As with any large organization, there can be trouble communicating between the top and the bottom. I frequently found that the managers in the middle did not wish to relay anything that might appear contrary to the opinions/beliefs of upper management. Frequently there is a specific expectation of what should be delivered and anything contrary will be rejected as unacceptable. Efforts to simplify were sometimes counterproductive as the low level detail reporting was removed by corporate, but middle management still required detail analysis for their comments. This results in increased work due to the change.
Caterpillar has a good model for promotions, but it has an interesting quirk. The model is based on a bell curve with normal distribution for performance ratings. You can NOT have too many people ranked in the highest percentile, and you MUST have some employees ranked in the lowest percentile. In spite of this they insist that it is not a 'forced' distribution of rankings among their employees with 'world class talent'. An interesting application of statistics for a Six Sigma company.
Six Sigma is a good thing for Caterpillar, but there are a few areas where it is being misapplied:
- The actual results of root cause investigation might not outweigh the force of a predetermined course of action. If someone has a pet-project there will be a Six-Sigma project that determines it is the required solution.
- Savings amounts grow in contrast to stated methodology. The first wave of Six Sigma was to go in and find the 'low hanging fruit' - the easy projects with biggest savings. After that, the hard-core detail analysis was to be used to find smaller levels of savings hidden in the detail. Largest to smallest? No - there is an ever-increasing amount of 'savings' that must be newly found each year. This drives a type of activity where measurement of improvement sometimes overshadows the concept of actual improvement.
