OF ATOS ORIGIN
Master thesis, MscBA, specialization Human Resource Management University of Groningen, Faculty of Economics and Business
December 8, 2009 MARLIES KOOPMAN Student number: 1734792 Berkelstraat 73 9725 GW Groningen Tel: +31 (0)6 28701783 e-mail: M.Koopman.5@student.rug.nl Supervisor – university
P.H. van der Meer J. van Polen
Supervisor – field of study R. van der Hoek
TABLE OF CONTENTS
ABSTRACT
1. INTRODUCTION………. 4
2. THEORETICAL FRAMEWORK……….…… 6
2.1 Context……….… 6
2.2 Performance management elements……….… 11
2.3 Incidental circumstances……….…. 16
2.4 Performance management outcomes………... 17
2.5 Recapitulation……….…. 19
3. RESEARCH METHODOLOGY………. 19
3.1 Participants……….. 19
3.2 Procedure………. 20
3.3 Measures and data analyses………. 21
4. RESULTS………. 21
4.1 Context……… 22
4.2 Performance management elements……… 25
4.3 Incidental circumstances………. 27
4.4 Performance management outcomes……….. 28
5. DISCUSSION……….. 30
5.1 Context……… 30
5.2 Performance management elements……… 32
5.3 Incidental circumstances………. 33
5.4 Performance management outcomes……….. 34
5.5 Research limitation………. 35
5.6 Conclusion……….. 35
REFERENCES……… 37
Appendix A, Organization chart Atos Origin Nederland B.V. MO IBS……….. 41
Appendix B, Framework: Requirements for good performance management…………. 42
ABSTRACT
1. INTRODUCTION
The increased competition, governmental pressure for the public sector, organizational restructuring with an emphasis upon decentralization of accountability, and an intent to reshape industrial relations so that individualization displaces collective arrangements have caused an increased attention to manage performance (Storey & Sisson, 1993). Also for Atos Origin, “an international information technology services company, whose business is turning client vision into results through the application of consulting, systems integration and managed operations”1, performance management has become more important in the last years. This thesis describes a research, concerning performance management, which is accomplished for this company.
Performance can be described as an output view or as a behavioral view. Today the behavioral view is more common (Williams, 2002). Murphy (1990 cited Williams 2002, p. 93) gives the following definition for the behavioral view of performance: “The set of behaviors that are relevant to the goals of the organization or the organizational unit in which a person works”. Managing performance has been recognized already for a long time ago (Williams, 2002). Over the years, a lot of different practices, techniques, tools, systems, and philosophies have passed. The term performance management came to a particular prominence in the late 1980s/ early 1990s. Performance management can be defined as “a systematic process for improving organizational performance by developing the performance of individuals and teams” (Armstrong, 2006). Another description of performance management, which suits with Armstrong’s definition, comes from Lebas (1995). With his description he focuses on the future, about the capability of someone being evaluated. In his opinion the purpose of management is about creating and shaping the future of the organization. He also describes in his article that performance management is partly being carried out by performance measurement, and that management could hardly exist without measurement. According to Lebas, measuring means “transforming a complex reality into a sequence of limited symbols that can be communicated and that can be reproduced under similar circumstances”. However, a contradiction arises here; measures can only be about the past, data about the past are used to evaluate the future (Lebas, 1995).
When development of employees and organizational improvement are the main goals of performance management, making use of a performance management system (PMS) is useful (Heathfield, 2009). A PMS should provide employees four basic benefits: a clear understanding
of job expectations, regular feedback about performance, advice and steps for improving performance, and rewards for good performance (All Business, 2009). Atos Origin works with a PMS, the performance management framework (PMF). The PMF has been implemented in 2007 as a tool for performance management and is a global, online performance management process ensuring a consistent standard of managing performance throughout the company. Besides judging performance, the PMF takes care of continuously developing and improving employee skills. Skepticism has emerged among the employees of the organization about the process around the PMF. The question has emerged if it does meet the requirements of an adequate PMS and if it works in the organization like it should. Consequently, the aim of this research is to find out which aspects are important for a good performance management process and to compare this with the situation of Atos Origin. Though, this research concerns not the applicability of the tool itself but the whole process which is included in performance management.
