As he pushed out from Fairfield, Conn., USA, the king of CEOs left us all holding the Six Sigma bag. In it, a host of tools, techniques, potential and promise clank against one another. It is a nice collection, and we are pretty sure it works, but it is the old proverb that bothers us: “Be careful what you ask for; you just might get it.”

What we asked for was a better way to do quality, customer focus, measurement and change than had been provided by the old warhorse Total Quality Management (TQM). What we got was a methodology suffused with just as much hope and aspiration as the loftiest of TQM programs. With greater numbers of companies putting their hats into the Six Sigma ring, for the cynical, the currency of data-driven change is suffering from inflated expectations.

But before filing Six Sigma away in the category of failed hopes and dreams and scanning the menu for another flavor, consider the following caveats. They fall into the category of been there, done that, and constitute a few wise saws that even the most defect-prone management can benefit from.

Key Issues to Avoid

Six Sigma is nearly identical to TQM, but with more resources and concentration. More than anything else, Six Sigma is a total business strategy, not just a program du jour hiding out in manufacturing. The following are three key issues that may have been a problem during past TQM deployments, and should be resolved when starting Six Sigma.

Key Issue 1: You Can Pick Your Friends, But You Can’t Pick Only Friendly Projects

In selecting Six Sigma opportunities, practitioners often turn to pet projects that have been sitting on the shelf for a long time. Maybe teams from the past have taken passes at them, or maybe they have been Band-Aided and victory declared a number of times. They are the old favorites that no one has succeeded in solving and that have pestered management for longer than anyone cares to remember. But putting new resources into old jugs instead of focusing on today’s critical problems may do a disservice. There are common reasons for leaving such projects on the shelf:

  1. The project may be too big and beyond a Six Sigma team’s abilities to handle.
  2. The reason this problem has not been solved is political and not subject to the bright light of human reason.
  3. It is strategic, not tactical, and thus beyond the purview of a Six Sigma team. For example, asking, “Are we in the right business or market?” is out of scope for a team.
  4. The voice of the customer, either internal or external, was not taken into account.
  5. There is little or no data associated with the project and the Six Sigma team will have to spend months just collecting it.


Choosing good Six Sigma projects is not easy. It is more art than science and depends on discernment and judgment. Here are a few tips that may help a Six Sigma effort get a good start:

  • Set a time limit on the project (probably less than four months). This forces modest projects and provides the team with a moderate, obtainable objective.
  • Make sure the Champion or sponsor has power and commitment to the solution.
  • Ensure the sponsor has responsibility for the success of the project, not the team leader or members.
  • Look at past successes before selecting projects. Ask, “What are we good at around here and how did we do it?”
  • Use cross-functional Six Sigma teams for what they are best at – interpreting and evaluating data and making creative suggestions, not grinding through old, lost or suspect data.
  • Get the data first, and then have the team apply creative problem-solving techniques.
  • If the customer says something, take it seriously. Too many customer surveys unearth information that is not acted on.
  • Segment the customer sample to adequately reflect the customer base. Practitioners should be talking to the right people about the right things.
  • If management wants to do something – enter a new market, bring out a new product – this task should not be assigned to the Six Sigma team, unless the team is specifically chartered to build a new product or process.

Key Issue 2: Senior Management on Parade

There has been enough said to last a lifetime about the importance of senior manager involvement in quality. And, for the most part, those engaged in Six Sigma have gotten the message. The challenge of senior manager involvement has morphed from getting them out of their offices to something quite new and strange: getting them off the teams’ backs. In many companies engaged in Six Sigma, senior management has done such a good job of conveying their seriousness and the importance of the program that teams are sometimes quaking in their boots, fearful of negotiating overzealous charters, pushing back for resources, or demanding help gathering far-flung data and measures.


When a Six Sigma guru served as vice president of quality at one of the GE Capital companies, all efforts to get senior management to “get it” were in vain until the senior team was tasked to develop, research and present their own projects. Some were frivolous, but many represented sincere efforts to understand the principles of Six Sigma and make things better. Six Sigma is complicated – and there is no substitute for personal experience. Managers must learn to treat it like any other powerful business initiative. Put the right resources and expectations in place, set goals of moderate difficulty, coach and manage performance. Six Sigma will fail if it is treated as some magic wand by which management responsibility can be delegated away. Here are a few actions management can take:

  • Get involved in training. Not the one- or two-day version. They should take the full two- or three-week DMAIC or DMADV program, select a project, and see it through to completion.
  • Spend time developing and verifying the business case and problem statement part of the team charter. Know it cold. Next, tease out a four-month project that is doable, and set it up for success by picking the right team and team leader.
  • Give the team leader or Black Belt the right of first refusal on project selection. Most will go ahead, but they know it is by choice not constraint.
  • It is OK to select projects based on payback, but big payback usually involves a big investment of time and resources. Build talent and experience first with moderate, small-scope projects, and then apply that experience to the big-ticket items.
  • One way to properly scope a project is to use a multigenerational project plan. This allows all concerned to view the project, even a big one as a progressive rollout rather than a one-shot deal. Although usually used in a DMADV project, the plan can be part of a DMAIC project and specify each phase, making the project more manageable and facilitating handoffs to other teams.
  • In any project, large or small, the proper role of the upper manager, sponsor or Champion is one of involvement and coaching. It takes skill and experience to coach and that experience can only come from having personally led projects and from having taken responsibility for getting things done. The responsibility for implementation should not be left to the Black Belt alone, but owned by the Champion as well. If team members are expected to spend upward of 100 percent of their time on the project, the Champion must follow suit. The table below is a thumbnail guide of time commitments for all relevant parties. Note: these times equate to hours. Thus, 20 percent would mean a full day per week.
Time Commitment Required for Members of Six Sigma Team During DMAIC Project Phases
Project Phase Team Leader (Black Belt) Team Member Master Black Belt Champion/Process Owner
Define 100 percent 50 – 75 percent 50 percent 75 percent
Measure 100 percent 100 percent 25 percent 10 percent
Analyze 100 percent 100 percent 75 percent 15 percent
Improve 100 percent 100 percent 10 percent 15 percent
Control 100 percent 50 percent 10 percent 25 percent

Key Issue 3: The Vigor of the Rigor

One of the strengths of Six Sigma over TQM, particularly in administrative settings, is that while TQM talked about data analysis, Six Sigma actually does it. Sometimes, however, it can over do it. Statistics are only necessary when taking samples from a population. With statistical analysis, it is possible to infer things about the population from a sample as small as a few percentage points. That is a blessing in big systems with thousands or millions of events. When teams work on big scope sigma projects, statistics can be a blessing because they can only gather a small amount of data during their investigations.

About the Author