|Date: Sun, 17 Feb 2002 17:48:50 -0800|
03 00050 61 02021701
Mr. Garold L. Johnson
City, St Zip
|Subject:||SDS implementation and promotion|
Excellent analysis in your letter today. I am sending a copy to Morris, since your comments reflect his worry that not enough people care about good management to make a market for SDS.
I am also sending a copy to Bill DeHart because he was the sponsor of SDS at PG&E and shares your perspective that good management is a worthwhile effort, per telecon with Morris and I on July 9, 2000.
You are correct that where people have tried SDS, Welch has not been invited back, which reflects the powerful attraction of bad management, commonly called feel good management that Morris discussed on November 23, 1991.
Probably, also, reflects my poor salesmanship on advantage of good management using intelligence for converting information into knowledge, along with strong cultural forces that...
These cultural forces make it difficult for a new way of working that saves time and money to take root and grow in an established culture, as Morris noted on May 27, 1999 On December 10 2001, you made a similar point relying on Buckminster Fuller.
Thus, there is a lot of evidence supporting your conclusion that selling SDS as a mainstream software program is a big challenge.
Keep in mind, as well, however, that the people paying the bills also have a problem. They have value at risk that requires good management in order realize objectives for getting things done on time and within budget. The only way to get good management is to specify it and then get tools and people who can perform.
That is why I asked in the letter yesterday for your comments on the scope of services for Com Metrics, based on your experience on the Boeing project.
It is not unusual for people to review contract performance to develop lessons learned and make suggestions for improvement. Requirements are incorporated into specifications to purchase performance. People would not prepare a CPM or a budget using C/SCSC, if they were not expressly required. People would not submit design details, if this was not required. Even when they are required, people try to get out of it, because it seems to take extra time.
So, one approach to solve the dilemma you pose is to come up with a scope of services that require people use Com Metrics, per Max Blodgett's comments on January 5, 1996.
Just wondered what you think about this idea.
THE WELCH COMPANY