iSixSigma

KBailey

Forum Replies Created

Forum Replies Created

Viewing 100 posts - 1 through 100 (of 195 total)
  • Author
    Posts
  • #110560

    KBailey
    Participant

    My personal favorite is “95% is pretty good!” when talking about one “opportunity” in a document with a dozen or more opportunities for error.
    My second favorite would have to be, “We’re too busy with all these other initiatives.” In some cases, it can be a valid concern. In our case the list of about 25 “initiatives” in progress included things like switching to a better tool to reduce strain and sprain injuries. It could have been done in a month, including training, but this was after 3 years of being pushed by the Best Practices Team and safety manager with minimal progress.
    (In case you’re wondering, the Best Practices Team was made up largely of actual users. We did look into why people weren’t switching and found that it was because they didn’t feel comfortable with the new tool. Upon further probing, we found that they simply hadn’t used it enough to get familiar with it. Those who gave it a chance found it to be superior.)

    0
    #110349

    KBailey
    Participant

    Delphine, you definitely need to focus better. You could do a project just on Return-to-Work, or on claim investigation, or on claim reporting. However, something that one of our instructors said keeps sticking in my mind. “Never spend your time improving at something you shouldn’t be doing in the first place.”
    In other words, work on reducing workplace injuries. If you’ve got a problem with fraudulant claims, you could also justify a project on investigation, but there may be some other factor that’s driving people to behave in that way.
    In tracking COPQ, make sure to watch other related outputs. Approach it the wrong way, and you may just shift a big chunk of workers’ comp costs to other areas including health plan, sick days, and lost productivity.
    This is an area where Six Sigma has huge potential, but it’s difficult to realize the benefits with a 3-6 month project approach. There’s nothing complicated about doing a DOE with people doing similar jobs but using different equipment in order to find out what’s ergonomically better. It’s simple enough to do, but it takes many people and a long time to get enough good data for a good model and cost-benefit analysis.
    It’s easier to say “You gotta pay thousands of dollars every year for workers’ comp” than it is to justify spending a couple hundred extra dollars for better equipment – especially when the furniture is also used as such a prominent status symbol to distinguish the bosses from the workers. We can’t let a peon who’s in their chair 8 hours per day have as high of a back on that chair as an exec who’s there 8 hours per week, after all. (That’s sarcasm with a purpose. It’s just one illustration of how we need to be willing to question what we do and why.)

    0
    #109890

    KBailey
    Participant

    Dan,
    Have you showed them the need for improvement? Have you measured the number of defects and the time spent reworking them, the “loops” in the process? Have you showed them the cycle time, breaking it out into a time-value map and showing the related cost?
    It sounds like you’re trying to get the point across that the workers are the problem when you haven’t demonstrated to them that there is a problem. Convince them of the need to improve, and then get them talking about how to achieve it. You may find underlying causes – like a need for information or approvals – contributing to their desire for flexibility.

    0
    #109691

    KBailey
    Participant

    Neel, I’d like to add something to the subjectivity discussion. There are two different cases that we call “subjective.”
    The first is when we rely on a human measurement system that we know isn’t as good as we’d like it to be. We see this a lot in sports, where referees or judges have to make calls or assign scores and everyone knows there is some subjectivity involved. Every measurement system has some degree of subjectivity, which we measure as reproducability and bias in a Measurement Systems Analysis. Sometimes it’s so difficult or expensive to measure that we accept a high level of subjectivity. For example, in a Six Sigma project, there’s only one best tool to use at a given time and one best way to use it, but it’s sometimes tough to get agreement on it. The only way to know 100% for sure would be to do the exact same project two or more times and see which way gets the best result. Difficult or expensive to measure…
    The second type of of subjectivity has to do with value judgments. Does something taste good? Is a particular color appealing? We can set up repeatable and reproducible measurement systems to quantify sweetness, tartness, saltiness, bitterness, etc., but that won’t tell us if it tastes good to a particular subject. Similarly, we can measure color in terms of precise wavelengths and intensities of light across the spectrum, but that doesn’t tell us if it’s appealing to a given subject.
    Management involves both types of subjectivity.
    On your new question: From a management perspective, Six Sigma is “easier” to use reactively. It’s easy to show cost justification by measuring how bad the process is before and after a project. When a process doesn’t exist, it’s hard to say how bad it would be if we don’t apply Six Sigma in the design. Even if we’re talking about preventing future defects, it’s easy to dismiss the benefits of Six Sigma by saying our managers and workers shouldn’t need Six Sigma to stop making so many defects or they must just be incompetent.
    Six Sigma does allow you to react, but the Control phase in DMAIC and the entire Design for Six Sigma approach are all about prevention.

    0
    #109019

    KBailey
    Participant

    The company you’re outsourcing to should be able to provide you with utilization % and similar metrics. You shouldn’t have to be afraid of “imposing” by asking for the info – you’re the customer and you’re asking for info they should be tracking for you anyway. Unfortunately, I am aware of at least one company out there who may be less cooperative. If they don’t want to provide info beyond what’s in the contract… good luck.
    You probably can’t hold the Call Center accountable right away for anything other than the Service Level Agreement(s) already in the contract. What you can and should do is determine what metrics are important to you and share that with them. If they want to renew the contract when it’s up, they should be willing to work with you now. When the contract comes up you can include new requirements including improvement goals. The types of measures should include things like closure rates (assuming it’s a revenue-generating call center), profitability (if they have any flexibility), customer satisfaction, order quality, etc.
    Paying per call received rewards the behavior of rushing through calls without caring about the outcome. Risky approach, depending on what other quality metrics are included in the SLAs. “Buying” agents puts you in a better position to push quality concerns, but it still doesn’t provide any incentive to be effective. Best approach is to include a of reward system tied to call quality as it affects your long-term profitability: closure rates, revenue generated, customer satisfaction, etc.
    In the long run, you’re going to pay for training in one way or another. I can’t tell you what’s common in call center outsourcing outside of my limited experience in the area. In general, training tends to be paid for by whoever stands to profit from it.

    0
    #108970

    KBailey
    Participant

    VGB, I find this fascinating. I may not be as much of a stats guru as some people on here. I would wager that if we did a pareto of factors in the success of Six Sigma in terms of impact to customer experience and bottom line benefit to the company in question, none of these “issues” would appear.
    I don’t know what the original poster was looking for, but I took it as being about the hot issues that will determine the future success and longevity of Six Sigma efforts. Give me good metrics for the key inputs and outputs and follow the DMAIC/DMADV roadmaps, and I couldn’t really care less about whether someone believes in the 1.5 sigma shift. Shift happens. The question is: what are we going to do about it?

    0
    #108963

    KBailey
    Participant

    Here’s a couple…

    Will Six Sigma be seen mostly just as a way to cut costs, or as a strategy to improve quality and innovation, gain market share, and grow the top line?
    Will management learn enough about SS to “pull” or only to “push” the discipline? (In other words, will they learn enough about it to at least know what questions to ask.)
     

    0
    #108916

    KBailey
    Participant

    Or as a sportscaster who worked on a local radio morning show said during his write-in campaign for governor:
    “A CD player in every home.” That was long ago, before DVDs and MP3s.

    0
    #108912

    KBailey
    Participant

    Some of you need to settle down a bit. Let’s get away from politics and back to the world of management and Six Sigma.
    “I have a PLAN to increase shareholder value. I’m going to cut prices for our small business customers so they won’t go out of business and they can keep buying from us. I’m going to raise pay for our low-income employees and add health benefits for our temporary, part-time, and contract workers so that everyone will be healthy and happy and more productive. To pay for it, I’m going to raise prices on our largest customers. I’m going to save money by putting an end to outsourcing and offshoring, and by not giving such big raises and bonuses to leaders who get results (including Six Sigma Black Belts) because we all need to think more about the common good of the company and our fellow employees. I have a PLAN.”

    0
    #108874

    KBailey
    Participant

    You could probably assume 5 is correct. One way to define “centered” is that the process mean equals the target.
    The trouble is that some testers would actually use such poorly-written questions and expect you to read their minds. I’m with aBBinMN that you shouldn’t assume mean to be the only/best measure of central tendency, especially given the way the question was worded.
    (If the question was intended as a measurement system, I’m predicting it to be a poor one.)

    0
    #108501

    KBailey
    Participant

    Success is a function of the opportunity present and your execution.
    If you’re dealing with a call center where you have a lot of paging through screens, looping back through scripts, and asking the person on the other end to repeat something they’ve told you, you have opportunity.
    Define your desired outputs. What is truly customer-driven and what are really the few key things you need to do call center management.  Make sure you have appropriate metrics and that they’re good enough for your needs. Put your scripts into a visual format and measure time, rework, jumping forward and backward, etc. Do a time value map.
    Do you have specialist groups for rework activities? If so, it indicates a problem with your up front process. Tough to say more without knowing whether you’re doing inbound, outbound, sales, service, or what.

