Practice Makes Perfect

So you think 99 percent accuracy in your business processes cuts the mustard? Proponents of Six Sigma say think again.

This article will help you:

Understand the long-term time and money commitments Six Sigma demands Evaluate whether Six Sigma could work at your organisation Learn about the payoffs of Six Sigma Six Sigma: A statistical term that refers to 3.4 defects per million opportunities (or 99.99966 percent accuracy), which is as close as anyone is likely to get to perfect. A defect can be anything from a faulty part to an incorrect customer bill. Six Sigma teams use extremely rigorous data collection and statistical analysis to ferret out sources of errors and to find ways to eliminate them.

Would you spend $450 million to get a return of $1.2 billion? General Electric Co. CEO Jack Welch sure would. That's the amount GE spent last year on its Six Sigma quality initiative. Welch expects to reap a hefty return for every dollar spent on Six Sigma, a comprehensive, statistics-based methodology that aims to achieve nothing less than perfection in every single company process and product.

And Welch has an excellent chance of achieving that return. In 1997 the Fairfield, Conn.-based company invested $380 million in Six Sigma-mostly for training-and received about $700 million in documented benefits from increased productivity and decreased waste. So it's no wonder Six Sigma devotees at GE and elsewhere tend to speak of it in terms bordering on miraculous.

Six Sigma can be a powerful tool for companies that compete on the basis of the quality of their products. However, in markets such as computer technology and retail, where innovation and speed are more important than quality, Six Sigma may not be worth the trouble. Even in companies where quality is a major driver, it's a long haul to realise tangible bottom-line benefits, and some companies put a halt to their Six Sigma projects before reaching that point. So it's not for everyone. For companies where quality rules the day, however, Six Sigma's proponents seem to outweigh its detractors.

A Brief Background

They say there's nothing new under the sun, and that applies to corporate quality programs, too. Despite the recent high-profile successes of Six Sigma at large companies like GE and AlliedSignal Inc., its tenets have been around a long time. Depending on whom you ask, Six Sigma was born anywhere from 10 to 20 years ago in corporate engineering communities, which are in the habit of reducing processes to statistically measurable phenomena (the cornerstone of Six Sigma). Motorola Inc. is the oft-cited creator of the formal Six Sigma methodology, which it has used since the late 1980s.

Six Sigma emphasises identifying and avoiding variation. For example, what causes a process to take 5 seconds when one person does it and 20 seconds when done by another? Six Sigma teams use a five-step approach to tackle specific problems such as inaccurate customer billing, sales policy errors, returned products, shipping errors, overtime expenses, rework expenses, inventory holding costs and long cycle times (see "Six Sigma's Five Steps," below).

Customer requirements, both external and internal, are paramount in choosing which Six Sigma projects to undertake.

Six Sigma has a language all its own, which seems particularly enigmatic to outsiders. Those who have completed the highest levels of training are called master black belts. Black belts spend all their working hours leading green belts-those who have completed a less rigorous training program-and other team members through Six Sigma projects. Six Sigma-trained people speak in terms of CTQs (factors that are "critical to quality"), the big Y (the customer's highest priority) and the vital few versus the trivial many (the projects that will have the most impact versus those that could stand improvement but are not critical).

It's enough to make anyone's head spin.

And many executives' heads are indeed turned by the breathtaking returns enjoyed by companies like GE, which is the biggest proponent of Six Sigma these days. But if you're a chief executive reading this article on an airplane, do yourself and your company a favour and don't jump aboard the Six Sigma bandwagon without considering the downsides. Six Sigma is an expensive, multiyear undertaking that demands an ironclad commitment-way beyond lip service-from the highest corporate executive.

In short, it's not a quick fix. Even though most market-leading companies operate at about Three Sigma, or about 99 percent accuracy, gaining that extra 1 percent accuracy doesn't happen in a matter of months. GE, for example, launched Six Sigma in 1995 and is pushing toward attaining that goal in all its divisions by 2000, when Welch is due to retire. Six Sigma pioneer Motorola began the program in 1987, and it took a full five years to see significant results, says Dennis Sester, vice president of quality in the Schaumburg, Ill., headquarters. Six Sigma has a long payback cycle at Motorola because it involves product and process redesign, and it takes years for a new product to permeate the marketplace. Sester says that if you're in it for the long run, it's worth it: Motorola attributes $15 billion in savings over the last 11 years to Six Sigma.