While Atos Origin is active in forty countries, this research is focused on the Dutch organization; Atos Origin Nederland B.V. (hereafter referred to as Atos Origin). Atos Origin exists of three units; Managed Operations (MO), System Integration (SI), and Atos Consulting (AC). In this research the sub-unit Infrastructure Business Solutions (IBS), which is part of MO, will be explored (the organization chart in Appendix A, shows the structure of MO in which IBS is outlined).
Hence, this research will evaluate the performance management process of a part of Atos Origin; IBS. The following research question is developed:
Which elements of the performance management process of the Atos Origin department IBS should be improved?
To answer this question, first two sub-questions need to be answered which are;
• what are the requirements for a good performance management process, and; • to what extent does Atos Origin – IBS meet these requirements?
describes how this new framework is used to carry out the research for answering the second sub-question. The research is conducted by interviewing managers and employees and using company documentation. The outcomes of this research will be described in the results section and, finally, discussed. The discussion section will conclude with the answer on the research question.
The framework which is used for this research is customized for Atos Origin, however, with some adjustments it is also applicable for other organizations. In this way, it is possible to generalize this method of research for many organizations to contribute to improvements of their PMS.
2. THEORETICAL FRAMEWORK
Spangenberg’s framework (1994) is divided in four categories, which are context, system characteristics, performance appraisal elements, and performance appraisal outcomes. For each category he gives several sub-categories which describe circumstances which can occur in the organization and can cause problems for performance management. These subjects have been a good inspiration to be a basis for this research. With additional literature, Spangenberg’s framework has been transformed into a framework which is more specific for Atos Origin. This new framework also consists of four categories which are context, performance management elements, incidental circumstances, and performance management outcomes. Besides that, instead of using sources which can occur, this new framework gives requirements for a good PMS; all four categories consist of several requirements.
This section will describe the theory of each category and all requirements which derive from it. Finally, the new resulting framework will be showed.
2.1 Context
First it is important to more accurately define performance management itself and how it should fit into the characteristics of the organization. Therefore, the first category of the framework exists of the context of the performance management process and the context of the organization itself.
Performance management characteristics. According to Williams (2002), there are three
This model shows that performance management consists of performance planning, performance improvement, and performance review. In the case of performance management as a system for managing employee performance, Williams argues, it is commonly presented as a cycle which he clarifies with an example of the model of Ainsworth and Smith (1993 cited William 2002, p. 13). This model is shown in Figure 1.
FIGURE 1
Cycle of performance management as a system for managing employee performance
Performance management as a system for integrating the management of organizational and employee performance is sort of a combination of the first two perspectives.
FIGURE 2
Nevertheless, renewing objectives periodically and linking those to PRP has a possible demotivating influence on the performance of employees, which can be explained as the ratchet-effect. More and more will be expected from an employee by renewing the objectives; rewards and incentives diminish when employees keep responding with additional effort (Gibbons, 1987).
Advantages of relating performance evaluation with payments are that it attracts employees with high abilities and encourages them to perform as high as possible (Booth & Frank, 1999) Furthermore, it focuses effort where the organization wants it; it strengthens the performance planning process and; it rewards the right people (Brading and Wright 1990 cited Storey & Sisson 1993, p. 140).
Besides these advantages, Storey and Sisson (1993) also sum some disadvantages concerning PRP. When giving a lot of attention to PRP, other important aspects of performance management can be forgotten and will not receive the needed attention. Furthermore, the introduction of PRP can be a problem; establishing formal performance criteria leaves a great deal to be desired. Third, PRP can be subjective and inconsistent, which is most of the time caused by a lack of attention to the training of managers how to carry out the performance evaluation. Finally, PRP would not fit with certain other policies and objectives like the importance of teamwork. In cases, especially where operations are inter-linked, individual PRP would appear to be totally inappropriate. In that case individual performance can undermine team spirit and cooperation, workers will focus more on their individual targets instead of the performance of the unit (Storey & Sisson, 1993).
Nevertheless, a reward system has the strength of attracting and retaining employees of the required quality, to underpin the drive to improve performance, and to support the ability to change (Collins 1991 cited Storey & Sisson 1993, p. 144). In addition, Williams describes that, according to surveys summarized in Brown and Armstrong (1999 cited Williams 2002, p. 183), personnel and HR managers believe that PRP has a positive influence on improving employee performance. Williams argues that these managers also believe that PRP is improving for organizational performance when it is used as part of some broader performance management policy (Williams, 2002).
When applying PRP, two main types can be explored. The first one involves linking PRP as measured by the achievement of specific individual objectives, the other one assesses performance in terms of certain behavioral traits such as problem-solving, reliability, initiative, cooperation, and so on.