    0
    #108329

    KBailey
    Participant

    I did a few related projects, one in the collection process and a couple related to different billing processes.
    Define your scope: will the project address the credit approval process, billing process, or collection process? Or do you first need to determine which of these processes is top priority? (Billing affects DSO, but the cost of carrying AR on the balance sheet may be very small compared to all the other COPQ of a bad billing process).
    For my first project we limited scope to the collection process of commercial accounts. We were working separately with a vendor to implement new credit approval/review processes. Our biggest savings were actually in bad debt write-offs and collection expense reduction. The main changes we made were:

    Work individual invoices, not “past due account”
    Work large accounts in-house but outsource many smaller accounts to company that was better able to contact customer’s at the appropriate time
    Accelerate contact schedule. We shortened the series of dunning letters, but started them earlier and followed up sooner with phone calls, and then accelerated the escalation to placement with a collection agency.
    Use Lean as well as SS tools. Look at the wait time between steps before the customer receives billing.
    It’s tough to do a meaningful DOE (and can take time), but you can do regression analysis using historical data. Analyzing a group of several hundred accounts outsourced at the same time, we found that the collectability dropped 6% for every 30 days from the last account activity. The age of the oldest invoice on the account was not a significant factor compared with the age of the newest invoice.
    Until your processes are consistent and timely and your billing is accurate, don’t get too hung up on the statistics.

    0
    #108165

    KBailey
    Participant

    RubberDude,
    You’re right about time as a quality factor. What we’re really saying is that caller perception and down time (or “customer satisfaction”) are CTQ’s we didn’t identify before.

    0
    #108162

    KBailey
    Participant

    Atul, I understand what you’re saying. However, from a cost perspective, the “acceptable limit” is itself a function of the return on a call. In most such outsourcing arrangements I’m aware of, the direct cost is born by the contact center operator and not directly by the client. It becomes less of a true customer need-driven CTQ and more of a cost control issue for the call center management.
    Because the acceptable limit is a function of expected return, we again get back to the success/failure of an individual call as a CTQ. Your cost to produce it is important to you, but it’s not a customer CTQ. To meet the customer requirement, the monthly average is a data point. To a call center manager who wants competitive advantage, it’s crucial to track performance to CTQs on a per-call basis.
    That doesn’t mean you ignore time. You just need to understand the bigger picture in order to set the right performance objectives with regard to time.

    0
    #108152

    KBailey
    Participant

    Reva, there’s no magic formula but there are some factors to consider, mostly related to caller satisfaction.

    Is there a time at which the rate of dropped calls increases?
    Is there a time at which satisfaction begins to drop off more quickly?
    Is there a call at which callers begin to become unhappy?
    What kind of satisfaction level does your competition achieve?
    Atul is partly correct. You have a USL of sorts, in that it’s a maximum for the average. However, it’s not a USL that lets you identify a particular call as defective – which is what you need to be able to do.
    The other approach goes back to the QFD approach someone mentioned. Your customer probably doesn’t really care about time – they care about dropped calls, unhappy callers, and success of calls (closure rate, revenue, etc. if sales). Determine what those true CTQs are and use that to measure your performance. Again, report your performance measures to the customer with a concise explanation of why youdo it in order to deliver more value.

    0
    #108131

    KBailey
    Participant

    Reva,
    Your customer doesn’t recognize the importance of that upper spec limit, which should be related to customer satisfaction.
    My advice is:

    Collect data on your distribution and on caller satisfaction
    Set a reasonable USL for yourselves
    In reporting to your customer include info about your USL and performance relative to it. Keep it short and simple – like a histogram with some satisfaction numbers for each point, and with a concise explanation of what it means and why you did it.
    If they’re smart, they’ll come around. If not, you still have to give them some credit – they were at least smart enough to contract you to help them out.

    0
    #107788

    KBailey
    Participant

    I think we’re talking about the same thing. Regardless of what euphemism management uses, if it involves getting people to work harder, take shorter breaks, eat lunch at their desk while they work (or skip it completely), and/or work longer days – I call it driving productivity. In my book, boosting efficiency means that you improve a process to accomplish more with less actual work – not just less dollars.
    For the record – wheeling and dealing to me includes legitimate selling and deal-making. If senior management is neglecting the parts of the business that actually deliver customer value, it’s a problem even if they are being ethical in their dealings.

    0
    #107717

    KBailey
    Participant

    I WHISH. Kinda. Kinda not.
    A previous supervisory position I had, about 95% of what my team did was reworking billing defects for both paper and EDI billing. My goal was to eliminate the need for us to exist. We never got to that point, but with some fair management support we did make some progress.
    Good consultants should look at it similarly. Managers are never going to be perfect – they’ll always need outside resources to help them. It’s a question of whether the consultants are going to catch them a fish or teach them to fish.

    0
    #107713

    KBailey
    Participant

    Am I a consultant? I wish!

    0
    #107706

    KBailey
    Participant

    Three other reasons:
    First: Sigma level can be calculated from a relatively small number of data points. Once you get past 3 sigma and get your defect rate < 1%, it becomes much more expensive to get enough data points to measure quality by actual % defective.
    Second: Taguchi loss function says that variation adds cost even if you’re within customer specs. Sigma level and Cp measurements give you an idea how much variation you have within specs – where most of your data points lie – instead of just the ones that fall outside specs.
    Third: Many people see 99%, 99.9%, or 99.99% quality as being pretty good. They just don’t naturally see a big difference between 1%, .1%, 0.1% defect rates, etc. Especially when it comes to processes with many steps and/or products with many components, it’s sometimes harder to get people motivated to improve when they’re looking at a number that they normall think of as “pretty good.”
     

    0
    #107703

    KBailey
    Participant

    One of my favorite things about Six Sigma is that if companies do it right, they can eliminate most of the need to bring in those expensive consultants.
    Instead, some management teams take the approach of cutting heads and driving productivity to the extent that the don’t allow their own people to spend any time on process improvement – let alone develop their knowledge and skills in that direction. Then they go out and bring in expensive consultants to figure out what they and their people could have figured out for themselves at a lower cost if they’d simply put together a plan and had the discipline to carry it out.
    IMHO, this is a breakdown in leadership. Management, in such organizations, is too busy wheeling and dealing to actually lead – or otherwise effectively manage – their areas of responsibility.

    0
    #107665

    KBailey
    Participant

    Rubber, I’m even more confused now.
    I’m saying that Six Sigma also has great potential value in areas where a sigma level of six probably is never going to be attainable – including politics, sales, marketing, sports, etc. How does that tie to your comment about posters who see SS as a “flavor of the month?”
    Six Sigma isn’t the end – it’s the means to the objective of competitive advantage and the goal of long term success. Either a) you’ll never reach a sigma level of 6 or b) you and the competition will both reach it. Six Sigma isn’t just about striving for a sigma level of 6 through DMAIC and DFSS, it’s also about redefining your CTQs and metrics based on what drives “customer’s” decisions and expectation levels in order to achieve competitive advantage – regardless of what you see as the “ideal” sigma level to aim for. If Six Sigma were just about striving to someday reach a sigma level of 6, then it would be a “flavor of the month.”

    0
    #107660

    KBailey
    Participant

    >many of the comments I see here come from misinformed people who have not read or studied the Six Sigma methodology and think it’s just another “flavor of the month.” and then, for some reason, want everyone else to think they know all about SS.
    What does this have to do with my remarks? I thought we were talking about the name, not the value or longevity of 6 Sigma.

    0
    #107653

    KBailey
    Participant

    RubberDude,
    For the record, my statistical knowledge pales in comparison to that of Stan and some of the other folks on here. I’m going to use a dirty word here, but I’d say I have more of an “intuitive” understanding of the implications of a concept, or how it can apply to a wide range of problems.
    Given the right tools, I can handle the basics. If you need the intensive statistical analysis… where’s that Stan guy when you need him? (I’m also surprised he hasn’t responded for so long. I hope he hasn’t flipped out too badly over some of my statements.)

    0
    #107651

    KBailey
    Participant

    >Statisticians will tell you that the six sigma level of performance is attainable and the best figure to work towards in any process – manufacturing or other.
    Is that so? What about the process of running a campaign for elective office? Or completing a pass in the NFL? Or the process of pitching a strike in baseball? What about a mass marketing process, where TV viewers and the output is people walking into your store?
    Two big reasons off the top of my head why Six Sigma won’t be the best level to work for in some processes: 1) competition striving to produce an output which is incompatible with your objective, 2) human involvement inherent to a process (sales, marketing, hiring, etc.) Both come down to some combination of too many factors that are too difficult to measure adequately and that interact in too complex of a way, and which even if you could measure and model adequately you still wouldn’t be able to control well enough.
    Some day, MAYBE someone will be able to achieve a sigma level of 6 in sales and marketing. That’s still not going to address the situation where competing parties are vying to achieve different and incompatible outputs. Six Sigma tools and methodology can still be invaluable, even if Six isn’t an attainable sigma level.

