A Consultant's Casebook
Full Circle Again
Many of today's jobs, like Army recruiting reflect the cumulative effects of many people working alone. This observation is disturbing. We know, for example, that there is a multiplication of effort when people work together as a cohesive team. Yet, recruiting has not been structured to have people working together. For this very reason, command within this organizational structure is difficult. It is difficult to discipline a unit when there are examples of exemplary performance as well as sub-standard performance within the same unit. (When behavior is organized and measured at the individual level, it should be commended and disciplined at the individual level.)
Worse, the one-man-show organization of recruiting, like many other jobs, is likely to reduce efficiency. Recall from our experimentation with the simulation that two factors have dramatic impact on contract achievement: (1) selling time and (2) enlistment processing time. If recruiters were organized to work in parallel on the same pools of prospects and applicants, it would have the effect of reducing selling time and processing time. The recruiting process, however, is organized for serial processing, not parallel processing. We anticipate that a re-organization of the work in recruiting stations would result in higher productivity. We emphasize that changes in work organization need not be done experimentally. Instead, changes can be examined through simulation models such as the one described here.
Finally, our studies suggest the need to re-think our capabilities of changing system performance from above when work is organized in the fashion we describe. Our model of recruiting does not have obvious leverage points where leadership can be brought to bear to influence recruiter performance. The model has a negligible interface with the officer corps (and management). If there is no vehicle by which managers can affect the system, it is important to reconsider how the work is organized. Again, experimentation on management organization is best left to computer simulation instead of field experiments.
What began as a simple study of factors that affected an observed drop in young persons' propensity to enlist ended up as an explanation of how system change is reflected in performance. In my lectures on statistics, I frequently emphasize how "N of 1 correlations" can be misinterpreted. In fact, General Wheeler relied on the simple co-occurrence of phenomena to conclude that they are related. Human behavior and performance in organizations is rarely a function of a single detail. It is important to understand behavior in context before we commit time, resources, and vast amounts of money to change them. From the General's point of view, increasing resources for advertising would have been desirable. In fact, increasing resources for hiring more recruiters, or re-organizing work procedures would have been more effective.
For an especially dramatic illustration, see The Face of Battle, by John Keegan, Penguin, New York, 198?. Keegan suggests that in battle, small cohesive units can withstand almost any odds, yet become completely vulnerable to weak assault when their cohesion disintegrates.
All material is copyrighted! Not to be used without the written permission of Andres Inn. ©2000.
Send mail to Andres.Inn@mail.ee with questions or comments about this web site.