BRIAN NEELY IS CTO OF A MAJOR SYSTEMS INTEGRATOR BASED OUT OF THE WASHINGTON D.C. AREA. I was speaking to Brian recently about his own focus on process improvement and how he sees process contributing to the success of large companies. "At one level, it's about expanded potential," he said, referring to the use of process programs within various company subsidiaries, "but at the executive level, it is about performance, especially performance measures."
Brian's focus is on empirical process improvement, the ability for operating units to demonstrate efficiencies through data and to establish performance trends over time through the analysis of data. That's a somewhat typical viewpoint for CFOs. Companies use financial standings, projected revenues, and burn rates as reliable predictors of success. The story is in those numbers. But that's a relatively advanced viewpoint for CIOs and CTOs. However, it's becoming more and more common. And more and more popular. As the IT industry begins to increase its appreciation that process lies at the heart of any technology development effort, the interest in process performance should increase proportionately.
That brings us to the subject of Six Sigma.
In the previous two chapters, you got an overview of ISO 9001 and CMMI. You saw that these programs were geared toward helping an organization implement process programs, perhaps to refine a current process position or to create a program from the ground up. In essence, those programs—those approaches—are all about instituting best practices. They help an organization set into place the structure for process improvement and, by default, quality management.
Six Sigma is different. Six Sigma has recently stepped into the field of IT as a Hot Topic. It has generated a lot of buzz, even more so than ISO or CMMI. People seem to be talking about it, discussing it, trying to figure out what place it might have in their organizations. Six Sigma is different from ISO 9001 and CMMI in that its focus is on measuring existing processes with a view to making them more efficient and effective.
Six Sigma assumes you have processes in place. Maybe they are formal, maybe they are informal, but you are definitely doing something to produce something. At its core, Six Sigma is a way to measure processes and then modify them to reduce the number of defects found in what you produce. With this program, you study the sources of defects and then analyze ways to make the processes more resilient, so that defects are not introduced or have fewer opportunities to creep in.
Many people think that the idea behind Six Sigma is to have a system that produces zero defects. That's not really true. But, statistically, the rote measure of "six sigma" means that your system will turn out only 3.4 defects per million opportunities for defects. (I'll look at this in more depth later.) But even that is not a really accurate representation of the intention of Six Sigma.
The real idea behind Six Sigma is to manage process improvement quantitatively. It seeks to put measures and controls in place so that you can readily and regularly monitor the performance of your processes and, using performance data, adjust them to maximize their ability to produce predictable, quality results.
You can think of Six Sigma as the evaluation side to a process improvement program. That's why many organizations pair Six Sigma with programs like ISO 9001, CMMI, or LEAN. Six Sigma gives you the tools you can use to rate how well these programs are performing for you. And this rating is not qualitative. It is not instinctive or intuitive. It is a rating based on hard data, on fact.
As this chapter looks at the high-level focus of Six Sigma, you'll see that it is a cycle of seven general steps:
Look at the product. Put a critical eye on what you produce. Continually examine what it is you make and how you make it so that you can always seek ways to make it better. There are few sacred cows in Six Sigma.
Identify defects. Examine your product and identify defects. Count them. Measure them. Know what you mean by the term "defect." You can think of a defect as anything that holds your product back from being the best it can be in the minds of your customers.
Look to the process. If the product is not all it can be, then chances are your processes could be improved. Examine your processes. What's happening with your current processes that might be letting defects in? What opportunities might you see to keep defects out?
Determine sources of defects. Analyze how the process works. Study its flows and structure to determine where in its operations defects are seeping in.
Improve the process. Based on your analysis of process performance and your understanding of the process structure, you now adjust the process with the intention of improving its performance. The goal is to lock defects out.
Use the new process. Now that you have improved the process, put it to work. Set it into the production environment and let the improvements make their mark.
Look at the product. Take a fresh look at the product. Did your improvement make a difference? Is the product better? If it is, look for new improvement opportunities. (And the cycle continues...)