    0
    #107636

    KBailey
    Participant

    Using this definition of arbitrary from Webster.com: “based on or determined by individual preference or convenience rather than by necessity or the intrinsic nature of something”
    Choosing to use the value six in Six Sigma may not have been arbitrary as in “arbitrary and capricious” but is was arbitrary in that it was convenient for Motorola at the time. It’s definitely not due to necessity or the intrinsic nature of something. If Bill Smith and others had been at some other company and dealing with transactional processes, they could just as well have settled on 4 sigma. If they worked in sales, marketing, or recruiting, they might even have settled on 1 or 2 sigma.
    I think I understand about the 3.4 defects per million just fine. Short term, we measure the process and calculate a probability or projected defect rate. We don’t actually run off a million widgets (or opportunities) and count to see if we have 3.4 defects. A casual reader of the “New to Six Sigma” section isn’t going to get that.
    The other thing they’re going to miss is that the way we measure processes and calculate sigma level is going to produce more than 3.4 defects per million. The 3.4 DPMO calculation doesn’t factor in the likelihood of special cause variation. What do we do when we identify special cause variation? Eliminate it, and then remove those data points from our control charts.
    Your point #3 hits on the key: 80%-90% of the manufacturing organizations in the world today. Maybe it’s because of my background outside of manufacturing. Some people get too hung up on that six sigma level and 3.4 defects per million and they conclude Six Sigma methodology can’t apply to their processes where 95%, 75%, or even 50% success rates seem (and may very well be) unattainable. They need someone to explain to them that Six was a good level to settle on at the time for the people who were involved, but that as a name for a business strategy and quality methodology it’s arbitrary.

    0
    #107628

    KBailey
    Participant

    Mike,
    Your post touches on some of points which I would say support the claim that it’s arbitrary. As we get closer to 4, 5, and 6 sigma quality being more common in both manufacturing and transactional processes, the bar will rise.
    I’m not sure if the chart I’ve seen is the same as the one you’re referring to, but the one I’ve seen shows that even 6 sigma isn’t really very good when you get into very complex products/processes with hundreds of thousands of opportunities for defects.

    0
    #107626

    KBailey
    Participant

    Stan, I didn’t suggest you were changing definitions. I just wanted to make sure 12 sigma might have some special meaning I wasn’t familiar with.
    I’m not disputing your statements about NASA or the apparent improbability of achieving a sigma level of 12. Practically speaking, I think it’s a bit unrealistic. However, if you had a process that’s performed often enough (like if lots of people would die from it), and if the cost of failure were high enough, and if the cost of preventing defect were low enough, you could conceivably want to set 12 as your sigma level target.
    Before you just dismiss this, consider: CTQ’s can be so commonplace and expected that we don’t even state them. If I’m having a house built, I’m not going to tell the builder that one of my CTQ’s is that the roof has to stay up. If I were shopping for baby formula, I wouldn’t bother to check for radiation levels below a certain spec (although I can conceive of a world where that would be a sad necessity.) The reality is that we don’t actually bother to measure all of the true CTQ’s or opportunities for defects, simply because the cost (or bother) is too much to justify based on how confident we are in the quality with regard to that particular CTQ.
    The relevance of this is that – especially in DFSS projects – we can’t assume that the customer’s stated requirements are comprehensive. If you don’t meet some requirement that they just assumed, you’re going to disappoint them badly.

    0
    #107618

    KBailey
    Participant

    RubberDude,
    Can you clarify which statements you take issue with and why?
    As for spelling and punctuation: I’m aware that many of the posters are from outside the US and UK. However, those of us from the US and UK also spell and punctuate our posts with a sigma level far below 6. Regardless of that, the point is that we don’t even try to achieve a sigma level of 6 in our spelling and punctuation. The reason we don’t is the same as the reason I say that the 6 in 6 sigma is arbitrary.

    0
    #107612

    KBailey
    Participant

    12 sigma is 2 * (6 sigma).
    Actually, you may have a fair question there, Stan.
    If you’re asking whether I know just how high of quality a sigma level of 12 would be and how much it would cost to attain it, the answer (in layman’s terms) is really good and a lot.
    If you’re suggesting that I didn’t answer the right question because sigma level is talking about standard deviations on one side of the process average so (some might think) 6 sigma could just as well have been called 12 sigma, then the answer is that sigma level refers to sigmas between process average and closest spec limit after adjusting for 1.5 sigma shift.
    If 12 sigma means something different to you, maybe you can tell me about it.

    0
    #107605

    KBailey
    Participant

    The 6 in 6 Sigma is arbitrary. For most manufacturing processes, 3.4 defects per million is pretty darned good. If the cost of a defect is high enough, maybe 12 sigma is an appropriate goal – depending on the cost of achieving it.
    Obviously, we don’t strive very hard for 6 sigma in everything we do, including. For example, looking at iSixSigma forum posts we’re at about 0.5 sigma in spelling and punctuation. Nobody seems to be complaining too much about that.

    0
    #107603

    KBailey
    Participant

    I suggest using DPM or DPMO, depending on whether COPQ is driven more by number of defects or number of defectives.
    Bigger picture, it sounds like you may need to look at your changeover/setup process and the other things leading up to production, more than the actual production process. In mass production processes, quality is largely a matter of reworking the engineering and setup. You can’t afford to operate that way with mass customization.

    0
    #107423

    KBailey
    Participant

    Yes, the concept of quality applies. The effectiveness of a marketing process can be measured by the number of customers who actually respond to the product campaign.
    There are still issues arising in the measurement and improvement. With marketing, you may be dealing with many significant factors outside your control, with significant interactions. Many companies approach take a “Lowest Common Denominator” approach which can work well with low-cost goods/services. Others approach it by identifying a niche to fill. Rare is the company that has great success at measuring and understanding the many factors well enough to be successful at filling many niches.

    0
    #107421

    KBailey
    Participant

    Rene, your situation reminds me of my experience with Microsoft and AOL. It’s never a defect in their product, the problem always must be on my end. I’m amazed at how often a problem they didn’t have goes away in their next update.
    I would add one thing to what Stan, Old, and the others have suggested:
    You may have a measurement systems problem, as someone else suggested, but the true root cause may be a failure to adequately capture customer requirements. Your product may need to be more robust to variation in temperature, humidity, cleanliness, voltage, etc.
    Old, in cause #3, seems to see this as the customer doing something wrong. If the customer needs something that operates at different voltage, is it their fault, or is it your fault for not finding that out? Don’t rush to pin it on them, because if your competition beats you in making their product more robust you lose a customer.
    kb

    0
    #107360

    KBailey
    Participant

    What are some key leadership outputs?

    High team productivity (quality and customer service are components of productivity)
    Optimal employee retention/turnover
    Hiring and developing employees for internal advancement
    Process improvement
    Cost and asset management
    Planning
    Most of these are actually easier to measure than some of the traditional “core competencies” for managers. They don’t get used more because an individual leader only has partial control over them. Still, they can be valuable for an overall organizational improvement effort.

    0
    #107354

    KBailey
    Participant

    I’m confused. You cannot measure productivity improvement initially, but you know the process is performing above the required target, and yet you know there’s an immediate need to improve productivity… it sounds like either you’re experiencing unplanned/poorly planned growth, or like someone decided they just needed to cut some heads.
    If there’s “an immediate need” to increase productivity, shouldn’t your goal be based on that need? On the other hand, what’s the point of setting an improvement goal at all if you can’t measure it from the start?
    I hope you’re looking at some direct electronic data transfer in some form in order to eliminate the need to put it in manually from scanned images. It’s tough to increase productivity while reducing variation in a manual data entry process at the same time unless you’re actually able to reduce the amount of manual work required.

    0
    #107174

    KBailey
    Participant

    Who actually sets the specifications? In my experience, the IT folks have set their own “specification” by guessing how long is the maximum it could possibly take them. It’s not the same as a customer specification.
    An appropriate analogy would be one of those situations on the Enterprise when Scotty (representing IT) would say he needed x but Jim (the customer) would only give him y, and Scotty would always come through. Your specification isn’t what Scotty says he needs, it what Jim says you get – and this Jim (me) is pretty demanding when it comes to IT.
    For this process, I’d say that unless you’re able to install every security patch with zero disruption to productivity, you’re not at 0 DPMO. If I have to close an application or move away from my desk or reboot or do anything else differently than I would on any other day, that’s a defect in your process.
    If you can install the patch while I’m at lunch or at home sleeping and I can’t even tell you did anything, then maybe you can claim 0 DPMO. Do you still measure up to 100% yield and >6 sigma?

    0
    #106315

    KBailey
    Participant

    Sorry, Stan. I’m not taking the bait. I don’t need to prove anything to you. Considering the differences between people, no matter what I say you can dismiss it as incomplete/shallow/misleading, and call it a dumb response to a stupid question.
    The only soapbox I’m on is the one about whether you should try something before you give up on it, just because one “expert” thinks it’s a waste of time. Where would we be if not for people pushing through skepticism, defeatism, and negativity to prove the “experts” wrong about whether there’s a better way of doing things?
    Regards.

