posted 05 May 1999 06:14 PM               

I'm looking for some suggestion. Thanks.

posted 07 May 1999 01:25 PM               

quote:

---

the managers don't have the SPC culture

---

Implementing an SPC method in administrative functions is difficult enough. Unless the managers involved are at least familiar with the concepts of SPC, the task is virtually impossible. Perhaps the first concern can be addressed in the second.

quote:

---

what variable can we put under control?

---

You would need to identify which areas are causing the greater concern and look at those. For example, purchase orders with errors, time cards not correctly processed, invoices issued to wrong customers, with wrong dates or wrong amounts, etc.

Next, you need to collect data over a period of time, say daily for four weeks. Plot these data not just to determine control, but to see if areas of improvement are needed. Pareto would be a good start. Using the Pareto analysis, determine the affect these errors are causing (lost money, lost time, cost of reprocessing, etc.).

Using these data, present to the managers the amount of losses they are experiencing in their departments (if they are on a bonus system for productivity or actual versus budget, this is costing THEM money). If the losses are of significance, this may push the managers to become more lured to the Îthe SPC culture.â Money is the language of management. They will listen when you speak it.

Regards,
Don

------------------
Just the ramblings of an Old Wizard Warrior.

Check Out dWizard's Lair:
*** Dead Link Removed ***

posted 10 May 1999 10:24 AM               

Good input. Don't do SPC just for the sake of doing SPC. Window dressing seldom accomplishes much.

Determine areas that need improvement and then determine if SPC should/could be used effectively. The ideas listed by Don are good examples of what can be measured by SPC. I have also seen some great efforts by "office folks" applying SPC effectively. I had a list, which had some 200+ possible processes (office and otherwise) which was given to me by a gentlemen who teaches Statistical Thinking. I just don't know where I put it, but I use to hand this out to "office folks" who were having a difficult time determining where and how to apply SPC. I'll see if I can find it.

posted 11 May 1999 12:14 PM               

quote:

---

Don't do SPC just for the sake of doing SPC.

---

This practice is particularly irritating. In the past, I worked for a firm doing work that was DoD monitored. The DoD reps insisted that SPC charts be used, which was troubling, but no big deal. My particular problem was that the charts they wanted to see were at the 'far-end' other than where the belonged, the inputs. After all, the 'P' stands for process. They would glorify the benefits of p-charts for rejects but were totally disinterested in control charts on the processes. Most annoying.

quote:

---

Determine areas that need improvement and then determine if SPC should/could be used effectively.

---

Key word: effectively. Too often I have seen SPC charts that were no more than wallpaper. Samples include p-bar at 5% (or more) with no efforts at improving the process or 3 Sigma limits that are wider than the specification limits. Waste, pure and simple, waste. Kevin, I forget where you posted them, but you had three definitions of 'value added' somewhere. Those are great examples of when or whether SPC should be used.

Regards,
Don

------------------
Just the ramblings of an Old Wizard Warrior.

Check Out dWizard's Lair:
*** Dead Link Removed ***

posted 11 May 1999 01:38 PM               

I checked out your Powerpoint presentation. Very nice job on the content and with the bells and whistles. Thanks for sharing that with us.

Funny how the supplier gave the customer (DoD) what they wanted. I wonder what they thought of their customer? I know what I am thinking.

Your example is a good one. Why do toilet seats cost so much when sold to the government? Maybe because they require their supplier to run Attribute and Variable Charts on toilet seats. Do you think the supplier absorbed those costs? Yeah, right!

The three classifications:

Value added - Is what it is. Adds value to the product or service as perceived by the Customer (or not, such as a properly implemented SPC program).

Nonvalue added (but necessary) - Again, is what it is. Does not add value to a product or service, but in its absence can lead to waste. Inspection is my example here.

Nonvalue added (waste) - Is what it is again. Done for no particular reason and detracts from the value of a product or service (plotting and calculating control charts beyond their usefulness). The customer pays for this, a cost he or she would not want to pay for.

posted 11 May 1999 04:41 PM               

Don, Kevin,
I agree with you, the SPC is not some "sexy pictures" to show to our customers.
Two years ago I began to collect the data about external complaints due to customer service mistakes.
It was very hard because the sales manager did belive in that new activity.
We discovered that we had 5 or 6 complaints per month. We put this variabile under SPC and we fixed a goal.
After 2 years we have in average less than 1 complaints per month. It is a good result. We didn't spend a lot of money or energy to improuve the system, I think that it was sufficient to involve the CS employees in this project. But the sales manager does not belive that the spc helped him.
Now I'm trying to collect data about the time we need to receive a good supplier's reply to our complaints (8D), but I have the same problem with the purchasing manager.
I think that spc can help us everywhere, it is not necessary to arrive to a control charts system, often the data or pareto analyst are the best.
In fact the 8D process the data are the most important step: without data you cannot understand and improuve your work

posted 12 May 1999 10:19 AM               

There are many statistical tools, SPC only being one of them. You mentioned your success with Pareto analysis. This is often the easiest statistical tool to deploy because it is straight forward and easy to create and interpret. The advantage of using control charts vs. the direct analysis of the data is that incorrect conclusions may be drawn when looking at a bunch of numbers on a sheet of paper. Better I think to see how these numbers relate to one another.

I like your systems approach in your organization. Forward thinking is good to see on the other side of the ocean (as well as anywhere here for that matter). Now how to get the buy in from your fellow associates. Have these folks participated in any type of SPC (or other statistical tools) training? If not, then you might want to geet them training (I would use an outside source even if you are qualified to do the training). By using an outside source, you have another person pushing your agenda. I don't like the term 'push' all that much as I am an avid fan of 'pull' systems, but the obsticals you are overcoming are of a personal nature for most folks (they don't know and do not want to admit inadequacy at any level). Outside sources (professionals) recognize the learning blocks and have the expertise to deal with many different folks. When folks realize that you are not forcing them to do it and adopt statistical thinking for themselves, then real strides may be accomplished. I wish you luck!

Regards,