Developing an incentive scheme for a project
Tel Aviv University
Motivational needs of individuals who work in project environments may differ from those of individuals working in different organizational environments; however, both managers and employees who are involved with research and development feel that incentive schemes should be developed for them as well (See references   ). Renwick and Lawler  found that lack of proper pay was ranked as a chief cause for dissatisfaction at work. This finding emphasizes that payment level is a significant factor affecting the amount and quality of employees’ efforts. The desire to earn more money may come from economic needs or from the individual equating monetary reward with the organization's recognition of his ability.
It is also well-known that employees can increase their productivity significantly if they receive a fair share of the added output. This was verified by a survey administered by Fein , who reported the average level of productivity jumped from 86.4 percent to 123.5 percent; that is, an increase of 42.9 percent was witnessed when an incentive scheme was installed. This indicates that although incentive schemes are not the only means of improving performance, they obviously can have a significant impact. It is important, however, to point out that most of the organizations that participated in Fein's survey were probably production departments in manufacturing and service organizations, as compared to high technology, research and development, and project oriented organizations.
Very few cases of using an incentive scheme in a project environment have been published, the one reported by Parkinson  being an exception.
The next section reviews the major stages in developing a wage incentive scheme; a definition of each stage is followed by a general discussion of implementation problems, ending with an implementation analysis in a project environment.
Stages in Developing an Incentive Scheme
Development of any incentive scheme requires the following stages:
1. Generating a relevant list of PCs (PCs = Performance Criteria, PC = Performance Criterion)
2. Choosing the preferred set
3. Measuring present performance
4. Assigning standards to the chosen criteria
5. Formulating the incentive equation
The following raises and discusses methodological issues in the implementation of the stages, as well as possible ways to overcome some of the methodological difficulties.
1. Generating a Relevant List
Since we are concerned with evaluating the performance of an organization, many PCs may be considered relevant. Examples of such criteria are cost per item, percent of defects, and turnover rate. Theoretically, one may argue that a long list of possible PCs may be generated for an organization, but for practical purposes we may find it difficult to generate a list of more than a few dozen relevant criteria.
An important issue in developing output oriented PCs is the generation of a PCs hierarchical system. First, this entails designing PCs for each elementary unit according to the Work Breakdown Structure (WBS) of the project. PCs are next generated at the upper levels by integrating some of these lower level units. However, as Martin  points out, it is important to make sure that there is only one set of a PCs systems which are interrelated one to the other without unnecessary duplication.
Since projects are generally one-time operations, they do not lend themselves as readily to the development of a PCs system. Woodward  identifies two complementary approaches which may be used for performance evaluation in this context; one is concerned with output and the other with the process.
The output oriented approach is based on measurement, using quantifiable inputs, outputs, and their derivatives. Typically, PCs belonging to this category relate to time, cost, and the quality of the project. Each PC is evaluated through comparison of the actual value to a designated standard, e.g., actual completion time (or cost) to the planned time, and the project's actual technological properties to its planned specifications. The importance of following the above PCs can be demonstrated by Baker and Fisher  who note that project expenditures that double the original budget are a common event or Goff , who supports this claim and stresses that many government projects reach expenditures of three times the original budget.
The process oriented approach stresses the importance and influence of the process throughout the project life cycle, and its successful completion. The work process refers to the activities and equipment required to convert input into outputs, taking into consideration the required resources, and the order and methods of performing the activities.
The characteristics of the process oriented approach depend to a large extent on behavioral dimemsions. Hence, performance assessment may be concentrated on an employee's evaluation procedure as it relates to attributes such as attitude, ability to communicate, accuracy, and creativity. Kerzner  presents different forms with possible attributes applicable to employees working in a project environment.
To sum up this section, we may conclude that output and process oriented PCs are both needed for a proper evaluation of the project.
2. Choosing the Preferred Set
Since a long list of PCs can be generated, questions arise as to how many PCs to include in the scheme, and how to choose the PCs to be included.
There is no magic number of PCs that a scheme should contain; however, since each PC creates some workload, it is common practice to avoid introducing too many. For example, Galinka, Globerson and Oron  found, among organizations using incentive schemes, a maximum of seven PCs being used, with three the most frequent number cited.
Although any selection process is based on the assumption that management has implicit knowledge of the best PCs to be selected, it is still beneficial to call management's attention to several selection guidelines. The following guidelines are based on the author's experience and on previous studies   , and are relevant to project environments:
• PCs must be derived from the company's objectives.
• PCs make possible the comparison to organizations that are in the same business.
• The purposes of the PCs must be clear.
• Data collection and calculation methods of the PCs must be clearly defined.
• Ratio PCs are preferred to absolute ones.
• PCs should be under the control of the evaluated organizational unit.
• PCs should be selected through discussion with the people involved.