    0
    #106263

    KBailey
    Participant

    NewBB, I picked up something in a project management class that may help.
    Draw out a little matrix with two scales, one for importance and one for urgency. Both scales can go low, medium, and high. Think of a few tasks and where they fit in the matrix. The normal tendency for most people is to spend a lot of time dealing with the urgent things, regardless of whether they’re important. What we need to do is to spend most of our time dealing with the important things before they become urgent, so we can do things right. Think of it as fire prevention instead of firefighting.
    One reason for a perception that SS takes too long is because there are too many fires burning that just *must* be handled right now. The reality is that we’d be better off in the long run if we just let some of those less-important fires burn themselves out and spend our time on the important things.
    Six Sigma projects need to move along, but we do need to allow adequate time to do the important jobs right. That’s what being proactive, rather than reactive, is all about. That’s part of the discipline it takes to achieve long-term success. Remember: the reason you have so many good “opportunities” for projects is because somebody didn’t take more time to define, measure, and analyze when they created the process.
    Practical approach: work with management to classify needs on basis of urgency and importance, to get their agreement that it’s important to take enough time on the important things. Try to come to some sort of consensus to balance urgent needs of organization with risk of poor implementation.

    0
    #106254

    KBailey
    Participant

    Kim,
    Be careful with this one. Yes, it’s important to keep unscheduled absenteeism under control. However, the last thing you need is someone coming to work when they shouldn’t and getting everyone else sick. Not only can you create even more unscheduled absenteeism, but you’ll have sick, unproductive people making more mistakes than usual.
    A big part of attendance management comes down to measurement systems: How do you know whether someone is really sick?

    0
    #106209

    KBailey
    Participant

    In third-world countries in general, in your country, or in another specific country?
    Unless you’re in a very influential international leadership position, my suggestion is this: don’t set out to save the world.
    The best way I know to promote real SS is to apply the tools and discipline where you are. Be successful with it yourself. People who really want to be successful will imitate others who are successful. It’s a waste of time trying to convince people who are more interested in being right than being successful.

    0
    #106208

    KBailey
    Participant

    Product & process change frequently, so measure training effectiveness for each product and process separately until you have enough understanding of how product and process factors affect the call center output. Examples: price, perishable vs durable.
    Also remember, training process is an improvement process, meaning that you need to compare before and after in order to evaluate effectiveness. Otherwise, you defects from other processes like hiring will look like training process defects when they’re really not.
     

    0
    #106185

    KBailey
    Participant

    Mike (and SVP),
    It’s not about one taking precedence in the sense that you neglect one. You don’t go into business because you want to please government regulators. You don’t study pharmacology because you want to make the shareholders rich.
    There’s an analogy to the Kano analysis. In the case of government regulators, you must meet expectations but there’s very little value in trying to delight them. With shareholders, it’s important to meet minimum expectations but more is better. Your customer customer is the one for whom you’re seeking to meet previously unknown needs, to provide what nobody else can do for them so well.
    The other thing is that in business terms – as in sales revenue generation – customer has a different meaning than just a process customer. If we’re not going to keep this in mind and be clear which type of customer we’re talking about, we’re going to have to find a new word to mean what “customer” used to mean. Out of all the parties who can be process customers, there’s only one that presents the opportunity for top line growth.

    0
    #106183

    KBailey
    Participant

    Ram,
    Good call on the “blind statements.” I’m not sure what everyone else is learning in school, but it’s always been my understanding that the p<5% rule is "a rule of thumb" based on 95% confidence being pretty good for most purposes. We need to remember that a useful generalization is a useful generalization, and part of our job is to recognize when it doesn't apply.

    0
    #106181

    KBailey
    Participant

    Stan, what Six Sigma planet are you on? “Completely successful?” Where are you getting that? “Completely successful” means 100% perfection. Every Six Sigma practioner I know recognizes that’s an impossibility.
    The reason you apply Six Sigma is because it’s more likely to get you closer to perfection than most other approaches to improvement out there and WAY more likely than no improvement method at all.
    Nobody ever said that Six Sigma is the end all and be all for human kinds’ woes. All we said is that it could be of value; that it could help.
    Applying your standard that Six Sigma has to be completely successful first, I’m going to throw it right back at you and tell you to drop your attitude until you can demonstrate a chemical dependency treatment or any other program that is “completely successful.”

    0
    #106147

    KBailey
    Participant

    If you want to disagree with the reasoning, stick to that and back it up with some substance.
    If you just want to call peoples’ answers “dumb” and questions “stupid”, you might want to keep this in mind: when you throw insults around and are then proven to be wrong, instead of others seeing you as mistaken they’re going to see you as deserving the very same adjectives you used and worse.

    0
    #106146

    KBailey
    Participant

    Stan, did someone stroke your fur the wrong way? I’d think you were joking, but since your response isn’t funny, I’ll assume you’re serious.
    “There are successful treatment programs in place.” What does that have to do with whether Six Sigma could add value? There are also a lot of UNsuccessful treatment programs in place. There are successful processes in place in every industry. That doesn’t mean there’s not room to improve some or all of them by applying the Six Sigma BOK and methodology.
    Furthermore, the original question wasn’t limited in scope to treatment programs. Addressing a problem suggests prevention, not just treatment.
    Finally, many of the tools a BB would use are the same ones already being applied in the field of chemical dependency treatment. Chemical dependency treatment, like every other field out there, involves smart people who know a lot about what they do but who don’t always know the best way to methodically approach the problem. Well, if they can define a process, Six Sigma gives them a methodical approach.

    0
    #106145

    KBailey
    Participant

    Mike, I try not to lose the distinction between a process customer – which could be anyone – and a business customer – the ones who buy the company’s goods/services. In the context of a discussion of a company’s survival, we’re usually talking about a business customer.

    0
    #106134

    KBailey
    Participant

    DOE and other analysis tools are helpful, but are not unique to Six Sigma. One could apply Six Sigma to treatment and measurement processes – or even to education/socialization processes if you were really ambitious.

    0
    #106133

    KBailey
    Participant

    The risk is that there’s some other problem elsewhere in the value stream that would prevent the savings from actually being achieved.

    0
    #106131

    KBailey
    Participant

    Phil, you are missing a few minor details necessary for our survival.
    Our customers may love us, but we’ll still go out of business if we fail to control costs, if we fail to price higher than cost, or if the government shuts us down for some evil like pumping toxic waste into the neighboring day care center.

    0
    #105914

    KBailey
    Participant

    There are a couple ways you can apply Six Sigma/Lean to reduce unsafe acts.
    One way is to do it within regular DMAIC or DFSS projects. Based on an analysis of safety issues, you can design your equipment and process, and write your SOPs, to eliminate factors leading up to unsafe behaviors. For example, if you eliminate the need to reach overhead, you can stop employees from climbing on their chairs.
    The other way is to do a SS project on your safety processes. Define the specific behaviors you want to measure and how to measure, define what is safe and unsafe (spec limits), etc. Depending on your situation, you could have one project to improve the process of identifying and measuring behaviors critical to safety, the safety training’s process ability to change behavior, or both.

    0
    #105886

    KBailey
    Participant

    What Ritz said.
    Some specific times when FMEA can be helpful:

    Outside DMAIC/DFSS projects to assess areas of risk, in order to prioritize project opportunities that wouldn’t come to the top based on a short-term run rate. Some people wrongly have a very narrow view about prioritizing based on experiential cost that they wouldn’t do a project to reduce risk until someone is killed.
    When a process is too out of control for a good DMAIC project, and you’re trying to just get it under control without doing DFSS.
    During Analyze phase of DMAIC and DMADV projects. FMEA can help you avoid unintended consequences in your improvement.
    During Control phase as a reference to help troubleshoot and correct when process shows out of control symptoms.

    0
    #105845

    KBailey
    Participant

    Saint, I haven’t actually done this yet, but I’m hoping to get the chance with our call center. I’d like to know what you end up with.
    I’m not surprised at your results. I would expect this type of measurement system to need frequent, intensive calibration. Have you had everyone evaluate a call separately, then come together as a group to do it, piece by piece, discussing the conclusions and coming to some sort of agreement about the rating?
    Simplifying the form may not help, but if you’re doing the Gage R&R for the total score, you may want to break it down and work on the measurement for each criteria individually.
    If you’d be so good as to let me know how you proceed and what you come up with, my e-mail is kelly dot bailey at dwyergroup dot com.
    Good luck!

    0
    #105844

    KBailey
    Participant

    George,
    A colleague of mine did a project similar to this. A few findings that may be relevant:

    Much of the overnighting was due to employees not ordering supplies (esp. sales reps not ordering brochures, etc.). Marketing Dep’t was freely overnighting. Decision was to stop overnighting such things.
    Some overnighting was considered necessary. Example: replacement computers/parts needed due to a crash.
    Overnight (AM delivery) often wasn’t delivered until PM. Decision was to use next day instead of overnight
    In our case, no heavy duty statistical analysis was required for this project – mostly just Lean.

