"Tyrrany of Targets" - An argument against Management by Objective

[Steve Prevette]

The attachment (hopefully I did this right) is an article I wrote a while back to provide an argument against Management by Objective, with a few illustrative examples, and some alternative logic to MBO.

- Steve Prevette
ASQ CQE, MS Operations Research
Fluor Hanford
City University

 

[Douglas E. Purdy]

Thanks for the read!

The attachment (hopefully I did this right) is an article I wrote a while back to provide an argument against Management by Objective, with a few illustrative examples, and some alternative logic to MBO.

- Steve Prevette
ASQ CQE, MS Operations Research
Fluor Hanford
City University

Steve,

I appreciated the article and the related resources.

Thanks,
Doug

 

[ccochran]

Nice article

Steve,

Nice work with the article. I have to admit I'm an adherent of reasonable and clear objectives, but it's hard to argue with your and Deming's logic.

Craig

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Craig Cochran
Center for International Standards and Quality
Georgia Institute of Technology

 

[amr1234]

Nice Article

I took this article to a meeting I had today with my lead auditor. He didn't and maybe couldn't even understand what was being imparted upon him. I wanted to tell him, in my most professional tone, "Read the **** thing, don't just roll it up." I didn't. The sound of my mother echoed in my head.. "You can lead a horse to water...." you all know the rest of that quote I'm sure. I went to the wall, after this meeting, and proceeded to thump my head a few times. Then continued on about my day. Thank God it's Thursday..

 

[Greg B]

Steve,

An excellent sell but would fall on deaf ears with our current management. They dsicovered 'balanced score cards' a few years back and are ardent fans of exact targets (98% operability, 98% throughputs, 20% reduction etc). I don't believe we have met all of the targets since the score card's inception.

We suffered similar Safety targets as your analogy with a required reduction from 18 to 3 (in one year) but we only achieved 4. So we recieved nothing because we failed. This was a massive reduction but did not meet the 'target'.

In saying al of this, I am still a fan of Objectives, Goals and Targets but they must be achievable, quantifiable and understandable. Here is my response to a question about personal goals from another thread and an extract from it below:
https://elsmar.com/elsmarqualityforum/posts/61107/

GOALS: A goal is an ultimate aim. It is what you want to achieve. It is a destination. You need to understand how you go from here to your goal. What do you need to do to get there? What skills do you require? What time frame? What help? What Cost? Etc
Goal: To be a millionaire by 30

OBJECTIVES: These are the stepping Stones to your Goals. They are short-term goals. An objective is measurable. What time frame and what indicators do you need?

· Objective 1: Graduate Business School in three years with Honours
· Objective 2:: Gain employment in the banking Industry by 2005

The Key to pulling all of this together is planning:

1. Establish realistic Goals and Targets, To achieve x by y,
2. Detail all requirements to ensure goals are achieved,
3. Draft a matrix of events (Timeline),
4. Discuss and communicate plan,
5. Implement the plan,
6. Monitor

Sound familiar? Plan, Do Check and Act


Greg B

 

[Govind]

This is a "must read" article not only for beginners also seasoned professionals.

Author in his conclusion says, "The key feature is the use of the control chart to determine if performance is stable or unstable (has trends). When the performance is stable, the responsible management should decide if improvement is needed".

This is something every organization, Quality professional should always keep in mind as Improvement 101.Today's world of tight project timelines people try to work on improvements without giving considerations to stability.

Another common error people commit working on improvement project is not validating if the improvement is statistically significant to declare.

Regards,
Govind.

 

[Wes Bucey]

As a fellow Demingite, I can hardly fault Steve Prevette's pronouncements, which mirror those of old W. Edwards "Management Did It" Deming.

If ever there was a case of preaching to the choir, this is it.

Good job, Steve!

 

[Rob Nix]

A very good read, Steve. Of course, it was like meditating on my OWN thoughts over the years and revalidating them. I especially appreciate your emphasis on using data to understand your system. Thanks.

 

[Mike S.]

I find the article to be well written, it is just that I an not so dead-set against setting "targets" -- and maybe Steve really isn't, either.

Consider this excerpt: "Establishing a Goal. With understanding of the system, we can then establish a goal. If we decide (through benchmarking, management review, and discussion with the customer) that 120 plus or minus 31 per year is an acceptable performance level and variation, then our goal is to maintain current performance."

Is 120 +/-31 not a target? In my small mind, goal is synonymous with target.

There are few times, IMO, that absolutes or the word "always" applies. Sure, some people or companies can misuse numbers, set unreasonable or impossible targets/goals, etc. for the reasons Steve so eloquently mentioned. I buy that completely. But to suggest that we should completely "eliminate management by numbers, numerical goals" is going too far. And based on some quotes from the article, I think that in reality, despite the title and the use of Demings quotes, Steve thinks so, too...

"The goal remains to achieve a significant improvement on the control chart. "

"An overall flowchart for trending, goal setting and achieving improvement is displayed above."

JMO

 

[Steve Prevette]

I find the article to be well written, it is just that I an not so dead-set against setting "targets" -- and maybe Steve really isn't, either.

Is 120 +/-31 not a target? In my small mind, goal is synonymous with target.

. . . But to suggest that we should completely "eliminate management by numbers, numerical goals" is going too far. And based on some quotes from the article, I think that in reality, despite the title and the use of Demings quotes, Steve thinks so, too...

JMO

Yes, it is necessarly to have goals. Even Dr. Deming did acknowledge that, in his higher level discussions of "aim" and also for goals. Goals I place upon myself may be worthwhile, goals imposed on others generally are not.

If I am having my kitchen remodeled, I certainly want the contractor to let me know up front how much it will cost, what will I get, and when will it be done. If I am knowledgeable about risks, I may want some what-if clauses - if my house is uninhabitable beyond the promised delivery, will the contractor pay my rent someplace else? If the contractor misses the deliverable, what is my recourse?

Those promises the contractor makes hopefully are based upon past experience, with the use of some statistical analysis of risks (control charts at least). I may even ask to see their past performance as the customer, which may help me in judging what my risks are.

But do I care how much the contractor pays his electrician? How much idle time there was during the project? What is the ratio of labor hours to material costs? No, I don't directly care. Does the contractor care? Yes. But my position is that it does little good for the contractor to establish internal targets/ numerical goals for those parameters. If they are of value, control chart them, understand them. Six sigma goes way overboard in requiring targets for EVERY measure (else how can you tell if you are six sigma capable?).

I work in government contracting. There does seem to be the expectation that targets will be established. For example, when working with radiation exposure, the rules call for estimates to be made of future exposure, and comparisons made. Unfortunately, the actuality of this has led to strange behaviors. A major improvement put into place was to control chart the actual exposures, and to provide a plus or minus band on the estimates. That at least allowed for the best (maximum likelihood) estimate to be supplied, and a worst case beyond which the guilty will be punished (I am being a little sarcastic here, but not much unfortunately).

If you take a look at https://www.osti.gov/dublincore/gpo/servlets/purl/807100-5YmXpk/native/ there is a paper on our use of SPC and radiation exposure analysis.

I still feel that the general rule for life (at least my life) is to eliminate targets and to use SPC to understand the data.

- Steve