• Objective PCs are preferred to subjective ones.
• Selected PCs must include productivity PCs (ratio of output to input), quality (nature of output), and process oriented PCs.
Although objective PCs are preferred to subjective ones, there is no choice but to use subjective PCs if the process oriented approach is desired. Results of the case study presented later in this article confirm that employees prefer to have a subjective PCs system to no PCs system at all. More relevant subjective PCs may be related to the activities which have to be performed throughout the project life cycle; Roberts and Fusfeld  identify six typical innovation project stages, each of which needs different behavioral attributes in order to be completed successfully. The six stages are: 1) pre-project; 2) project possibilities; 3) project initiation; 4) project execution; 5) project outcome evaluation; and 6) project transfer. These six stages need direct inputs from the following five critical behavioral functions: (1) idea generating; (2) entrepreneuring; (3) project leading and coordinating; (4) gatekeeping (information gathering and channeling); and (5) coaching less experienced personnel. These functions represent the various roles that must be carried out for the successful completion of innovation projects. It is important to recognize the functions among the different project stages, and to identify some relevant measures for each of them.
3. Measuring Present Performance
Present performance of each of the chosen PCs must be measured before standards, or satisfactory levels of performance, can be established. There are two reasons for this. First, if a PC proves to be difficult to measure either an alternative means of measuring must be employed or a more easily measured PC must be used. Second, a knowledge of the present level of performance for each of the chosen PCs must be identified, since this information has to support the decision concerning the desired standard for each criterion.
Although a proper performance measurement system is a very important managerial tool for effective project management, an organization must avoid performance measurement systems which are too detailed. Gerloff , who analyzed 108 projects, concluded that too tight and detailed project control may not be effective, and may actually lead to undesired changes in the project's objective.
Since process related PCs involve human attributes, there is a need to develop proper employee evalution routines and forms. Kerzner  points out that a major difficulty in evaluating employees working in a project environment is due to the fact that they usually have several bosses, depending on the number of projects in which they are involved. An employee evaluation usually involves a discussion between the supervisor and the subordinate, together with a written summary which is referred to by the name, “the evaluation form.” Examples of various types of project oriented evaluation forms are presented by Kerzner .
4. Assigning Standards to the Chosen Criteria
Standards are used to indicate the satisfactory level of performance, above which an incentive is paid. An appropriate standard has to be higher than the present average performance, but still be attainable if it is to encourage continuous improvement. Also, there must be simple and logical techniques by which standards can be revised.
Possible techniques for assigning standards to PCs include:
• Work measurement techniques: methods such as time study, predetermined time standards, and work sampling (See Mundel ). These techniques are used for determining standards for only one PC, mainly efficiency as expressed by the ratio of assigned time to actual time.
• Analysis of past organizational performance: data for each PC may show deviations of the value of the PC, and also trends that can be depicted by a learning curve. Based on this analysis a relevant standard may be assigned.
• Management by objectives approach: assignment of standards or objectives is based on different considerations such as needs, abilities, and availability of resources. The specified objectives are usually a result of a group analysis and discussion.
• Data comparison: between organizations with similar characteristics such as type of business, technology, or environment (e.g., insurance companies, textile factories). An organization may assign standards based on its own standing as compared to other organizations.
• Economic considerations: profit may be obtained if expenses do not exceed a certain level. This level (or levels), if the operation is broken down according to needed resources (manpower, energy, material, etc.), may be considered a standard or standards.
• Legal considerations: the law may state required performance levels for specific PCs. In this case, the standards have to meet those stated by law.
Since subjective estimates are probably very common in project standard setting, it is important to investigate the discrepancy between the standards as they were set by the people involved, and the actual performance level. A survey conducted by Baker and Fisher  shows that among 450 project budget planners, the main reasons for mistaken estimates of project cost are insufficient data, vague definition of project objectives, and lack of experienced personnel.
Another issue which may complicate the standard setting in a project environment is the frequent revisions which may occur in design or scheduling, and result in different time and cost structures. Standards have to be revised accordingly, and it becomes difficult to evaluate the performance to date on the basis of the previously valid standards. As standards for the output oriented PCs tended to be dynamic, and therefore harder to deal with, they should be supported by process oriented PCs which use behavioral attributes. The standards for the behavioral attributes may be based either on the average of all the employees, or on the previous performance of the evaluated employee.
5. Formulating the Incentive Equation
The incentive, which is the value added to the base salary, is a function of the value of all the PCs and their assigned standards. A very important part of incentive scheme design is the formulation of the equation used to calculate the incentive. Some examples of equations can be found in Globerson  and Fein .