    0
    #105824

    KBailey
    Participant

    Amit,
    This is going to depend a lot on your business. Factors to control:

    Pricing structure. At least have some rules in place.
    Sales/negotiation policies, procedures, and training. It’s a challenge, because of the nature of the complex interaction between your proposal/rep (contollable factors) and the customer (uncontrollable factor)
    Procedures for researching customer and assessing needs/opportunities (include Kano analysis)
    I don’t know if this applies for you, but if you’re having trouble maintaining margins, I suggest looking for ways to add more value profitably (or sell features/benefits) before giving up pricing. Again, Kano can be very helpful to successfully differentiate yourself from your competition.
    kb

    0
    #105785

    KBailey
    Participant

    Is RFP “Request for Price” or “Request for Proposal”?
    Nitin Sahni’s proposed CTQ’s of timeliness, accuracy, and completeness should all be used. However, you mentioned conversion ratio, which is one of the most important metrics to your business – the other being profitability of the deal (a high conversion rate doesn’t help if you’re losing money on the deal). You could go with a single, high level process quality metric of Gross Profit per proposal.
    To determine if a specific RFP is good or bad:

    If you submit a proposal and don’t get the deal, that’s a defect
    If you get the deal but it’s unprofitable, that’s a defect
    If you get the deal and it’s profitable, but you only captured a portion of what you could have, that’s probably not a “defect” but it’s still lost opportunity
    Part of what you really need to get to is this: did your RFP process assess the customer’s true needs, the opportunity for you to meet those needs, the cost of meeting those needs, and communicate to the customer how you will meet those needs. Closure rate and profitability metrics will reflect this, but they’re not enough to tell you where you’re missing opportunities.
    Timeliness, accuracy, and completeness will help identify some areas of failure, but you really need to get to metrics for details like did you even understand what the customer needed and were you dealing with the right decision-makers. This is where it helps to know who/what the customer chose and why, and what their satisfaction level is. It also comes partly from tracking customer satisfaction on the deals you win. The results of processes like your RFP/Proposal often don’t show up until later.
    I hope this helps…

    0
    #105721

    KBailey
    Participant

    I forgot to make the point: real world learning is the more efficient and effective. When classroom instruction is needed, the best teachers are often those with up-to-date real world experience.

    0
    #105720

    KBailey
    Participant

    If you ask me, part of our problem as a society is that we rely too much on academic education as preparation for the job. The real reasons so many companies put so much stock in academic degrees? Because it saves their HR folks from having to come up with a measurement system that works.
    Academics can’t teach how to use things in the real world because they’re so isolated from the real world. Even if they once knew, they don’t keep up. Most of what they have to go by is what they learn from other professional academics.
    It all comes back to the fact that the outputs are a function of the inputs.

    0
    #105719

    KBailey
    Participant

    I’m not sure what you’re reading.
    Transactional quality is simply the quality of transactional processes. Transactional processes are – according to one common definition – any business processes other than manufacturing processes. I don’t use the term “tranformational” to refer to manufacturing processes, because many transactional processes also transform information or money or people or something in some way.
    When we say things like, “transactional quality usually stinks” what we mean is that the efficiency and effectiveness of transactional processes is usually very poor. Figure it out. What do you think the defect rate is in HR’s hiring process? What about a typical sales process? Invoicing, in a business that does manual invoice entry? A common mindset in office environments is, “If you get it right 95% of the time, that’s pretty good!” If you get 95% FPY on customer selection, 95% on Purchase Order # entry, 95% on parts/service selection, and 95% on pricing, you’ve got almost 20% of your invoice defective.
    In contrast, you wouldn’t tolerate that kind of poor quality in the bank processessing your checks and credit card transactions, or in your employer paying you correctly. As a business operator, you wouldn’t be very happy with a vendor that messed up your billing 15-20% of the time.

    0
    #105649

    KBailey
    Participant

    You’re not a BB, are you? Working on a paper for some class?
    Use DFSS when:

    You’ve done DMAIC and are unable to meet required quality level
    You’re designing a new process, product, or service
    You’re investing significant capital
    What are you looking for in expert opinions on transactional quality? Transactional quality is important but it usually doesn’t get the attention it should. Even in manufacturing processes, most COPQ can be traced back to transactional quality: hiring, training, identifying customer requirements, planning, purchasing, writing SOPs, process analysis (a transactional process in itself) etc.
    How can best practices be defined for something as broad as “transactional quality?” That’s like asking what’s the best play to call in sports. Best practice is to treat it as you would any other process – define, measure, etc.

    0
    #105637

    KBailey
    Participant

    Good answer, Ian.
    Once you apply Lean, it’s a lot less costly to put appropriate metrics and controls in place. Better yet, put the metrics and controls in place as you implement the changes from your lean assessment.
    k

    0
    #105554

    KBailey
    Participant

    Angela,
    Two things: First, the metrics are much the same as in a manufacturing or other production environment: money, defects, first pass yield, defect rate, etc. Second, don’t forget to apply Lean.
    The hard part is often getting people to actually care about measuring these things. Office people often tend to think in terms of “I need to review/do it manually to make sure it’s right,” or “My work isn’t COPQ” (even if it’s 95% rework). You’ll need leadership support, and you’ll need good political skills to overcome the cultural challenges.
    You’ve got to think of it like a factory. Everything that is produced should have a customer. Everything that has to get fixed is a defect and rework. Everything that has to be reviewed, that’s COPQ. All the time that a document is sitting around in somebody’s inbox or on their desk has a cost associated with it. Every time someone has to take or make a call to a customer/vendor that isn’t for a new order, the process didn’t work (there are exceptions, but the tendency is not to classify callbacks as waste when they really should be.)
    So how do you measure? If there’s any paper floating around, you’ll need people to manually help measure. Follow the flow of documents, and manually capture a sampling. Anywhere you can, capture date/time and other info electonically. A good place to start is wherever you have the most people doing pretty much the same thing, and where lots of dollars flow through – perhaps in billing, receivables, collections, or payables. Again, go back to the basic metrics I mentioned earlier. You’ll have to figure out what is a defect and what is rework for your specific processes.

    0
    #105553

    KBailey
    Participant

    Instead of Excel’s Data Analysis toolpack, you might want to just do a line chart. On the Options tab of the Add Trendline window you can select to display the equation and R-squared.
    Height/shoe size is a good example to teach how, but some people still won’t see how to apply the concept to transactional processes. Think of the key metrics you would apply to transactional processes: time, touches, dollars, defects, success rate, reject rate, etc.

    0
    #105548

    KBailey
    Participant

    Regarding the statement:  There is no need for SS if a company is profitable and you have repeat business.
    It’s not uncommon for companies in growth industries to be profitable and have repeat business, even if they’re not well-managed. The question is whether they’re maximizing shareholder value without breaking the rules.
    Well-managed companies, on the other hand, already apply sound problem-solving methods, measure many of the right things, make decisions based on data more than at poorly-managed companies, and have better organizational discipline – even if they’re not “Six Sigma.” Again, the question is not whether they’re doing OK, but whether they could be doing better.
    kb

    0
    #105367

    KBailey
    Participant

    Alifa,
    Don’t limit yourself to the other answers you’ve gotten. If your p-value is borderline, there are other things to look at before making your determination.

    Look for sources of variation that should be included in your model.
    Tighten up process control. Make sure everyone’s following SOPs and that those SOPs are adequate.
    If you didn’t already, make sure to check for interactions and non-linear relationships.
    Good look!
    k
     
     

    0
    #105313

    KBailey
    Participant

    John, the article recommended by RK is helpful, but there’s a couple things you can do differently.
    You can treat these numerical surveys as continuous data, rather than discrete, for many purposes.
    The other point relates to your preference for reporting proportion satisfied and proportion dissatisfied. As you obviously know, average just doesn’t tell the story. There’s a huge difference between an average of 7 where you have 25% 6, 50% 7, and 25% 8, and an average of 7 where you have 70% 9, 10% 5, and 20% 1.
    Easiest way I know to show management the difference is to get all the data and put it into histograms to compare the different groups. The significance to management  is that the dissatisfied people probably cost the organization more than what you gain from the supersatisfied ones.
    The negative is that the there’s a good chance that the average by itself will be misleading.

    0
    #105166

    KBailey
    Participant

    John, it works like this:

    Time is Money
    Knowledge is Power
    From basic physics: Power = Work/Time
    Substitute Money for Time and Knowledge for Power to get: Knowledge = Work/Money
    Solving for Money, we get: Money = Work/Knowledge
    Therefore: as knowledge approaches 0, money approaches infinity, regardless of work done.
    Does that help?