Organizational characteristics. The strategy an organization decides to pursue is a central
strategy is a definition of the direction in which an organization wants to go to achieve the mission; strategies are about vision and direction (Armstrong, 2006). Armstrong describes that the strategies may be set out in different headings. One of those headings, which is important for performance management, is the human resources strategy (HR-strategy). Armstrong defines the HR-strategy as ‘the acquisition, motivation and development of the human resources required by the organization and the steps required to create commitment and constructive and cooperative relationships with employees’. Boxall and Purcell (2008) argue that it is not assumable that HR strategies are uniform within firms. Mostly it is not a single set of critical practices for managing work and people, “it is better to think of HR strategy as a cluster of HR systems” (Boxall & Purcell, 2008). Usually, an organization has more overlapping HR practices which should result in legitimacy and take care of internal political pressures.
Besides the strategy of an organization, the structure of an organization is important. Mintzberg (1979) and Bakker, Paauwe, and Immerzeel (1989) describe five different types of organizational structures which all have a different effect on performance management. First they mention the straightforward structure, which is about a scarce division of labor and a low developed organizational structure where supervision takes place in a direct form. Often the final product is visible for the employees and the relation with work and the company is strong. Like the structure, also performance management is simple. It has some authoritative and paternalistic characteristics because most of its contents are decided by the director. Terms of employment are strong pronounced, while education, job evaluations, and job appraisals are limited.
The second structure is the machine bureaucracy. The most import issue with this structure is standardization of work processes which are instructed by staff departments. Job activities are most of the time a routine and are accomplished according standards and procedures. Performance management is guided strongly by standards and procedures, all personal activities are described precisely.
The professional bureaucracy is the third structure according to Mintzberg and to Bakker and others. In this structure not the job is standardized but the required skills, which makes selection, training, and education more important. Work activities are adjusted by standardized desired skills. When an employee has the desired skills he or she has a lot of autonomy in the organization.
the company is to decide what the desired output is. The division structure is an arrangement of different work units, therefore, terms which are common in this structure are unit-, result-, and contract management. Performance management exists of extraordinary demands. Because success of the organization with this structure is strongly dependent on quality, flexibility, and cooperation of the employees, employees are a strategic factor. A conflict of interest arises here for performance management. On the one hand, it has to be very decentralized to enable it to adapt easily to specific circumstances of the different work units. On the other hand, it is essential to build a strong shared identity to attract and bind employees.
2.2 Performance management elements
The following section describes the main characteristics of performance management itself and how this is part of the organization. This second category of the framework will elaborate on requirements about important aspects of performance management and on issues which can go wrong.
Important issues of performance management. Who are the players in performance
management? Williams (2002) claims that both the line-manager and the employee are responsible for the performance review. Possible sources to measure the performance of employees can be managers, peers, employees themselves, and customers. The most frequently used source are managers because with their knowledge about the requirements of a job they have the ability to rate, they have the opportunities to observe, and they are motivated to rate their employees because it is their concern how their employees are doing their job (Heneman, Wexley, & Moore, 1987). Nevertheless, also some problems can occur with the manager as rater. In some jobs, an employee can work out of sight of the manager, like sales persons. Another problem can be bias against particular employees (Axline, 1991), which will be elaborated later on in this part.
Peers, the second possible source for performance management, also have good knowledge about the needed skills of their coworkers and, therefore, are a good supervising option when the manager is not always able to observe the employee. A negative side of using peers is the potential for friendship which can bias ratings (Landy & Farr, 1983). Besides that, peers find it often uncomfortable to be rater and ratee at the same time (McEvoy & Buller, 1987).
less helpful when it is going to be used for administrative purposes (Steel & Ovalle, 1984; Atwater, 1998).
The fourth source is the use of the opinion of the customer. Especially in the service industries, where “the product is produced and consumed on the spot”, it is often difficult for supervisors to observe their employee. In this way the customer is a good source of performance information (Bernardin, Hagan, Kane, & Villanova, 1998). An inconvenience of using customers is that the surveys, which have to be used for the customer evaluation, are expensive.
In conclusion, the optimal way to collect performance information depends on the job. The results and behavior of every employee in the organization have to be managed, although, how often and which information to use depends on the function (Murphy & Cleveland, 1991). However, most of the time, using a variety of sources is the best. A 360-degree appraisal is a good instrument to accomplish this. This is a technique which combines the information of multiple raters like the manager, colleagues, subordinates, and customers (Hoffman, 1995).
