Dr. Deming, and the Folly of "Management by Objectives"

Dr. Deming takes a dim view of Management by Objectives. And he has a point. It seems logical to emphasize the meeting of objectives, but managing by them has difficulties. Your objective evidence of meeting objectives is historical data, so you end up "managing" much like if you painted your car's windshield black, and drove simply from the view in your rear-view mirror. You would be able to make progress, slowly and with many fender scrapes, but it would hardly be satisfactory. You need to have some kind of "forward-looking" metrics. Now, any "forward-looking" metrics will carry some risk, perhaps a great deal of risk. Imagine a windshield heavily smeared with Vaseline; you would lose all depth perception because all objects would image as smeared, flat blobs of color. But you would do better than driving with the rear-view mirror alone. "Forward-looking" metrics are based on some model of causality, based on incomplete knowledge and a simplified model of the world. It may not be any better than "viewing through a Vaseline smeared sheet of glass". But how much more satisfactory than managing by historical data alone. So last night I thought about what the exact components of compensation and management you would desire.

1. Historical evidence of meeting objectives: like bookings of a salesman
2. Data on activities that have a "causal" relationship to the desired objectives: like number of meetings and follow-up activities with decision makers
3. Data on activities that have a presumed, perhaps tenuous causal relationship to the desired objectives: like hours spent researching a possible new product offering. There is a significant chance of complete failure with a new product offering - that is why the causal relationship to the desired objectives is tenuous.
4. "Positive" human factors: I sometimes eat at a cafeteria salad restaurant. The cafeteria trays are handed out by a handicapped man. Part of his salary is due to the restaurant living its stated values. (But it isn't all "altruism", he brings value, he is the voice of the restaurant's handicapped patrons). Note - What isn't a "positive" human factor - keeping someone in a job because you feel sorry for them is not a "positive" human factor - it has everything to do with a manager who finds it easy to be generous with other peoples' money, who cannot deal with the anxiety of terminating someone for cause, and who would not hesitate to terminate that person during an economic downturn, which is the exact time they would find it hardest to find a new job.
5. "Negative" human factors: the percentage of salary that is based on kissing up, is another example.