New directions for management control of project plans

Estech, Incorporated

Elmer H. Burack

Illinois Institute of Technology

Just recently, our economic problems seemed almost insurmountable. We were still deep in the throes of double digit inflation, unemployment was climbing at an ever increasing pace. The accumulation of petrodollars weighed heavily on our minds. Raw materials were in short supply because of new cartels and rising prices of suppliers.

It is for these reaons that United States industry stands at the threshold of a new era. Perhaps never before has it had, at the same time, so many problems and such bright prospects for the future. Excellent business planning and management will be required to solve the problems and realize the prospects. Actions taken by industry during the next two to three years will set the course of events for the next decade. For some, imminent decisions will determine survival. For others, the quality of planning and decisionmaking will set the stage for unprecedented growth — or for stagnation and decline. The confluence of events has placed and will continue to place unprecedented demands on strategic planners.

Nevertheless, industry’s organizational planners commonly submit proposals that cannot be adequately funded by set budgets even after a careful examination of planned projects. The results is that these proposals are funded, and as development begins, some projects start to overrun their cost estimates. Constraints then have to be imposed on other projects in order to meet overall program budgets.

The problem is not so much the actual cost of the programs, but the inability to adhere to the cost and schedule estimates, or put another way, the inability to predict future program deviations early enough and with sufficient confidence to do something about them. The primary problem, therefore, is to recognize quickly when significant deviations are occurring in a project so that overall control of the program can still be exercised. In addition, management should be informed as to the degree and origin of the uncertainty and complexity accompanying project plans before implementation.

The Project Planning Continuum

The purpose of this paper is to discuss briefly the application of a new management tool called the project planning continuum. The project planning continuum is composed of two major elements (See Figure 1). The first element is theUncertainty and Complexity Model (U/C), which identifies and translates the influence of people, the environment, and organizational characteristics into uncertainty and complexity measures. These are then associated with major elements comprising the plan. Uncertainty (U) is defined as the predictability or likelihood that a particular event or occurrence will happen; whereas complexity (C) is defined as the processes, interrelationships, activities, and organizational units that have to interreact to obtain specified results. This model allows management to audit a proposed project plan in terms of the uncertainty and complexity surrounding its major elements before its implementation.

The second major element of the project planning continuum is a series of models, called Resource Appraisement Models (RAM), which monitor and predict strategic turning pionts of an ongoing project plan (with its associated uncertainty and complexity levels) each reporting period, using the same timephasing of project events as in the original plan but with actual reported data. The models anticipate the future levels of expenditures, as each month’s reported expenditures are added to the data base.

FIGURE 1
Graphic Portrayal of the Project Planning Continuum

The balance of this paper is devoted to a brief explanation of how the uncertainty and complexity model and RAMs were constructed, coupled with the results of administering the models to one corporate plan.

The Uncertainty and Complexity Process

Figure 1 shows schematically the U/C model and the way individual job roles are related to work carried out in the development of the project plan, paying particular attention to whether one is a member of the staff who developed the plan or of the management team who had appraised it and suggested revisions. Once these organizational groups have been identified, the modeling process starts by administering Creighton’s (1972) professional preference questionnaire (PPQ) to the management group, in order to categorize individuals as opinion leaders, mid-influencers, or stabilizers.

An opinion leader is defined as a strategic intermediary between the source of knowledge and the application of knowledge. He is an individual from whom others seek information and advice, and he is able to influence individuals in a desired way with some degree of frequency. Stabilizers are individuals who perform in the opposite polar position of opinion leaders.

After the management group is categorized, the modeling process continues by identifying key environmental elements comprising the plan and then constructing an Uncertainty and Complexity Profile Questionnaire (UCPQ) to be administered to both the management and planning teams. The elements of the UCPQ are ascertained by means of semistructured and unstructured interviews with participating individuals who comprised each team. After the UCPQ has been constructed, it is administered to both management and planning specialists, using the Delphi Technique.

Last, it is necessary to measure the amount of influence exerted on the planning specialists. This is done through the creation of an Influence Profile Questionnaire (IPQ) which is administered to the planning specialists. In their analysis of the questions shown in the UCPQ, the planners are asked to rate the amount of adverse influence from organization activities they feel or perceive. Since individual management members have direct influence in each of the questions shown on the IPQ, the authors are indirectly asking the planning specialists to rank management members according to their influence. After all these data have been accumulated, the uncertainty and complexity measures are then computed. The results of applying the U/C model follow.

Project Plan X (PPX)

The history of Company X spans and encompasses the history of the modern container industry. The corporation (over 12,000 employees) traces its history to the turn of the century through the convolution of acquisitions and mergers. During the last decade, Company X began to expand its business outside the container market into the field of edible products. It was suggested that management consider developing a specialty production operation, and this appeared to be a reasonable objective. However, some management personnel felt that this venture was not intended for the fainthearted and might not be correct for Company X because key executives, who would be measuring actions and results of the development team, lacked knowledge about the new business.