Evaluation of employees can be done with objective or with subjective measures. Objective measures are based on figures like sales, profits, or production, while subjective measures are about the estimated value of an employee to the organization. A disadvantage of objective measures is that in most jobs the individual output is hard to measure because of joint production and no possibility to observe the employees (Baker, Jensen, & Murphy, 1988).
For subjective measurement a rating scale can be used to assist in making judgments. It also enables to categorize the judgments to inform performance or contribution pay decisions (Armstrong, 2006). Armstrong explains that rating scales can be defined alphabetically, numerically, with initials (to disguise the hierarchical nature of the scale), and verbally. For all kind of scales, the levels may be defined positive – negative (definitions regress downwards from a highly positive to a negative description) or just positive (strong positive to less strong positive description). The latter one is more popular because it is in line with a culture of continuous improvement.
Further, it has a better motivational value because when an employee is evaluated good and receives the third-level rating (from the positive side) there are still three levels below. This gives a better feeling than when receiving the third-level rating in a five-level scale where only two levels are passed.
Another option for subjective measurement is ranking; grading employees against each other (Bates, 2003). This subject will be elaborated more extensive later on in this part. The essay-approach is also a common used subjective measurement instrument. This approach can vary from a description of the performance of the employee to a structured list of questions which can be completed in the own words of the rater. Advantages of the essay-approach for the rater are that it is not very time consuming and easy. However, this approach is not structured and difficult to administer which makes it less reliable and hard to use the available information (Hoogstad & Weststeijn, 1996).
For good performance management several activities should contribute to an overall systematic approach. Armstrong (2006) describes the main activities. First he describes the role should be defined clearly by agreeing key result areas and competence requirements. Second, the performance agreement is important. Expectations have to be defined with the help of individual objectives, it has to be described in what way performance will be measured, and the competences, needed to deliver the required results, have to be clear. The third activity Armstrong describes is the performance improvement plan, which is about what individuals should do to improve their performance when this is needed. Besides the performance improvement plan, also a personal development plan can be used. This fourth instrument describes how individuals can develop their skills and knowledge and increase their level of competence. The aforementioned activities together have to be managed through the year. This contains a continuous process of providing feedback on performance, conducting informal progress reviews, updated objectives and, where necessary, dealing with performance problems. Finally, the performance review is the formal evaluation over a period of performing. In this evaluation, achievements, progress, and problems will be discussed. On the basis of this review, the performance agreement, performance improvement plan, and the personal development plan will be revised. Besides that, the review can result in a performance rating (Armstrong, 2006).
Errors in rating employees. When rating employees, the quality of rating is very important.
Leniency is about rating individuals higher than they are supposed to be rated (severity is the opposite tendency); the tendency to be favorably biased against to other individuals (Tsui & Barry, 1986; Villanova, Bernardin, Dahmus, & Sims, 1993; Drory & Ben-Porat, 1980). According to Kane, Bernardin, Villanova, and Peyrefitte (1995) it is one of the most troublesome rating errors.
Tsui and Barry (1986) describe the halo-effect as “a rater’s tendency to give similar ratings on all performance dimensions for a single rate”. This can be done both in a positive as in a negative way (Hoogstad & Westeijn, 1996).
Range restriction reveals the spread of the range in which raters rate their ratees. This gives information about how raters differentiate performance levels among ratees. A comparable concept of this is central tendency, which is about clustering the ratings of ratees around the center of a rating scale. The difference is that range restriction assumes a natural center which is the same or not the same as the authentic center (Tsui & Barry, 1986).
Jones, Johnson, Butler, and Main (1983 cited Tsui & Barry 1986, p. 588) explain interrater agreement as the degree to which raters do agree with each other about the performance of a ratee. The degree in which they do not agree shows there is rater error somewhere in the process.
To control leniency, Villanova et al. (1993) give a solution. They state that using multiple observations seems appropriate; each rater should appraise more ratees, the average of all ratings should be used. Like mentioned under the heading ‘Important issues of performance management’ (2.2 performance management elements), 360-degree feedback would be a good instrument because the input of multiple raters is used (Hoffman, 1995).
