2 sigma. It generally takes some type of DFSS activity to break through this level.
    You should also watch your metrics. DPMO should be used as a management metric. The opportunity count is the complexity factor which lets you compare dissimilar processes. At the project level you should be running on DPU. It is much easier for your team to understand and you will not get involved in some stupid discussion about what is an opportunity.
    The guy that thinks GE pioneered SS for transactions is incorrect.

    0
    #69578

    john beaudoin
    Participant

    Good Response,
    In the training I have received, I have been told that GE built upon what Motorola had accomplished by converting Six Sigma to transactional processes.  Some of the largest profits for GE come from their financing and credit areas, which are heavily transactional.  As a result of their focus in this area, they have been more successful transactionally than any other company that I am aware of.  I am currious to know who the real pioneer of Transactional Greenbelts and Blackbelts is?

    0
    #69596

    Mike Carnell
    Participant

    If you look at the documentation from the 86-88 time frame when Motorola launched SS the directive was that we would be SS at everything we do. There was specific language that it was not indigenous to manufacturing. If you look at a history of the TCS competition (Total Customer Satisfaction) you will see a lot of transaction processes. A prime example is closing the company books. Look at how quickly Motorola reports quartely earnings (to bad the haven’t reported good numbers lately) – direct relationship.
    You need to remember Motorola never did Six Sigma to popularize it. It was just how we did business so there was never much celebration around it. The GE people definately took it to a new level. Although Bob Galvin was a great leader and will some day be recognized for his contributions to Manufacturing and Business, he has never been as high a profile as Jack Welch.
    GE Capital is such a large part of the GE picture that the number of projects and opportunities around transactions is a complete order of magnitude different than Motorola. There are many things which have made GE great such as focus, intelligence and resources. They poured all three into the Transaction process. Simple Pareto principle. If you look at who contributes to GE profitability it is _____. Transaction was their leverage variable. It only made sense. Even in an area which doesn’t appear to be Transaction driven such as Aircraft Engines it has a huge impact. Ken Meyer, drove the AE deployment and insisted on transaction projects from the first wave.
    It is like any other piece of SS. Nobody has done it all. Motorola didn’t ignore it. GE certainly did shine it up a great deal.

    0
    #69622

    Ted S
    Member

    I’ve read with interest the responses that were posted to this simple question and was amazed in all the different opinions as to the merits of reaching 3.4 DPMO.  I am also astonished that everyone seems to have processes that have a true entitlement at that level.
    To answer the question, you can improve a process using DMAIC only to the entitlement of the process.  If that exceeds 3.4 DPMO you will never be sucessful if that is the only target you can invision. 
    A better question would be – What is the capability of the process, can it meet the customer expectation at that level, given the cost it would take to improve to entitlement and how does that level match against the competion.  6 Sigma and management of projects are business issues and the application of the limited resources has to be applied where the greatest reward can be achieved. 

    0
    #69804

    walden
    Participant

    Who is talking about Toyota – I said that most human based process entitlement is around 4.5 – 5 Sigma. In non-manufacturing sectors this is a widely understood fact. If you want to drive processes to 6 sigma then you will most likely have to look at signficant process automation.

    0
Viewing 22 posts - 1 through 22 (of 22 total)

The forum ‘General’ is closed to new topics and replies.