Management did recognize that manpower to design and develop such a plan would be hard to find and that Company X was entering this business from scratch. A study team was assembled when another inexperienced company built an uneconomical specialty production facility in the midwest.

At this point everything pointed up; however, management had failed to recognize that the examination team did not have sufficient expertise for such an analysis. Nevertheless, surveys of the existing facility continued and capital investment of $1.6 million was estimated to overhaul it. It was felt that the refitting could be done in less than a year, with production starting soon after. Based on this recommendation, the facility was acquired and refitting was started.

When the plans for the plant were originally offered, all costs were predicted on then available second hand equipment. A year later, when it was actually required, some of this equipment was not available and other pieces did not meet new laws and regulations, so equipment had to be purchased new rather than used. Also, due to the recently developed fuel shortage, PPX had to add a gas dryer to assure uninterrupted production in case the utility cuts off the regular fuel supply.

Labor rates of $6.50 per hour were used in the engineers’ computations, based on local practice. After a nonunion laborer was electrocuted on the premises, a steel contractor brought in union help; PPX was then forced into the same practice by all contractors, with the labor rates increasing to approximately $15.00 per hour.

The original plans for the site did not contemplate any locker room or lunch room space and had assumed the use of then planned central lab facilities for minimum required testing. Both of these assumptions changed, and a new design had to be prepared with cost estimates based on new specifications to allow for the additional facilities.

Top management, however, was not aware of what was happening. Reviews reported everything going fine, and even periodic visits by management did not uncover the seriousness of the situation. Consultants were covertly brought in by the development team in order to help out. The authors interviewed the consultants who said they were shocked at the state of affairs. They suggested that the PPX planning group did not analyze the operation in the usual manner. The consultants said that generally the process is laid out with equipment drawings and specifications. Next, a cost forecast of the necessary equipment is made and cost estimates of the labor and bids for construction are obtained. The consultants said that in this case everything was done in reverse order. For example, building construction had taken place before it was known if the equipment could be utilized in the space built.

One year after the start of the operation some members of management were initially alerted to the problems; however, even these people believed that the apparent cost overruns would balance out against projected savings. This belief that everything would balance out was still held 20 months after the start of the operation, even though the original planned expenditures were overrun and the operation was nowhere, near finished.

Approximately two years after the start of PPX, the true seriousness of the situation was recognized and relations between members of the management team became strained. Periodic confessional reviews of PPX became unbearable exercises in futility. Members of the management team started to avoid each other, with internal communications becoming less verbal and friendly. Correspondence between management members over the course of this project verified what was said to the authors in private. Finally, the actual amount of capital expenditures was over $3.5 million.

It was in this environment that the authors started the proposed field work. Each member of the development and management teams was asked, to the best of his ability, to place himself back at the time of the original development of PPX and to fill out his questionnaire accordingly. It was recognized that some bias would enter, but it was felt that the group response, utilizing the Delphi technique, would minimize this bias.

UC Findings for PPX

Eleven management individuals were selected for participation in this effort. These individuals were directly associated with PPX and had influence in the project from its early beginnings. The titles of the individuals ranged from the chairman of the board through the sales manager.

Each individual was administered both the Professional Preference Questionnaire (PPQ), and the Uncertainty and Complexity Profile Questionnaire (UCPQ), and a Personnel Profile was formulated. Planning specialists were adminstered the same UCPQ along with the Influence Profile Questionnaire.

Table 1 exhibits the U/C model findings. The final uncertainty and complexity measures associated with the planned environmental units range from a high of 35 to a low of zero. A response of zero indicates a very simple and routine task to accomplish; whereas, a measure of 9 suggests that there will be some difficulty in accomplishing the task. A rating of 18 signals that some sophistication and technology will be needed to meet these goals. However, a rating of 26 implies that even with sophistication and technology being applied, this task will ultimately have problems; therefore, contingencies beyond the scope of the task should be planned for. A 35 measure signals that the planned goals, even with total corporate support and expanded contingencies, will not be met.

Table 1 indicates that more uncertainty and complexity were associated with categories of cost, vendors, scheduling, and information flow than shown in management’s original UCPQ. In addition, it also shows that in some sections of the life cycle other elements were of importance (for example, there was a large complexity rating given equipment purchase and installation during period two). The last step in the process is to add all columns of the U/C covering in order to compute overall uncertainty and complexity measures.