The following is a simple example of an incentive scheme based on two PCs:
|where INC:||percent of incentive or the value added to the base salary|
|Eff:||the efficiency, which is the ratio of assigned time to actual time multiplied by 100|
|R:||the response time factor expressed as a function of the project length. A possible function is described in Figure 1|
Response Time Factor R as a Function of Project Length (months)
To generate the response time factor function, management has to answer the following questions:
• What is the critical project length (PL1) above which the incentive achieved by the efficiency factor has to be reduced? For example, let us assume that Eff = 120. The bonus, based only on efficiency, is 120 – 100 = 20%. If management decided that PL1 = 1.0, and the actual project length was longer than 1.0 month, the bonus paid would be less than 20%.
• Is management willing to pay more for faster completion than PL,? In this example, management decided not to pay an additional bonus for completing the project before PL1.
• What is the critical project length (PL2) after which a bonus will not be paid despite high efficiency? In this example, if PL2 = 6.0 months and the actual length is 7.5, nothing would be paid even though a 20% bonus would be indicated based on efficiency alone.
• What reduction function should be used between the point PL, at which there is no penalty, and PL2 at which there is maximum penalty (no bonus paid)? This function may be a linear or nonlinear function.
The example described above is relevant to output oriented PCs. The next section describes a case study of a process oriented incentive scheme based on the results of employees’ evaluation. A complete description of the study may be found in Shapira and Globerson ,
The Case Study
This section describes an existing individual incentive scheme, which is used by a project oriented research and development organization which develops electromechanical products. The organization employs about 500 persons, including scientists, engineers, technicians, and administrative staff. The incentive scheme is based on the results of employees’ evaluation with regard to chosen behavioral attributes. The scheme includes two subschemes, one for employees and one for managers, with a common part. Only the employees’ subscheme is reported in this section. A form (See Figure 2) is used for documenting the evaluation results.
Group 1 of PC. There are five general weighted performance criteria (weight is given in brackets):
1. Discipline at work (15)
2. Professional knowledge relevant to the last period's projects (30)
3. Human relations (15)
4. Production and quality of performance (20)
5. Dedication at work and responsibility (20)
A relative weight is assigned to each criterion so that the total weight equals 100. The weights of the criteria were assigned by a management committee, and the results were approved by the employees. Assignment of weights to the criteria was based on relative importance and the potential impact that they may have on the success of performing the job under evaluation.
Each criterion consists of 4 subcriteria (As can be seen in Figure 2). A 7-point scale for each sub-criterion is used for evaluation, where 4 is the lowest grade and 10 is the highest. The meaning of each grade is defined in Figure 2. The grade is assigned by the manager of the evaluated person.
The following equation is used for calculating the Group 1 efficiency factor for each employee k:
|PC(k, i, j):||grade of employee k on subcriterion j within crtierion i|
|W(i):||relative weight of criterion i (Wi= 100)|
Since each criterion may obtain a maximum grade of 40 points (4 subcriteria with a maximum of 10 points each), the value of the denominator is 40. Thus, GlEFF(k) will be within the range
For example, the numerical value of the efficiency factor of equation (2), when one uses the values circled in Figure 2, is:
(Note that the values circled represent the performance of a specific employee.)
Group 1 of Performance Criteria (for employees and managers)
Employee's Incentive Scheme. The following are the steps in calculating the employee's incentive:
a. Calculate GlEFF(k) for employee k by using equation (2).
b. Evaluate the employee's work load and assign a load factor (LF); See Figure 2.
c. Calculate the efficiency of employee k, EFF(k), by using the following equation:
Continuing the previous numerical example, in which G1EFF = 72.875, if the employee load factor is LF = 1.1, then
d. Calculate the bonus, BONUS (k), for employee k by using the following formula:
|M:||number of employees in that department|
|CAKE:||the amount of money assigned for bonuses in that department|
In order to find the employee's bonus, the EFF(i) for all the employees in that department has to be calculated, and the amount of the “CAKE” to be divided among all the employees has to be determined. For example, if CAKE = $15,000 and there are 6 employees in the department with EFF(i) as follows,
|EFF(l) = 115||EFF(2) = 120||EFF(3) = 80|
|EFF (4) = 60||EFF(5) = 95||EFF(6) = 110|
then the bonus for employee 3 is:
After using the incentive scheme for two years, the company decided to evaluate the impact on its employees and managers. The following conclusions were derived from questionnaires that the participants were requested to complete.
1. Participants believe that the scheme increases the manager's power significantly.
2. Although a majority of both employees and managers are in favor of continuing the scheme, the managers’ view is more favorable than the employees’. A possible explanation is that the incentive scheme benefits managers more than employees; managers and employees both get bonuses, but managers also have power since they decide on the reward level.