To also prevent other sorts of rater bias and errors, Borman (1978 cited Smith 1986, p. 22) describes a model in which he suggests that performance appraisal judgment should follow three separate steps. The first step says that the rater observes behaviors that are relevant to the job, the rater has to be able to identify and attend relevant behaviors for effective performance of the employee. Secondly, the rater has to make an evaluation of each behavior, independent of other behaviors. The third step of Borman’s model is that each evaluation is weighted to arrive at a single rating for a performance dimension. According to Borman, a performance appraisal should be reliable and accurate when these steps are effectively dealt with. Smith (1986) and Amstrong (2006) add that rater training is a very effective tool for improving the effectiveness of performance ratings.
this option also has some difficulties, which are described by Meisler (2003). The first difficulty is statistical validity, which is about the needed size of the group to rank in a valid way. Opinions are diverging from a minimum of twenty-three (Handfield-Jones cited Meisler 2003, p. 46) to a minimum of thirty-seven employees (Dick Grote cited Meisler 2003, p. 46). The second difficulty Meisler states, is to be objective, which is very important when rating people. To be objective, a set of clear criteria is needed. However, it is difficult to define criteria in a clear way. Forced ranking can help to make the performance-review system more objective because it forces managers to watch out for not falling in a fixed pattern. Another difficulty, is cannibalization. Meisler explains that the employees which are non-performers will leave the organization after an evaluation. Thereafter, the level of low performing moves up because again there need to be low performers. When doing this again and again, according to Grote (cited Meisler 2003, p. 48), forced ranking looses its effectiveness. The final difficulty Meisler sums in his article is diversification, which provides the strongest argument against forced ranking: bias against women, minorities, and older workers.
For the second difficulty – objectivity – Meisler gives forced ranking (ranking towards a forced distribution) as a solution. However, the other abovementioned difficulties of ranking may also appear with forced ranking. Bates (2003) states that this subject has both proponents and critics, there is no generally excepted research that gives either side clear superiority in the debate. Proponents argue that forced distribution is adequate to identify high-potentials who should receive training, promotions and financial incentives. Also bottom performers can be recognized to help them up or let them leave. On the other hand, critics affirm that the practice can be arbitrary, unfair, illegal, a moral killer, and death to teamwork. Critics say that groups can have a diverse composition; while one group can largely exist of low performers, the other group can comprise mainly of high performers. In this way, it is not fair to judge groups the same and it is hard to identify poor performers (Bates, 2003). Forced ranking systems can hurt productivity; it can negatively affect teamwork because employees will only act in their own interests (Machine Design, 2007). Armstrong (2006) calls it “demeaning and demotivating”.
distributed can differ (bell-shaped, several equal groups, or a 20-70-10 distribution). According to Bates, requirements for a good use of forced ranking are: articulate clear goals and specific criteria, train raters well, merge ranking with other HR metrics, and reward top performers. 2.3 Incidental circumstances
For a PMS to be effective, several factors are important. As mentioned in the introduction, Atos Origin has implemented the PMF since 2007. Therefore, first important aspects of implementing changes will be discussed. Secondly, the legitimacy of a system will be discussed.
Implementing changes. When a new PMS is implemented, it is very important to pay much
attention to this process. According to Williams (2002), it is a mistake to think that when a PMS has been designed, it is ready to be introduced in an organization immediately. The way changes are implemented is very important. It would be sensible to work with a phased approach.
Conger, Spreitzer, and Lawler (1999) describe that managing change can be organized in five categories. First, it is important to motivate the change, which includes creating readiness for change along with the organization members. This also has to prevent resistance to change. The second step is creating a vision. The organization members must understand why the change is needed and what the desired future state is. Third, developing political support for change is important. This means that managers need to gain support from powerful individuals and groups within the organization. Managing the transition from the current to the future state is the fourth activity. This is about planning how to manage the change. Finally, the fifth activity is concerned with sustaining momentum for change, which means it will be carried to completion. All these steps are very important and have to be performed in this order. Organizational leaders have to give careful attention to each activity.
Beer, 1980; Spector; 1989; Kirkpatrick, 1985; Kanter, Stein, & Jick, 1992 all cited by Ford & Ford 1995, p. 542).