Looking at these measures more closely, one can see that management and planning personnel were 70 percent less sure of what was going to happen in period two than in period one. The U/C model approach has shown that there was a high degree of uncertainty and complexity associated with elements of cost, vendors, scheduling, and information flow. And after final examination, these were exactly the areas which caused the greatest problems in PPX. In addition, these U/C measures are used as inputs in the RAM models.

Resource Appraisement Models

Experts emphasize that it has been a difficult problem for management to recognize and anticipate significant variances in a planned organizational activity. Recently, exponential smoothing has been developed for forecasting future trends.

The main difficulty in applying exponential and adaptive smoothing approaches to forecasting has been the inability of these techniques to simulate all functional demand patterns. Since project plans may take any variety of functional contours, a general procedure to match these infinite patterns had to be developed to simulate these functions. These techniques, called Resource Appraisement Models (RAM), differ significantly from the previous models in that (a) they are created from piecewise linear segments and are able to match any contour, (b) the smoothing constant is a function of the uncertainty and complexity measures and not the forecasting error, (c) the deviations between the plan and actual, rather than the smoothing constant, are used to predict turning points in the project plan. A discussion of each model is found in Appendix I.

TABLE 1 Uncertainty/Complexity Life Cycle Covering for PPX

Major Elements Contracting/ Construction Equipment Purchasing/ Installation Major Wiring Test Phase/Manufacturing
U C U C U C U C
Cost guidelines 12 18 17 20 14 12 11 10
Quality and reliability 3 3 9 6 8 3 8 4
Vendors 6 5 16 12 13 12 13 10
Equipment and installation 7 13 11 13 10 8 10 7
Scheduling 13 19 20 25 21 19 19 17
State of the art 8 13 7 13 5 6 4 8
Information flow 7 7 16 14 16 11 16 9
Resource evaluation 3 5 18 18 11 8 9 9
Total 59 83 114 121 98 79 90 74
U/C combined and
   normalized 1 1.7 1.3 1.2

FIGURE 3 RAM Predicts Dollar Expenditure for PPX

FIGURE 3 RAM Predicts Cumulative Work Accomplished for PPX

RAM Analysis of PPX

The plan expenditure was $1.6 million spread across a 12-period life cycle covering design, development and beginning manufacturing. However, PPX took approximately two and one half years to complete, with expenditures over $3.5 million. Management members didn’t begin to recognize the seriousness of the situation until approximately 18 months after the start of the operation, and only after the original planning expenditures were spent.

Actual expenditures were reported by period from the start of the project. Figures 2 and 3 show the original planned cost, the cumulative work accomplished (percent expenditures), the actual reported expenditures and the RAM I predictions. Table 2 reports, in the first three columns, the time history of the project (that is, the periods), the original planned cost, and the reported actual expenditures. Table 3 presents the original planned costs and actual reported costs per period as percentages of work accomplished.

TABLE 2 Project Plan X — RAM I Dollar Expenditure Predictions ($000)

Project Plan X — RAM I Dollar Expenditure Predictions ($000)

TABLE 3 RAM I Summary Output for PPX Work Accomplished (%)

RAM I Summary Output for PPX Work Accomplished (%)

From the updated plan, anticipated expenditures for one month ahead, two months ahead, three months ahead, and for the total program are exhibited. These are shown, for any month, in columns four, five, and six of Table 2 as the anticipated expenditures that were computed one, two and three months ago. Alongside each prediction is the accuracy which finally resulted — the ratio of the anticipated and the actual expenditures. Thus (1.00) means that the anticipated expenditures were 1.00 of what was actually reported. This ratio is used later in the analysis as a measure of confidence in the RAM I computed plan.

The seventh column of Table 2 gives the anticipated total project forecast each month, while column eight gives the uncertainty for the total program. This is computed by keeping a running average, starting in period j, on the forecasting accuracy over the last three reporting periods and then applying this average to the plan forecast. Column nine gives the ratio of the anticipated total to the originally planned total expenditure. All dollar expenditures for PPX are expressed in thousands unless specifically labeled. The uncertainty measures developed for both plans were put into the models; an analysis of RAM output follows.

First Period — The reported dollar expenditure was exactly as stated in the original project plan. The actual work accomplished was reported at 1.2 percent compared with the planned value of 10.0 percent. The total work accomplished, at the end of the project period, is forecasted at 42 percent.