3. The evaluation form is considered to be a reliable source of information and it reflects actual achievements.
To conclude the case study it may be stated that it is possible to develop an incentive scheme even for research and development projects which may be considered more complicated then other types of projects, such as construction projects. The case study also points out that in spite of all the problems raised by the use of a subjective evaluation system, people prefer to work with an incentive scheme which is seriously presented and implemented, rather than work without such a scheme. Provided there is any kind of periodic evaluation procedure in the organization, an incentive scheme can be tied to it. Connecting the incentive scheme to the evaluation procedure assures, among other things, that the evaluation will be taken seriously by both employees and managers with all the consequences attached to it, such as specific feedback to the employee concerning potential areas for improvement.
Finally, a serious drawback is the difficulty that the company faces if it wants to evaluate the impact of introducing the scheme. It is difficult to compare the “before” and “after,” because the complete nature of the work changes continuously.
Wage incentive schemes can be applied to organizations dealing with nonrepetitive activities such as projects, but it remains difficult to evaluate the impact of such a scheme on performance. This is especially true if the PCs are derived only from the process oriented approach. Further research is needed in the area of investigating the relationship between behavioral patterns of employees involved in a project environment, and the level of performance expressed in input-output terms. Questions such as, “Is there a significant correlation between ‘positive’ changes in behavioral attributes and successful completion of the project?” have to be answered before managers can be convinced that the efforts are justified. In spite of the lack of evidence regarding the amount of improvement that a company can look forward to after introducing an incentive scheme, it is obvious that such a scheme increases the motivation of the employee to use a greater part of his potential for the company's needs. Consequently, a company may find that the benefits reaped by introducing a wage incentive scheme far outweigh any increases incurred in salary costs.
1. Baker, B., & Fisher, D., Cost Growth: Can It Be Controlled?, Project Management Quarterly, 1974, 5, 2-3.
2. Berczi, A., Improving Public Sector Management Performance, Management International Review, 1978, 8.
3. Dessaver, J.H., How a Large Corporation Motivates its R&D People, Research Management, 1974, 14.
4. Eilon, S., Some Useful Ratios in Evaluating Performance, Omega, 1979, 7.
5. Fein, M., Work Measurement and Wage Incentive, Industrial Engineering, 1973, 5.
6. Fein, M., Financial Incentives, in Handbook of Industrial Engineering, edited by Gariel Salvendy, New York: John Wiley and Sons, 1982.
7. Galinka, O., Globerson, S., Oron, N., The Use of Multiple Incentive Schemes, American Institute for Decision Sciences Conference Proceedings, San Antonio, 1983.
8. Gerloff, L.A., Performance Contol in Government R&D Projects, IEEE Transactions, 1973, EM-20.
9. Gerstenfeld, A., & Rosen, G., Why Engineers Transfer, Business Horizons, 1970, 13.
10. Globerson, S., Developing A Multiple Factor Incentive Plan, Industrial Engineering, 1982, 14.
11. Goff, N.S., Development Project Costs, Journal of Systems Management, 1975, VI.
12. Hitt, M., & Middlemist, R.D., A Methodology to Develop Criteria and Criteria Weightings for Assessing Subunit Effectiveness in Organizations, Academy of Management Journal, 1979, 22.
13. Hurst, G.E., Choosing Performance Measures, Department of Decision Sciences, The Wharton School, University of Pennsylvania, 77-11-01, 1977.
14. Kerzner, H., Evaluation Techniques in Project Management, Journal of Systems Management, 1980, 31, 10-19.
15. Martin, C.C., Project Management, American Management Association, 1976.
16. Martin, M.D., Lenz, J., & Glover, W., Uncertainty Analysis for Program Management, Project Management Quarterly, 1975, 3.
17. Mundel, M.E., Motion and Time Study, Englewood Cliffs, NJ: Prentice-Hall, 1978.
18. Noyce, R.N. (an interview with), Creativity by Numbers, Harvard Business Review, 1980, 58.
19. Parkinson, D., Productivity Incentive on Engineering Construction Sites, Cost Engineering, 1979, 21.
20. Renwick, P.A., & Lawler, E.E., What You Really Want From Your Job, Psychology Today, 1978.
21. Roberts, E.B., & Fusfeld, A.R., Staffing Innovative Technology-Based Organization, Sloan Management Review, 1981, 22.
22. Shapira, R., & Globerson, S., Incentive Scheme for a Research and Development Organization, Research Management, in press.
23. Woodword, S.W., Performance in Planning A Large Project, Journal of Management Studies, 1982, 19.
Dr. Shlomo Globerson is a faculty member of the Graduate School of Business Administration, Tel Aviv University. At present, he is a visiting professor at Northeastern University. His current research and consultation work is in developing performance criteria systems and bonus schemes for project oriented organizations.
I would like to thank the following individuals for their assistance and response to my requests (June and September PMQs) for back issues of the Project Mangement Quarterly:
Through their efforts and others identified previously, a complete set of original PMQs has been developed.
Terry L. Kinnear