I am no longer surprised by the number of people who say, “I’m working on my Green Belt/Black Belt certification, but I’m not sure how to apply Six Sigma to [health care, telecom, hotels, food, transportation]—insert your service industry here. They just can’t seem to figure out how to translate manufacturing-oriented training and case studies to their industry. It’s not that difficult, but that’s what I do most of the time: help people translate the “map” used in manufacturing to their particular industry.
While manufacturing focuses on reducing variation (I call it “deviation” because it deviates from the customer’s requirements), services can best benefit by focusing on eliminating defects, mistakes, and errors in the service delivery process.
Where’s the pain?
The first step is to identify the pain caused by defects, mistakes, and errors. Is it:
• Medication errors, hospital-acquired infections, surgical complications, ventilator-assisted pneumonia, repeat radiology or lab tests?
• Incorrect or inaccurate orders that lead to incorrect service delivery that leads to waste, rework, and lost profit?
• Inaccurate bills or invoices? Improperly applied payments? Denied insurance claims?
• Defects in delivered software?
• Preventable adjustments and credits?
Find the data
While define, measure, analyze, improve and control (DMAIC) lead most people to believe they need to start by defining and measuring something, data about the number or cost of these types of defects are already collected somewhere by someone. Find the data! As described in my previous article about Control Charts for Services, get the data into a format that can be graphed as a control chart.
Most corporate pain revolves around money: adjustments, credits, and and so on. Because Six Sigma focuses on cutting the cost of waste and rework, it’s often easiest to start with lost revenue or the cost of waste. Here’s an example of using the cost instead of a count of defects. This hospital system’s pain was denied insurance claims. I simply summarized and charted the monthly cost of denied claims.
The process was stable, so the hospital could plan on losing more than a million dollars a month. Until I combined the $100 denied here and the $1,000 denied there, it had no idea how much it was losing.
Mine the data
Once you know the level of pain, it’s time to start using Pareto charts to drill down. In this case, I started with various reason codes. Timely filing (i.e., not filing the claim within 30 days) was the main culprit.
But I rarely stop at the first level of Pareto chart. I drill down into the “big bar” to see if there’s a Pareto pattern within that. In this case, one insurer accounted for almost two-thirds of timely filing denials: 
Create the problem statement and gather a SWAT team
Once the problem has been narrowed to the most granular level, it’s easy to write a problem statement. The “big bar” on the Pareto chart becomes the head of the fishbone diagram: During 2002–2003, Insurer 1 accounted for 64 percent of denials for timely filing ($7,849,569), which was higher than desired and caused lost profit:
Most companies make the mistake of convening a team before they know what problem they are trying to solve. By narrowing the focus using data, it’s easy to figure out who should be on the team. After showing this analysis to the sponsor, she immediately identified the five people that should be on the root cause team.
This “SWAT team” of experts was able to diagnose the root causes and define countermeasures in fewer than four hours.
Implement and verify countermeasures
The team figured out the problem on a Friday and implemented changes to work around this particular insurer on Monday. Getting the insurer to comply with its contract took a while longer.
To verify that the countermeasures actually reduced the pain, add data to the control chart begun when finding the data. In this case, the change resulted in savings of almost $5 million per year: 