Third Period — The reported dollar expenditure is very low at 60 versus 130, even though this is 109 percent of what was predicted only one period ago. RAM I now predicts $95,000, $49,000 and $128,000 for periods 4, 5 and 6. A total expenditure of $1.2 million is anticipated, which is 73 percent of planned costs. This RAM output without any other information might indicate that PPX is being developed at less than plan costs. But the 0.8 percent actual reported work accomplished is far less than the original plan value of 7.9 percent, and RAM anticipates at least a 54 percent underrun with the possibility of only 18.2 percent of the work being finished at the end of the 12- period horizon. To the management and planning specialists, it is very clear that little, if any, relationship exists between the original plan and the actual performance, since RAM predicts 73 percent of the money will be spent for only 46 percent of the work to be accomplished at the end of 12-periods. Also, the accuracy of the previous two periods’ predictions for work accomplished was only 56 percent and 29 percent. RAM’s output does indicate that a project review was necessary, even though the expenditures to date were still only 18.2 percent of that budgeted.

Conclusions

This research has presented a new tool for management described as the U/C continuum which provides an important complement to the experience and judgment of the practitioner. The models not only seek to expand the insight of the management and planning specialist into the environmental factors which influence organizational management or system procedures, but also provide an early warning system when a corporate plan is deviating from preset patterns. The strength of the approaches is found in their ability to identify the relevant environmental factors, the interpretation of their uncertainty and complexity characteristics, and the ability to highlight serious deviations between the project plan and actual performance.

The U/C continuum displays a wide range of versatility in points of application, but like any other tool, it is no better than the skill of the management employing it. Because the organizational technology underlying the use of this system is still in a developmental stage, considerable adjustments or refinement should be expected as a new research comes to light.

REFERENCES

1. Brown, R. G. Smoothing Forecasting and Prediction of Discrete Time Series, (Englewood Cliffs, N.J.: Prentice-Hall, 1963).

2. Burack, Elmer H. Prospectus on Matrix Management: The U-C Technique, (Paper presented at the annual meeting of the Academy of Management, 1973).

3. Burack, Elmer H., and H. Zia Hassan. A Newer Managerial Tool: U-C Analysis, (Paper in process).

4. Chambers, O. C., S. K. Mullick, and D. Smith. “How to Choose the Right Forecasting Techniques,” Harvard Business Review, Vol. 49, No. 4 (1971), 45-74.

5. Chow, Wen M. “Adaptive Control of the Exponential Smoothing Constant,” The Journal of Industrial Engineering, Vol. 16 (1965), 314-317.

6. Creight, J. W. Enhancement of Research and Development Output Utilization Efficiencies; Linker Concept Methodology in the Technology Transfer Process, (Monterey, Calif.: Naval Postgraduate School, 1972).

7. Duncan, Robert B. “Characteristics of Organizational Environment and Environmental Uncertainty,” Administrative Science Quarterly, Vol. 17 (1972), 313-327.

8. Eilon, S., and J. Elmaleh. “Adaptive Limits in Inventory Control,” Management Science, Vol. 16 (1970), 533-548.

9. Galbraith, Jay. Organization Design: An Informative Processing View, MIT Paper No. 425, (Cambridge, Mass.: MIT, 1969).

10. Lawrence, Paul R., and Jay W. Lorsch. Organization and Environment, (Cambridge, Mass.: Harvard Business School, Division of Research, 1967).

11. Souder, William. “Experience with an R & D Project Control Model,” Monsanto IEEE, Vol. FM-15, No. 1 (1968), 39-49.

12. Souder, William. “Budgeting for R & D — A Case for Management Science Method,” Business Horizons, Vol. 13, No. 3 (1970) 31-38.

13.Vancil, Richard F. “The Accuracy of Long Range Planning,” Harvard Business Review, Vol. 48, No. 5 (1970), 98-101.

14.Whybark, D. Clay. Testing an Adaptive Inventory Control Model, Institute Paper, No. 289, (Lafayette, Ind.: Purdue University, Krannert School, 1970).

15.Whybark, D. Clay. A Comparison of Adaptive Forecasting Techniques, Institute Paper, No. 302, (Lafayette, Ind.: Purdue University, Krannert School, 1971).

16.Woodward, Joan. Industrial Organization: Theory and Practice, (London: Oxford University Press, 1965).

APPENDIX I

RAM I Unconstrained*

The purpose of RAM I is to modify the overall RAP, using the same functional RAP relationships coupled with the life cycle uncertainty and complexity measures, but predicting on the basis of the latest data with the deviation. If the integrated deviations predicted for the whole plan are then large enough to cause significant impact on the overall program (sum of many projects), then the plan is reviewed. RAM I follows below:

img

RAM II Solid Constraint*

The particular use of RAM II is concerned with holding RAP to its initial budget and work accomplished limits. In other words, this is a smoothing out of the RAP from the last reported actual to the end of project using the RAP life cycle as a guideline. RAM II follows below:


*For a more detailed discussion of RAM I and RAM II see the American Association of Cost Engineers Bulletin, February and April 1976, Volume 18, Numbers 1 and 2.

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI.

Advertisement

Advertisement

Related Content

Advertisement