    0
    #105160

    KBailey
    Participant

    Like I said: With “perfect” forecasting…
    With perfect financial forecasting you wouldn’t need cash reserves because you’d be able to apply JIT to your borrowing and other cash management processes. You wouldn’t need the cash reserves because you’d know exactly when you needed to either convert other assets to cash or borrow, and exactly how much.
    Is it possible in the real world? No, not any more than perfection is possible in any other process. However, we’re not talking about setting an obtainable goal here; we’re talking about calculating COPQ. COPQ should be measured relative to absolute perfection. If you measure COPQ relative to what you think is practical, you’re limiting yourself. You’ll never know what’s truly possible until your competition achieves it.
    In your example, the weather is simply an uncontrolled, external source of variation. COPQ is COPQ, regardless of whether it’s the result of factors you control or those you don’t. If you can’t control it, you can still find ways to make your process more robust to it – but only if you recognize it as COPQ.

    0
    #105150

    KBailey
    Participant

    With “perfect” financial forecasting, the business could (theoretically) operate with $0 cash reserves, without hurting your ability to pay vendors or creditors in a timely fashion. The benefit could go to reducing need for investment capital (you could buy back stock), reduced need for borrowing (save on interest costs), earlier payment of vendors, or get invested in something that will generate higher ROIC.
    Hard COPQ calculation for this type of project should use the lowest of these values: interest paid on borrowed cash (assuming credit is available), expected ROIC of investors, or ROIC for additional investment that was made possible. It doesn’t matter if you could invest the savings from your project and achieve a 20% ROIC – if you could have borrowed the money at 6%, your project savings is only the 6%. You only get to count the higher ROI if the thing that’s stopping you now is that you can’t get the cash for it.
    There are all kinds of other major costs associated with poor financial forecasting. If headcount could be cut based on financial performance, good forecasting lets you reduce by natural attrition and other methods that are less costly (in dollars, morale, and productivity) than layoffs. Especially if the stock is publicly traded, the biggest cost may be to stockholders – related to investor confidence in management’s projections. You may be able to quantify this by calculating the relationship between P/E ratio and accuracy of financial forecasting for a sampling of publicly traded companies.

    0
    #105027

    KBailey
    Participant

    Nick,
    Forget the ice water or boiling for this. There’s a fundamental question you need to answer: What are you calibrating against? In other words, how were the specs calibrated?
    There’s little or no value in calibrating against some “objective standard” unless the specs are calibrated against the same standard. I’m guessing your range of 16° 29°C was set at “ideal” room temperature +/- 6.5 degrees to get a work environment that’s comfortable enough by the boss’s standards. If so, you can calibrate your thermometers against the average readings of all of the thermometers. If one or two thermometers are more than a degree or two from the average, don’t include it in your average calculation. Only if several thermometers are more than a degree or two off average is there a compelling need to bring in a calibrated thermometer, unless you have some reason to believe that your average is off in one direction or the other by more than a degree.
    If you can adjust the measurements or controls across the entire temperature range, you should calibrate them across the range from 16 to 29 degrees (forget the ice water and boiling points.) If you can only adjust up or down by some number of degrees, calibrate either at ~22.5 or using an average of the measurements at 16 and 29.
    If your temperature range was actually set by customer spec, calibrate against the customer’s thermometer. If it’s because temperatures outside that range adversely affect your process, you can still calibrate your thermometers against themselves and then calibrate your specs.
    IF the specs were actually set because of actual measured process performance, and IF you can’t adjust the specs, then you may actually have to look for a quality, calibrated thermometer from outside – but it should still be calibrated by the same standard as the thermometer(s) used to set the specs.

    0
    #104946

    KBailey
    Participant

    Like Darth says, you do need to frame your question more specifically. However, framing the question can be considered part of the Define phase of a Six Sigma project.
    Define what the good thing is that you’re afraid might not happen, or the bad thing that might happen. What are the input factors that can contribute to the risk? Define it, then either MAIC or MADV it.

    0
    #104460

    KBailey
    Participant

    Andy U has got the idea. Perhaps I can add something to help clear through the mud on this topic.
    One criteria for metrics validation is, “Does it drive the desired behavior?” DPMO can drive quality improvement in process steps or in particular opportunities. However, DPMO does NOT drive lean in the same way defects per unit/million or defectives per unit/million do. Overemphasizing DPMO rewards the behavior of including unnecessary complexity and opportunities, and it discourages driving process quality for each component to the level necessary for the desired customer experience.
    If I buy a new car, I don’t really give a darn how many opportunities there are. If I have to bring it back to the dealer to be fixed, I’m going to be unhappy – regardless of whether it needs a new windshield wiper blade or a new battery or a new engine and transmission.
    It’s good to keep score and be able to compare the effectiveness of different processes, but what I’ve seen on this thread makes me wonder if some people are focused too much on that, and not enough on delivering a defect-free product or service.
    Other than that, the question relating to the Taguchi Loss Function is interesting. About the only thing I can think to add to that topic is that DPMO can again be misleading, because the cost of a single defect (or even variation within specs) will tend to increase with the number of opportunities or level of complexity of a product due to cascading through the system. It also tends to be more difficult to isolate and correct problems that have more complexity/opportunities.
    kb

    0
    #104443

    KBailey
    Participant

    Yes it’s possible, although the normal Six Sigma project methodology was designed for processes that you perform over and over, more than for a one-time project. Six Sigma can be applied very effectively to the project-related processes such as requirements definition, design, development, testing, implementation, training, project management, etc. Once you’re into a specific project, there are tools that can help (FMEA, Value Stream Mapping, QFD) but one might question whether it’s true Six Sigma methodology to apply them to a project.

    0
    #104395

    KBailey
    Participant

    This is an interesting problem. To measure repeatability, does the “operator” have to forget the entire call, or is it good enough if they just can’t remember how they scored it?
    Because there’s so much value in knowing, and it’s so difficult to get it perfect, I say go ahead and have them listen to the exact same tapes at least twice. Just make sure there’s enough separation to ensure they can’t remember how they evaluated it previously. It may help if they don’t know beforehand that the call in question is part of the MSA. For example, you could select your sample from calls that have already been evaluated once, and slip them back into the mix later a week or 2 or 3 later depending on how many calls the evaluators listen to daily.
    If you want to be really thorough, you could do a DOE to determine how long you need to wait and what other factors you need to account for in order to effectively determine repeatability in your system.

    0
    #104341

    KBailey
    Participant

    I’m not being cynical, just realistic. I’m aware of the Forum’s “rule” I think we may have a problem with unrealistic expectations here. How many people actually read and follow all the instructions, warnings, disclaimers, and license agreements they’re “supposed” to? Very few. We can either blame them and think they’re a bunch of lazy idiots, or we can question whether we’re writing good rules.
    ISixSigma has advertising customers. My guess is that they’re paying to have their ads viewed by many people – not over and over by the same 6 people. That means that features like the Forum need to be customer-friendly to a broad audience including new people, not just focused on what’s convenient and entertaining for the moderator and a few regulars. So again, I raise the questions:

    Who are the forum’s customers?
    What do they need and want?
    IMHO, the reason people don’t follow the rule is because it’s a bad rule.

    0
    #104321

    KBailey
    Participant

    If I understand the question right, it’s “How should we handle the problem of posters who didn’t read past posts first, because it’s not interesting for us to read the same things being rehashed?”
    What’s the problem? The first questions that come to my mind are these:

    Who are the forum customers?
    What are their needs and wants?
    In this digital age, why should somebody with a question spend hours searching through thousands of past posts that may or may not answer their question? It’s easier to just ask the question anew. If you understand people, you’ll recognize that even if Stan or Darth or Phil has answered a question previously, the forum gives others a chance to answer – developing their communications skills while building their confidence and self-esteem. Answering other peoples’ questions lets us feel smart, it lets us be helpful, it helps us feel good about ourselves.
    Wouldn’t you agree? Or would you say that the forum exists to entertain those of us who don’t need help with serious questions?

    0
    #104318

    KBailey
    Participant

    Hi Newbie,
    Sheri’s advice was good, but before deciding to put this project aside I’d like to explore something. This may be getting into the costly solutions you’re trying to avoid, but humor me. From your description of the problem and process, it seems that one thing we’d want to do is push the quality control and error detection closer to your customers. In other words, have the customer key the information themselves or have your agent/rep key it with the customer there to verify spelling. (I refer to keying, recognizing that there are also other ways to capture the info electronically.)
    One thing you didn’t tell us is the cost per defect or cost per defective. Use that as you look at possible solutions. Before you do give up on this to follow Sheri’s advice, let people know the general problem you’ve got so they can watch for solutions. Technology is changing fairly quickly, and a solution that isn’t cost-effective now may be in a year. Also, there’s so much technology out there that it’s tough even for the techies to keep up on everything available.

    0
    #103013

    KBailey
    Participant

    Multiply COPQ associated with a defect in each CtQ by the defect rate for that CtQ. The relative importance of each CtQ should be reflected in your COPQ measurement. If some CtQ’s are extremely important to customer in a way that is difficult to quantify in monetary terms, make sure that is factored into COPQ. (Example: risk of injury or death, or risk of losing large repeat customer account over a defect.)
    You can put this in a Pareto chart. Caution: a Pareto Chart of COPQ by CtQ, won’t reflect the possibility of a single root cause being responsible for defects in multiple CtQs. For this reason, it’s still important to identify root causes and then do a Pareto Chart by cause in order to determine which cause(s) to correct first.