For such a long-term effort, every company needs major help from consultants to create and implement employee training programs. Mikel J. Harry, founder and CEO of Six Sigma Academy Inc., a consulting firm based in Scottsdale, Ariz., reportedly charges fees of $1 million and up to train the first wave of Six Sigma team members onsite, according to former clients. That breaks down to an investment of about $50,000 to train the 20 or so black and green belts that typically compose the first wave. (Several other waves of about 20 people each follow the first.) Board-Room Backing Those who have done Six Sigma agree the most critical success factor is top management support. "This is not the quality program du jour. We are serious about this," says Blair Souder, manager of PPDC (polymer processing development centre) injection moulding programs at GE Plastics in Pittsfield, Mass. (See "Plastic Surgery," below) That seriousness comes through loud and clear at the top of the enterprise. "Just look at the text of anything Jack [Welch] has said in the last three years. I'll bet he mentions Six Sigma more than 80 percent of the time," says Souder.

"The top executive must be part of Six Sigma. [He or she] must change the agenda of upper management meetings so the quality initiative is right near the top," says Dean Kropp, associate dean and professor of operations and manufacturing management at the John M. Olin School of Business at Washington University in St. Louis.

Mark Palazzo, general manager of the RF Electronics business unit at Scientific-Atlanta Inc., a $1.2 billion broadband and satellite network equipment provider in Atlanta, goes further. "Six Sigma has to be part of every discussion on the performance of the business," he says. Palazzo discusses Six Sigma results on a daily basis with his boss, Dwight Duke, vice president of terrestrial network systems.

The Latest Fad?

Employees can be forgiven for being wary of anything that smacks of slogans and T-shirts, which Six Sigma undeniably does. James R. Stanley of Howmet International Corp. says his company was already heavily involved in other strategic quality initiatives-such as Integrated Process Management (IPM)-two years ago when its biggest customer, GE, asked him to consider Six Sigma as well.

Six Sigma's Five Steps

Define: A Six Sigma project team identifies a project suitable for Six Sigma efforts based on business objectives as well as customer needs and feedback. As part of the definition phase, the team identifies those attributes, called CTQs (critical to quality characteristics), that the customer considers to have the most impact on quality.

Measure: The team identifies the key internal processes that influence CTQs and measures the defects currently generated relative to those processes.

Analyse: The team discovers why defects are generated by identifying the key variables that are most likely to create process variation.

Improve: The team confirms the key variables and quantifies their effects on the CTQs. It also identifies the maximum acceptable ranges of the key variables and validates a system for measuring deviations of the variables. The team modifies the process to stay within the acceptable range.

Control: Tools are put in place to ensure that under the modified process the key variables remain within the maximum acceptable ranges over time.

Source: General Electric Co.

"At the midlevels, they said, 'Why are we doing this?' I'm sure it seemed like the program of the day," says Stanley, senior vice president of U.S. operations for Greenwich, Conn.-based Howmet, a $1.3 billion maker of investment castings used in jet aircraft and gas turbine engines. Stanley admits at first he thought Six Sigma would be redundant. But he gamely shelled out $1 million for Six Sigma training for the first 18 months. That investment has paid off.

In June 1998 a Six Sigma team at Howmet discovered a way of controlling the silica percentage inside the ceramic baths in its casting operations. The silica percentage, which can get out of whack easily, must remain at between 28 and 34 percent. Otherwise the castings will be defective, causing costly rework. Previously, the manufacturing team would do the testing about every two weeks using a complex method. Under the new process, the team checks the silica percentage using density, which can be done in a few hours. Stanley estimates that this process improvement could save $8 million to $10 million in costs.

