How Success At TQM Can Be Failure At Managing
Reviewed: John Sterman, Nelson Repenning, and Fred Kofman. “Unanticipated Side Effects of Successful Quality Programs: Exploring a Paradox of Organizational Improvement”. Management Science, 43(4), April 1997. Available online.
Sterman et al. show how a fairly basic model can show the results of a TQM implementation at a largish company. They also discuss the problems inherent in the implementation of TQM, including the systemic imbalances created by most implementations.
The authors do not use RO language in their article, although I will be using it here to discuss the problems they raise.
One of the issues addressed is the need for quick results to bolster support of the program. These inevitably come in the form of Stratum 1 processes, since you can experiment and see results on them quickly. Higher Stratum (or Work Level) processes, such as product development, take longer to see results and are therefore not looked at initially.
The authors point out that the problem comes when you increase production in a Stratum 1 process but not in the higher-level processes that feed it. Production, always a lower stratum process, now can produce faster but there is no new demand that needs it. Capacity now outstrips demand resulting in redundancies / layoffs / cannings.
The other problem is that by working on the low stratum processes first, you now have a longer wait to create any dent in the higher stratum processes. While I am a democratic advocate, this is a strong support for those who say that change has to start at the top of the organization. Without that, you get things moving too fast at the bottom, resulting in the top thinning it out.
Of course, starting at the top and forgetting the bottom also creates problems. The higher processes, flush with faster times, is now trying to push these new needs to a pipeline that is not in any way ready to receive them. “Tony Stark’s” problems that I described awhile back probably fit here. He started at the top but ignored the plant-level issues. While it will probably work out in the end (just as ADI’s TQM did), it has caused him a few problems as the top grows impatient with the bottom.
Many of ADI’s problems seemed related to time-horizon, not thinking things through at a high enough level. Since they started and had phenomenal growth, it’s likely that ADI management were not big enough to be able to see through the likely problems, which the authors point out were clearly evident. However, hindsight is always 20/20. I’d still like to see how they line up. I’m betting that their roles grew faster than they did.