    0
    #103012

    KBailey
    Participant

    I have to disagree with Ron on this. A Run Chart can be very helpful for a process that is out of control and/or undergoing improvement. Once a process is in control, a control chart is more powerful. Until then it can just clutter the picture.
    If you have too much variation, especially from external sources which you can’t control in the short term, you may even want to consider including a rolling average. In Excel, you can do that right in a run chart by adding an appropriate trendline.

    0
    #102855

    KBailey
    Participant

    If current sales process is oriented around selling by referrals from existing customers, and marketing process is oriented toward existing customers, you’re talking about major redesign or creating new processes – DFSS.
    Things to keep in mind:

    Referral selling can be great, but depending on the time it takes you may be leaving segments or entire markets untapped for some time, leaving room for your competition to move into those areas uncontested.
    Cold-call and related selling methods require different strengths than referral/relationship selling. Pushing your current sales force to do things they’re not suited for could turn excellent sales people into failures.
    DOE comes to mind, if I may state the obvious. Find out what characteristics your customers have that cause them to buy something from you and brainstorm about who else in the world shares those characteristics. Assuming your customers are rational, there’s something about your products/services/delivery that distinguishes you from your competition in being able to meet some need of theirs.

    0
    #102836

    KBailey
    Participant

    Two things, Bram,
    First, you didn’t say if any of those solutions they’ve been coming up with for the past >20 years have ever been implemented. If not, they’re going to think it’s a waste of time to try to come up with anything new.
    Second, most innovation comes not from originating new ideas but from adapting other peoples’ ideas to your needs. The best way to get people to “think outside the box” is to get them to “look outside the box” at what is going on around them within the organization and outside. If your company has managed productivity by locking people in their cubes or chaining them to their stations on the line for 20 years, that’s a problem. Get people out to see what’s going on at your customers, at your vendors, at your competitors, and anywhere else where they can see people doing things differently.

    0
    #100427

    KBailey
    Participant

    Cath, to answer your question…
    Yes.
    Calculating sigma level gives you a nice number to compare one process to another, but adds no real, direct value. Sigma level tells you nothing about the dollar value of COPQ involved. Although you can calculate it for virtually any process, it is of limited value – especially for processes involving internal customers. There is always some ambiguity in defining CTQ’s and specs, and sigma level does not differentiate between defect types that may have vastly different associated costs.
    Focus on using the Six Sigma roadmap to reduce COPQ and/or increase value, even if you measure COPQ/value with “soft” metrics like warm fuzzy vs. cold prickly feelings for a process (e.g., employee birthday recognition).
    That’s my not-so-humble opinion.

    0
    #100237

    KBailey
    Participant

    Mike, I’m explaining causes for the current state, not justifying it. Long-term success of Six Sigma will require the strong leadership you and others have described. However, if management isn’t prepared to do their part, it’s not helpful to point our fingers at them and complain that they’re a bunch of pointy-haired idiots.
    It’s not about backing down on what they need to do, it’s about picking your battles, and picking the time, manner, and place in which to fight them. I’m not going to learn how their process really works so they don’t have to. I’m not going to manage their process for them. I may take on a bigger role “doing their job” with regard to project selection, partly because some managers will select SS projects to work on things they don’t have time for, rather than on their key strategic goals, because they think they know how to implement their own initiatives. This doesn’t work well.
    If you go back and read some of the posts, putting yourself in the shoes of a manager who just hasn’t had the right training to be a good Process Owner in a Six Sigma company, I believe you’ll see how the language would be perceived as whining and fingerpointing. My argument is that one shouldn’t conclude that managers are incompetent just because they don’t already have the level of process knowledge we’d like them to have. They absolutely need to be willing and able to learn, but they’re not going to be very open-minded about it if they feel their BBs are just whining or playing the blame game.

    0
    #100234

    KBailey
    Participant

    It requires several characteristics, even before the skills come into play.

    Zero tolerance for poor quality and waste. That doesn’t mean you dump on people. It does mean that you refuse to be content with “pretty good.”
    People who are driven to learn new things, understand how things work, and improve themselves.
    Proactive thinking. This is the aspect of time management that says, “Spend more time preventing fires so you can spend less time putting them out.”
    Pay attention to innovation taking place in the world around you. It’s easier to use other peoples’ inventions than to invent your own. There are two skills here – recognizing other peoples’ ideas that have value to you, and adapting other peoples’ ideas in different ways to your business. And no, I’m not talking about pirating intellectual property here.
    Think long-term. Breakthrough improvement only comes with risk. You must be willing to take a risk to achieve large long-term rewards.

    0
    #100168

    KBailey
    Participant

    In the interest of trying to identify a solution, I’m going to suggest doing what we did. Bring in someone to provide a couple days of Project Management training.
    Instead of the cookie cutter approach to BB selection so prevalent on this site, do a little KSA analysis (Knowledge, Skills, Attributes). You need to do is make sure your SS team covers all key KSA’s. BB selection is one way, effective time management and project management are skills that can be learned.
    As for expecting all BBs and BB candidates to be good project managers coming in, I suggest reading what I wrote about expecting all managers to be fluent in the language of Process Capability. https://www.isixsigma.com/forum/showmessage.asp?messageID=46028

    0
    #100164

    KBailey
    Participant

    Mia, I agree with your sentiments to a large degree – it IS their job to know their processes and learn to communicate effectively about them.
    However, I think you’re overlooking an important reality which actually ties into our educational system here in the U.S. Many of our managers, especially in service industries and transactional fields, were liberal arts majors. That doesn’t make them incompetent, it just means they may never have learned to think about processes the way we’d like them to.
    If we shut down most of our liberal arts schools and pushed everyone to study business, your expectations (and Mike’s, and others’) would be much more realistic. However, given that our education establishment has been telling young people for years how companies like “well-rounded” people with a “liberal” education, compounded by employers’ tendency to hire and promote someone with any degree ahead of someone with the analytic skills and process understanding but no degree, it shouldn’t be surprising that so many managers lack fluency in the language of Process Capability.
    My SS instructors always said, “It’s the process, not the people.” When we stop rushing to blame the managers, and instead recognize how the process broke down, we’ll be much more effective in doing our job and correcting the problem.

    0
    #100130

    KBailey
    Participant

    It’s reasonable to expect someone in charge of the line at a large manufacturing facility to be conversant in Process Capability. That expectation may not hold up as well when you get into transactional processes and smaller, struggling companies.
    Some companies need Six Sigma so badly precisely because their management team isn’t conversant in Process Capability; lacks the discipline to methodically solve problems and attack waste proactively; can’t define and their processes and problems; doesn’t know what metrics they need, even for the y’s. We’re not all like Motorola, GE, 3M, or Allied Signal with a plethora of MBA’s and Ph.D.’s managing everything.
    There’s a lot of really smart, talented managers who aren’t conversant in Process Capability. They can learn it with a little help. They know something’s broken, but they just haven’t learned process language enough to measure it properly or define it for a BB project without help from a BB or GB taking the time to learn the project and work them through it.
    What it comes down to is that if you’re a BB and don’t feel your management team is doing their part, you can either work to “change” them, or you can sit around and complain. Maybe it’s not *supposed* to be your job. But if you don’t do it, is it going to get done? It’s always tough to drive change, especially in your boss(es). So what! Acknowledge the grim reality of the current state, and figure out how to move toward the desired future state. That’s what BBs do.

    0
    #100128

    KBailey
    Participant

    No rule of thumb. We can’t necessarily assume that all orders are uniform in the quantity and quality of information required. Depending on the situation, the customer may also be introducting variation which we don’t have much control over. If order entry time affects customer, talk to the customers to find out what the specs should be.
    One thing you should look for is variation in entering identical orders, both for the same operator and between operators. Beyond that, break down the time into value-added and non-value added. If the variation is due to waste such as inspection or reworking defects, reduce it based on COPQ – not on some arbitrary internal spec.

    0
    #100127

    KBailey
    Participant

    Start by defining the processes, from the top down. Define the desired output(s) for each process/step; define what constitutes an opportunity (e.g, a customer walking in the door or calling on the phone); define what activities are rework. Measure all those things, and then calculate yield, first pass yield, and COPQ.

    0
    #100126

    KBailey
    Participant

    Champions and Process Owners need to pick the projects and write the problem/charter statement. However, they don’t necessarily know how to define the problem/project in process terms. Champions should be able to. In a “Six Sigma world” the process owners should also be able to. In the real world, BBs and GBs need to develop the definition enough so that the needed measurements can be obtained.