Middle managers may be especially resistant to the changes wrought by Six Sigma. The way to cut through middle management's fear, uncertainty and doubt is to link their pay and promotion prospects to getting training on Six Sigma, says Jim Martin, president of Six Sigma Integration Inc., a consulting firm in Rehoboth, Mass. "It's the only way to get their attention. If they don't have incentive to change, they won't." Indeed the high-profile Six Sigma implementers all link career paths closely to Six Sigma. GE's Welch says his people will not be promoted unless they complete 10 days of training spread out over two to three months, and that training involves finishing at least one project.

Common Pitfalls

As all employees quickly learn, metrics lie at the heart of Six Sigma. The basic approach is to measure performance on an existing process, compare it with a statistically valid ideal and figure out how to eliminate any variation.

Project teams might speak in terms of reducing cycle time, improving customer satisfaction, cutting down on returns and improving the speed and accuracy of order fulfillment. No project is considered complete until the benefit has been shown and a team of financial auditors signs off.

But even with rigorous measurements, it may be difficult to tell in advance whether an increase in customer satisfaction and quality will result-and to what degree-in a positive effect on the bottom line. "In the early days, it's easy to identify the costs of getting into [Six Sigma]. To identify the benefits early on may require a leap of faith," says Washington University's Kropp.

Another problem is most companies do not have enough financial people to be able to verify the benefits for the thousands of Six Sigma projects that are going on at the same time. As a result, the verification piece tends to get delayed, says Kropp.

Is Your Company Six Sigma Material?

If quality is the number-one priority at your company, Six Sigma will be appropriate as long as you also have management commitment and lots of cash for training. But not all companies compete primarily on quality, points out Shawn Bohner, vice president for performance engineering and measurement strategies at the Reston, Va., office of Meta Group Inc., a research and advisory services company. Every successful enterprise must decide which of three basic types of company it wants to be, says Bohner, citing Michael Treacy's The Discipline of Market Leaders: Choose Your Customers, Narrow Your Focus, Dominate Your Markets (Addison Wesley Longman Inc., 1995). The first type, companies that are operationally excellent like GE, lends itself to Six Sigma. For these companies, squeezing out defects and increasing efficiency are primary concerns.

The second category, companies that strive for customer intimacy like a consulting firm, does not lend itself to Six Sigma. Relationships, which is what this category is about, are often built on intangibles and not subject to statistical analysis as the processes in the first type of company are. "In this kind of organisation, if you try to figure out your customers in a systematic way, they might feel very uncomfortable," says Bohner.

The third type of company is the product leader. It can use Six Sigma, but only to the degree that it competes primarily on quality. For example, Nike makes high-quality shoes, but it competes primarily on being first to market with innovative designs. Companies like Nike are "looking for the home run, the big innovations. Their money goes into research but not so much quality improvement," Bohner says.

Software development companies, which typically strive to be product leaders, must likewise decide if a Six Sigma level of quality is desired. "You know how much it would cost to get only 3.4 defects per million lines of code?" asks Matthew Hotle, vice president and research director of application development for GartnerGroup Inc. in Stamford, Conn. "For a software company, you have to ask if the product to be delivered requires a Six Sigma level of quality. You have to apply the right amount of rigor [to controlling quality] but not so much you price yourself out of the market." Plastic SurgeryGE plastics has more than 3,000 six sigma projects completed or in progress.

Completed projects include

Quicker colour matching:

A Six Sigma team in Singapore reduced the lead time for matching colours of GE resins to customer requirements by 85 percent.

Reduced cycle time:

A new Cycoloy resin product was developed that reduced cycle time in the manufacturing of automotive instrument panels.

Quicker deliveries:

A Six Sigma project cut the time it took to get material across the border from the United States to Mexico from three to four days down to one day.

Enhanced productivity:

A team developed a new grade of Noryl GTX resin that reduced the number of painting steps needed and improved paint efficiency on top coats in automotive applications.

Source: GE Plastics

The Trickledown Effect

To date, it has generally been large companies that have implemented Six Sigma.