Before we take a look at the push and structure of Six Sigma, let's go through a brief recap of the history of Six Sigma.
The foundation of Six Sigma was created at Motorola around 1979. Two engineers, Art Sundry and Bill Smith, were the pioneers. They were involved in the company's pager business at the time. And the business was having quality problems. The demand for pagers was skyrocketing, and Motorola was cranking up production to take advantage of the demand. But the pace proved to be somewhat hectic, and problems began to pop up.
Out on the factory floor, when a pager rolled off the assembly line, before it was packaged, it went to Test. If it passed there, OK. If it failed, it was rerouted to Repair. Once Repair fixed what was wrong, it went on to packaging and then shipping. When more and more pagers started finding their way to Repair, Sundry and Smith, both intrigued by this growing quality problem, began to track the lives of the troublesome pagers once they got out into the field, and they noticed an interesting fact. The pagers that failed initially in Test failed more often in the field even though these were the ones that had gotten that extra shot of quality control by going through Repair. The pagers that passed Test originally tended to operate defect-free in the field.
Sundry and Smith knew that Motorola was committed to customer satisfaction and looked around to see how the company was dealing with this. The model they saw was from the Classical School. Motorola was committed to a large in-house repair facility to fix anything it could find before it got into the field. And it also invested in many repair shops in the field, so customers wouldn't have to deal with dead pagers for long. This was the American business trend of the early 80s: Motorola was keeping its customers happy by reacting to (correcting) problems in the field. But Sundry and Smith had the following revelation. They saw (just like Philip Crosby knew) that quality through reaction is expensive. It takes a lot to support a reaction strategy. More people, more materials, more steps, more time, especially more money.
That's when they realized that the defect-free pagers were not only superior and imminently more reliable, they actually cost the company less money. The lesson was clear: improving quality reduces costs. It reduces costs by reducing activities and materials. The American business climate at the time thought pretty much the opposite: it costs too much money to build quality in; it's cheaper in the long run to deal with it in the field.
Sundry and Smith—both experienced engineers—began to develop statistical measures so that they could empirically analyze Motorola's production process to find out why these defects were creeping in. This analytical, numerical approach was essential because the pager-making process was large. It was complex. It had lots of interrelations. It was clear in this case that "gut feeling" was not the way to make improvements. What was needed was hard data. Lots of it.
Over a period of three years, Sundry and Smith implemented measures and techniques, casually labeled Six Sigma, across Motorola, measuring, measuring, analyzing, and improving. This new approach—the new program they had worked out—worked. It worked out very well. From 1985 on until today, Motorola has documented over $16 billion in savings from their Six Sigma efforts.
In 1995, Jack Welch adopted Six Sigma for General Electric. Jack Welch made Six Sigma a part of GE culture. He is largely credited with widely promoting the success of Six Sigma at GE and its potential for successful use throughout corporate America. His famous quote is that Six Sigma "changed the DNA of GE."
Jack Welch and his team are the people who shaped Six Sigma into the program we know today. GE popularized and formalized Six Sigma through such concepts as Critical to Quality, the Voice of the Customer, and the DMAIC methodology. Jack Welch has credited Six Sigma with saving GE hundreds of millions of dollars in 10 years of use. Six Sigma is so important at GE that today you can't be promoted at the company without being Six Sigma-trained and -certified.
In the last 10 years, Six Sigma has risen to the top as one of the most talked about process improvement and quality management programs available. It rivals ISO 9001 and CMMI in interest and adoption. But of the three, it is often the least understood. There are a couple of reasons for this. The first is that compared to ISO and CMMI, Six Sigma has the potential to be imminently more complex. If you move seriously into its statistical and quantitative aspects, it can be both powerful (to the informed) and powerfully daunting (to the uninformed). And then there is the question of what Six Sigma actually is. Is it GE's property? Did Motorola get a patent on it? Who is the owner of Six Sigma?