Page 208 - Lean six sigma demystified

186 Lean Six Sigma DemystifieD

Deviation
Variation from the ideal target causes higher costs and lower profit margins.
Common types of deviation include

• Too long or too short
• Too big or too small

• Too wide or too narrow
• Too dense or too porous
• Too fast or too slow

Get the idea? It’s okay to brainstorm problems about one of these three templates
for improvement, but it’s usually worthless to brainstorm without these focuses.
Worst of all, most teams hesitate to identify the really pressing problems
because they don’t want to be on the hook for fixing them. In contrast, they
should focus on the worst first. Fix those problems, and everything else starts
falling into place. We have to stop majoring in minor things.
Recently, I facilitated a team that had been in existence for 6 months. All they had
to show for their time was a flowchart of a process that was mainly rework. I’d been
calling for weeks nagging the team for data about process performance. I got part of
the data the night before the meeting and the rest of the data by lunch. But after a
morning of trying to sort through the issues surrounding the process, the team had
fallen into “storming” about the whole process. They were frustrated and so was I.
Pitfall 2. Starting a team when you have no data (control chart and Pareto chart
minimum) indicates you have a problem that cannot be solved using Six
Sigma. Without data to guide you, you don’t know who should be on the

team, so you end up with different people trying to solve different problems.
Solution: Set the team up for success. (1) Work with data you already have;
don’t start a team to collect a bunch of new data. (2) Refine your problem
before you let a group of people get in a room to analyze root causes. You can
guarantee a team’s success by laser-focusing the problem to be solved. One
person can do this analysis in a few days using the QI Macros.
Pitfall 3. Question data. To throw a team off its tracks, some member who doesn’t
like the implications of the data will state in a congruent voice that the data is
clearly wrong. If you let it, this will derail the team into further data analysis. I
know from experience that all data is imperfect. It has been systematically dis-
torted to make the key players look good and to manipulate the reward system,
but it is the systematic distortion that allows you to use the data anyway.