How I Saved Millions Using a 1960s Movie Title
When I worked in the phone company in the 1990s, we had slow data transmission between our Denver and Albuquerque data centers. Our team knew there was something wrong, but we didn't quite know how to approach it with TQM. So we started looking, one by one, at each transmission line. We quickly found that some lines were performing as expected while others were running at half their potential speed. There were a dozen slow lines and we dubbed them The Dirty Dozen after the movie with Lee Marvin as a colonel that trains a dozen convicted murders to undertake a suicide mission in World War II.
One-by-one we investigated why each line was running slower than expected. In every case, one or both modems on either end were configured to run at a lower speed. On further investigation, we found that the original order for each line had been configured that way. They were installed as ordered! By doing “root cause analysis” on each line we were able to identify and verify the root causes as we discovered them.
We issued change orders and in a matter of days, all lines were running at maximum speed.
This step-by-step, line-by-line, defect-by-defect approach to analysis became one of the most effective tools in my arsenal of Six Sigma tools.
The Dirty 30 (Thirty) for Postage Costs
A few years later, I worked on a project to reduce postage costs. The phone company sends out millions of bills each month and postage costs had been climbing. On investigation, we found that the increased postage was caused by bills that went over the one-ounce limit. I remembered my experience with the Dirty Dozen and wondered if it would apply to this problem.
Fortunately (or unfortunately), the phone company also received 150,000 returned bills each month, so it was easy to find some two and three ounce bills to examine. We dug through the returned mail and started opening overweight bills one-by-one. Some were two ounces or more because they were business bills, but some were not.
By the time we’d opened 30 bills a pattern had emerged—small long-distance companies had been springing up and we’d started billing for them, just like AT&T. But each small charge added another page to the bill. It only took one or two pages to push the bill over the one-ounce limit.
The opening and analysis of 30 bills took no more than a couple of hours.
We opened another 30 bills just to confirm our theory. Then, armed with the evidence we went to the Finance VP to brainstorm countermeasures. It took awhile, but a redesigned bill, printed on both sides of the paper, slashed the weight of all bills saving $20 million a year.
The Dirty 30 for Service Order Errors
A few years later, I worked on a 17% error rate on telephone company order errors. We were able to identify the six most common types of errors, but there were thousands of these a day. What to do? Then I remembered The Dirty 30.
A few of us sat around a computer terminal while a savvy operator helped us investigate 30 examples of each type of error. I used a check sheet to keep track of errors found. Invariably, the process started slowly with various types of errors, but by 30, the root cause had appeared. And because we knew exactly what caused the error, it was easy to design system changes to prevent the error. We did this for each of the six main errors which accounted for 90 percent of all order errors.
It took only four hours to investigate and verify the root causes for each type of error.
It took several months to implement all of the changes in the system and when we were done, we’d completely eliminated five of the six errors saving $3 Million per year. That’s the power of the Dirty 30.
The Dirty 30 for Healthcare Denied Claims
A few years later, I worked with a hospital group that had over a million dollars a month in denied insurance claims. This time, I didn’t hesitate. We dove right into the Dirty 30 process.
Again, about six errors out of hundreds dominated the denied claims. We huddled around a computer terminal and investigated 30 of each type of error. In just a few hours, we found and verified the root causes for each type of error.
We even found that one insurance provider caused two-thirds of denials for “timely filing.” The team figured this out on a Friday, implemented process changes on Monday.
This one change saved $5 million per year.
We also found $24 million in miscoded denials. And the list goes on and on.
The Dirty 30 Process
The Dirty 30 can be used anytime there are “silent killers” of productivity and profitability. And you don’t have to be Lee Marvin to get employees to solve these problems.
- Just get 30 or more examples of each type of error (don’t try to do them all at once or you’ll get confused).
- Then analyze the root cause of each one, keeping track of the types and quantities for each cause.
- By 30, a pattern usually emerges. By 50, it’s clear as it can be, or the initial data analysis was off target.
Benefits of the Dirty 30 Process
- You can do it on almost any type of defect.
- You can’t go wrong. Analysis of each example delivers both a root cause and verification.
- By 30, not only are the root causes obvious, but the countermeasures to prevent the problem are also obvious.
- You can do it in a matter of hours, not weeks or months.
- It does take some Pareto analysis to narrow your focus to where the Dirty 30 reside, but this usually takes no more than eight hours regardless of how much data exists.
Where could you apply the Dirty 30 process to start getting results immediately?
Rights to reprint this article in company periodicals is freely given with the inclusion of the following tag line: "© 2008 Jay Arthur, the KnowWare® Man, (888) 468-1537, firstname.lastname@example.org."