________________________Articles| ________________________Articles |
|
Certification of Software Engineers Litigation of Software-Intensive Projects O-O-Design: ganz einfach oder sehr kompliziert? Professionalism in Software Engineering Reusing the Products of Analysis UML: A Curse or Blessing? (pdf file) Wer verträgt wieviel Abstraktion? Wohin mit den Funktionen im Objektmodell?
|
Measuring and Managing Performance in Organizations
One of the insights that comes early in a reading of Rob Austin's book, Measuring and Managing Performance in Organizations is that measurement is a potentially dangerous business. When you measure any indicator of performance, you incur a risk of worsening that performance. This is what Rob calls dysfunction. In order to see why this happens, you need to remember that measurement is almost always part of an effort to achieve some goal. You can't always measure all aspects of progress against the goal, so you settle for some surrogate parameter, one that seems to represent the goal closely and is simple enough to measure. So, for example, if the goal is long-term profitability, you may seek to achieve that goal by measuring and tracking productivity. What you're doing, in the abstract, is this:
When dysfunction occurs, the values of <parameter> go up comfortingly, but the values of <goal> get worse. You probably understood long ago that dysfunction was a possibility, but thought -- as we did -- that it was nothing more than a rare, freakish anomaly. But as Rob pursues the subject with persuasive thoroughness, it gradually begins to dawn on you that dysfunction is not an exception to the rule; it is the rule: Anything you measure is likely to exhibit at least some dysfunction. When you try to measure performance, particularly the performance of knowledge workers, you're positively courting dysfunction. Our first real understanding of dysfunction came from reading Rob's Ph.D. thesis. The thesis arrived at our offices over the transom (Tom had been interviewed for Rob's research project a year before, but did not even remember that when the thesis arrived), along with a pile of other hopeful contenders for our attention -- brochures, monographs, phone and credit card bills, and a few book manuscripts. We dutifully looked through the first few pages, and then read the rest avidly. Recorded below, in abbreviated form, is the sequence of reactions that each of us had:
We called Rob for permission to make a few copies and sent them out, first to our fellow Guild members, and then to several influential members of the measurement community. As calls and email messages came back, we found that others were having the same response: "At first, I wondered why you wanted me to spend time on this huge pile of paper; by about page 30, I began to understand. . . ." When you realize that dysfunction will probably accompany almost any kind of measurement, you're inclined to ask questions like, Why and when is it likely to occur? What are the underlying causes? What are the indicators that it is happening? and, most of all, What can I do about it? Satisfying answers to these and other allied questions were provided by Rob's thesis, but by no other source that we knew of. That made us believe that the work needed to be made available in some more accessible form. We began to encourage and cajole Rob to develop his work into a book. Measuring and Managing Performance in Organizations is the admirable result. We believe this is a book that needs to be on the desk of just about anyone who manages anything.
June 1996: Tom DeMarco (Camden, Maine) & Timothy Lister (New York) Copyright © 1996 by Dorset House Publishing Co.
|