But, as was the case with Howmet, giants like $91 billion GE are increasingly getting their suppliers into the act. After all, the quality of a product is no greater than the sum of its parts. What good does it do to achieve perfection with one's own components only to have the overall quality of the product spoiled by a bad component from a supplier? "We're about four to five years away from large companies mandating their suppliers and partners get involved with this," says William Bartkowski, chairman and managing director of CorCom Co., a Minneapolis consultancy that provides Six Sigma training.

Some companies reportedly give their suppliers a lot of resources and other help to implement Six Sigma, according to industry consultants. Free training for the supplier company has occurred in some instances, and in some cases the large customer may receive a piece of the savings realised from the Six Sigma initiative.

Impact at Scientific-Atlanta

Although the press and commentary regarding Six Sigma to date is overwhelmingly positive, it is not all rosy. Sherita Ceasar, vice president of quality at Scientific-Atlanta, spent many years at Motorola during its Six Sigma heyday.

Although Motorola had a lot of highly publicised Six Sigma successes in the early years, that has not insulated the company from its current financial woes, which have been exacerbated by the collapse in global financial markets.

During the early 1990s at Motorola, Ceasar felt that the company was missing the boat by applying Six Sigma exclusively to the manufacturing process. This focus on manufacturing did not ensure the company was designing products according to customers' notions of quality or even developing the right products at all. When Ceasar came to Scientific-Atlanta two years ago, she made it her business to focus on the engineering and product design process rather than manufacturing.

Ceasar chose a new name, Impact (improving performance-excellence and cycle time), for Scientific-Atlanta's quality initiative. Ceasar sees a focus on engineering and new product development as the main difference between Impact and Six Sigma. "I wanted to start with new product development and ensure our products were reliable and manageable," she says.

The payoffs are beginning to show. For example, using Impact the Satellite Television Networks (STN) division was able to reduce the costs of two integrated receiver decoder products by 10 percent. It also dramatically improved the reliability of those products.

Today, Scientific-Atlanta has five-year goals of a tenfold performance increase per year and a fivefold cycle-time reduction per year. Six Sigma veteran Ceasar believes they will make the numbers. "There's a sense of urgency around the process," she says.

Of course, even Six Sigma's die-hard proponents agree the timing has to be right to attempt such a sweeping effort. If your company is in complete turmoil or you're waiting for the board to pick a new CEO, now is not the time to undertake Six Sigma. "The companies that have gone on record with documented multimillion dollar savings have had a lucky confluence of a charismatic CEO and an intact management team that could execute," says Jim Jubelirer, senior consultant at Burke Customer Satisfaction Associates in Chapel Hill, N.C. If you're in chaos, don't try this at home.

Lauren Gibbons Paul is a freelance writer in Belmont, Mass. She can be reached at

Don't Forget Technology

Six Sigma efforts won't succeed without a stable, robust IT infrastructure The collection of data is of major importance," says James R. Stanley, senior vice president of U.S. operations for Howmet International Corp., a $1.3 billion maker of investment castings used in jet aircraft and gas turbine engine components in Greenwich, Conn. After all, data is the essence of comparing one process to an idealised standard to determine variation, which is the heart of Six Sigma improvements.

Data collection is the Achilles' heel of many a Six Sigma effort, says Jim Jubelirer, a senior consultant at Burke Customer Satisfaction Associates in Chapel Hill, N.C. Jubelirer has worked with many companies whose Six Sigma initiatives stumbled due to a faulty IT infrastructure. "I have unlimited horror stories about companies' lousy computer systems. Say all I need is customer records. Many companies don't have those records. Or there is one system that keeps purchase and accounting data and another system for name and address, and you can't merge the two," he says.

This sort of woe, often caused by an amalgamation of disparate computer platforms thrown together in a merger or acquisition, is common at the large companies (several billions of dollars in revenues) that have been early adopters of Six Sigma. Making matters worse, there is currently no software program that goes across a variety of computing platforms to collect data for Six Sigma analysis.

Stanley knows IT's role is crucial. "Systems are vitally important. The systems have to give you data instantaneously," he says. Howmet CIO Chuck Elledge adds, "The IT infrastructure could make or break [the Six Sigma effort]. Data gathering [is] key." -L. Paul