    0
    #100123

    KBailey
    Participant

    Casey, I have to question what you said about common cause and special cause being exhaustive and mutually exclusive sources of variability.
    It seems to me that there are two key measurements that define a cause of variability: frequency and magnitude of effect. We can also think of frequency as a measure of the probability of an occurrence within a given period of time or per million opportunities. Magnitude of effect is simply how big of an impact the cause has on the output. There will be variation in both frequency and magnitude of effect, but for our purposes we’ll consider that variation to be part of the measurement of frequency and magnitude of effect.
    We tend to think of special causes as discrete, or even unique events. The magnitude of the effect of such a cause may be discrete or continuously variable, although it must be large enough to produce an “out of control” warning on our control chart, in order for us to notice it. I submit, however, that if a process is allowed to run long enough, virtually any special cause will repeat. Isn’t that why we take steps to prevent recurrence of special causes, when we identify them? If we wanted, we could eventually capture enough data points of the special cause to model it statistically. Special causes can be discrete events, or they can be continuous like common waves which occasionally combine together to suddenly create an extraordinarily high peak.
    Common cause variation, we’re more likely to visualize as continuously variable in both frequency and magnitude of effect. However, this is also inaccurate. Any cause can be called common as long as it doesn’t produce one of the out of control symptoms. In reality, common cause variation includes a multitude of discrete “special causes” which mask each other or – by sheer random chance – don’t happen to have enough impact on the output to show up as out of control at a given time.
    In other words, there’s really no difference in essence between what we label as common cause and what we call special cause variation. It’s a continuum. I agree with you that “Dynamic Fluctuation” is not a third type of variation. Where I would differ with you is that I would say it is just a name for sources of variation which fall in the middle area of the continuum I described.

    0
    #99769

    KBailey
    Participant

    First, it helps to be clear on what you want to measure. I’m assuming you really want to measure overall job performance.
    Measuring quality on project work can include looking at tools used, benefits, and project schedule, as the other responses suggest. All of these are helpful, but keep in mind that benefits and project schedule are hugely dependent on project selection and the process owners. Evaluate tool useage, benefits, and schedule, but take into account the nature of the project, benefit potential, and obstacles, and the level of support/participation by others.
    I would also look for cultural change. Do the process owners and team members take some of the knowledge and tools with them, to apply in other things they’re involved with? Do they believe in the value of SS more than they did before? Do they sustain the improvements and pursue continuous improvement? Do they “pull” support from the SS team to help them understand and improve their processes? If not, SS is just another consulting program, not a culture or business strategy.

    0
    #99413

    KBailey
    Participant

    Am I missing the boat on SS, or is it other people?
    To me, Six Sigma is:

    Basic problem-solving methodology
    Organizational discipline
    DMAIC is not really some special new technique. Basic problem-solving, from 1st grade: first you define the problem. That’s DM in DMAIC. Then you identify possible solutions. That’s the A. Then you choose and implement the best one. That’s I.
    The first part of organizational discipline is simply remembering and applying what you learned, the C. The practical definition of intelligence requires a change in behavior as proof of learning. In other words, the lack of discipline to apply the C in DMAIC is proof of a lack of organizational intelligence. The second part of organizational discipline is recognizing the limitations of intution, no matter how smart you think you are, and looking for good information in decision-making; dedicating some resources to proactively apply basic problem-solving techniques.
    Where do people get the idea that Six Sigma is a tool or a set of tools? Most of the tools have been around for years and have application outside of formal Six Sigma programs. When I hear someone say Six Sigma doesn’t fit their needs, I interpret it to mean one of two things:
    Either they’re so confident in their intuition – in their ability to make it up as they go – that they don’t see a need for basic problem solving technique; or they’ve acquired a very narrow view of Six Sigma and don’t see how to adapt it to their situation. Probably 95% of innovation is taking other peoples’ ideas and adapting them to your situation. Part of being a BB is to innovate in applying the right tools to the situation, adapting the tools when appropriate.

    0
    #99363

    KBailey
    Participant

    I’m all for being flexible. However, Kliner described a situation in which they’d had 3 large successful projects, and didn’t see value in VSM because the returns are too small.
    Maybe I’m just stubborn, but I would not give up looking for bigger systemic opportunities. If you have lots of small “opportunities” in short run processes, that means you’ve got lots of defects in the processes leading up to those runs.
    Perhaps both you and Kliner have exhausted the possibility to use VSM to back out and get the higher level picture, but I didn’t see anything in Kliner’s writing to assure me of that. Especially in a forum like this where other people are looking for ideas, I think it’s better to cover the possibilities rather than risk missing what might be the key to success.

    0
    #99354

    KBailey
    Participant

    I can see how this is possible if you’re running a lean business with few quality problems. However, I think we all know that most companies don’t have that “problem.”
    We had trouble identifying good projects. Management wasn’t thinking in process terms. They assigned one team to reduce returns, another to reduce EDI billing rejects, and another to improve closure rate of a particular type of incoming call. The trouble was that there were several different processes that all contributed significantly to all three types of defects. The defect could actually happen in the process of taking an order, the process of verifying third-party billing information, the process of routing/scheduling the technician, or the process of ordering the necessary parts/supplies. Billing rejects and conversion rate both were also affected by the invoicing process.
    It’s possible to be looking so closely at the coins that we miss the fact that many different types of them are spilling from the same pot. Mapping helps you define the processes, and see all of their corresponding outputs. When you apply appropriate metrics, you get a better picture of the total COPQ and thus, “the pot of gold.”
    Maybe that doesn’t help you, but I’ve been around enough companies to know that it’s very common to have processes that are built haphazardly, due to the fact that nobody bothered to map out the big picture to truly understand all of the key requirements of the process and THEN figure out the best and leanest way to meet those needs.
    In this case, Kliner said the returns are too low to justify value stream mapping. That’s not the same situation as someone who’s mapped and measured and just doesn’t have the big opportunities. Without the mapping, you can see the “big picture” in a P&L sense, but how can you see the big picture in a process sense? Without the mapping to show the big picture, how can you really be sure that your process metrics tie to the bottom line the way you think they do?
    I’m not trying to be argumentative. It’s just that I’m convinced that VSM can have tremendous value in effective project identification in most organizations.

    0
    #99335

    KBailey
    Participant

    Kliner,
    I can think of two different causes for your problem.
    1) Lack of adequate process definition and metrics. Solution: work on value stream mapping. Identify and implement key metrics throughout in order to identify waste. It’s sometimes hard to get the resources. If you’re fortunate, your management team will see that the metrics will help them manage their business better and support implementation.
    2) Management wants you to work on the things they’re too busy for. I’m sure they’re working on their own inititatives in areas where they think they know what’s wrong and how to fix it, even though they don’t have adequate process definition or metrics (see #1). They assume the process owners know their own processes better than they really do, and they think about improvement in terms of either changing the people or changing the machines/technology. Solution: instead of looking for an SS project, ask to participate in the key initiatives management is trying to drive by themselves. Let them drive the project, you just provide some tactful backseat driving; you can offer them a roadmap, some suggestions, and when you think they’re approaching a bad decision, ask for permission to collect and analyze some data for them. Remember – you’re there to help them succeed; not to challenge their authority or make them look bad. With the right approach, you’ll get them wanting more of this knowledge-based management.

    0
    #98997

    KBailey
    Participant

    I’m going to throw another thought out here…
    Black Belts don’t save money. Process owners and teams save money. I’d bet money that if we really dug into it, we’d find project savings to be more a function of project selection and management commitment/participation than of the BB. A good BB may get you there a little faster, and may even be responsible for modestly better savings than a lesser BB (or even a GB) would have produced. However, I’d estimate the project selection and management as responsible for 80% of savings.
    So what do you rate BB on? Partly on savings, but also on the 360 evaluation suggested. Did the BB help the process owner and team members truly understand their process better? Did they come away from the project with knowledge of some tools they can use to help keep their process under control and achieve continuous improvement? Or did the BB just crunch numbers and dictate different SOPs to them?
    BBs are supposed to be agents for cultural change, more than statisticians. Don’t downplay the importance of getting others in the organization to see the value of the tools and methodology, or of being able to actually apply the basic ones.
    k

    0
    #98995

    KBailey
    Participant

    Mia,
    Don’t get too distracted by the turnover numbers. Focus on the resources that are currently being wasted but that can be freed up. If you free up resources, it’s management’s job to decide how to put them to good use. Long-term success requires some sort of growth or evolution. You don’t get there if management’s vision is limited to cutting costs.
    Talk with management. What would they do to grow sales, improve customer satisfaction, improve morale, or increase profitability if they had $x of capital to do anything they want, or x hours of free labor? Chances are they’ve got ideas, but they’re thinking they’d have to spend a lot of money to implement any of them.
    Good luck!

    0
    #98865

    KBailey
    Participant

    That helps for counting the immediate hard savings. You still need to be careful about driving people to avoid reporting. If you’re self-insured for your health plan as well as work comp, you may also want to look for strains, sprains, repetitive motion, and stress-related claims showing up in health plan costs. It’s harder to get the data, largely due to privacy concerns, but you might be able to get the administrator to give you some helpful, anonymous data.

    0
Viewing 100 posts - 1 through 100 (of 195 total)