Legitimacy and discrimination. Organizations have to find congruence between the social
values associated with or implied by their activities and the norms of acceptable behavior in the larger social system of which they are a part. When these two systems are congruent, there is organizational legitimacy (Dowling & Pfeffer, 1975). In discussions of legitimacy, the focus is more on the cognitive side than the evaluative side (Meyer & Scott, 1983). Suchman (1995) combines these two facts in a new definition of legitimacy. According to him, legitimacy is “a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions”. For a performance management process in an organization it is important to meet legitimacy to prevent a threat to organizational legitimacy. To assess legitimacy, it can be examined which norms and values are prevalent in a society (Dowling & Pfeffer, 1975).
Besides legitimacy, also discrimination is an important issue to consider when working with a PMS. Bernardin, Hennessey, and Peyrefitte (1995 cited Arvey & Murphy 1998, p. 157) have observed discrimination (in age, race, gender) occurs in performance appraisal systems that are too subjective or insufficiently specific. To overcome this, Folger, Konovsky, and Cropanzano (1992 cited Arvey & Murphy 1998, p. 157) have developed three characteristics for a fair appraisal system. First, adequate notice should be given such that employing organizations publish, distribute, and explain performance evaluation standards and processes. Secondly, fair hearing should be provided with a formal review meeting in which an employee is informed of his/her evaluation. Finally, the third characteristic is that raters apply performance standards consistently across employees, without distortion by external pressure, corruption, or personal biases. This, all together, refers to judgment based on evidence (Arvey & Murphy, 1998).
2.4 Performance management outcomes
Finally, the way in which the performance management outcomes can be managed will be discussed. First, the possibilities for motivating performance will be elaborated. Thereafter, how to manage low performers will be addressed.
Possibilities to motivate performance. To attract, motivate, and retain employees, people
forms of rewarding are both extrinsic; they are not directly connected to the job. A form which is directly connected to the job is intrinsic rewarding. This can be enjoyment, freedom, responsibility, and possibilities to develop within the job (Hancké & Kluytmans, 1996; Bartol & Srivastava, 2002).
Besides the kind of job, other factors the reward can depend on are the performance, the seniority, and the competencies or skills of an individual (Gerhart & Milkovich, 1992; Gerhart & Rynes, 2003). When rewarding depends on the performance, this is referred to as performance related pay (PRP) which is discussed earlier in this chapter. It motivates employees to work harder to reach higher payments (Booth & Frank, 1999). However, it also de-motivates employees when they are rated high, but perceive their level of pay to be unsatisfactory (Heneman, 1992 cited Brown 2001, p.889). Therefore, for reward systems to motivate the individual performance, it is important that the rewards have to be perceived as fair. To reach this the employees have to set challenging goals in order to achieve the attractive rewards, and there need to be facilities that insure that employees possess high self-efficacy for performing the tasks. For the reward giver there are the requirements that it should be possible to observe or record the target behavior and that it should be possible to assess the value of this behavior (Bartol & Srivastava, 2002).
Managing low performance. According to Bates (2003), most companies do not
immediately fire anyone judged to be in the bottom category. Most of the time, those at the bottom will receive a warning while employees at the top of the rankings are rewarded with promotions.
performance and provide feedback to the employee and, when necessary, agree on which further actions are necessary (Armstrong, 2006).
2.5 Recapitulation
The first sub-question of this research, about what the requirements are for a good performance management process, was answered in this section; the four categories discussed above, resulted in the new framework with those requirements (appendix B). This framework can be used as an evaluation model to answer the second sub-question about to what extent Atos Origin meets the requirements. Finally, the research question can be answered. How this evaluation was conducted for Atos Origin, is described in the next section, the research methodology.
3. RESEARCH METHODOLOY
The framework, which is the result of the theory section, provides the requirements for good performance management. To find an answer to the research question, about which elements of the performance management process of the Atos Origin department IBS should be improved, first the second sub-question needed to be answered. This sub-question was about to what extent Atos Origin – IBS meets the requirements. Therefore, the requirements of the framework were tested partly by company documentation and partly by information from the experience and knowledge of certain employees in the organization. For the latter one, interviews were used because in this way it was possible to elaborate on the answers immediately to collect highly qualitative information.
This section will elaborate on the participants and the procedure of the research, and the measures and data analysis used to evaluate the PMS of Atos Origin.
3.1 Participants
To delimit the research area, the focus was on the IBS sub-unit. IBS consists of 1058 employees and is divided in six departments through the country, which are: IBS North, IBS Middle, IBS West, IBS South, Operations Support Systems (OSS), and Transitions & Projects (T&P).
Management Team member (HR-MT member), a manager, and an employee from all six departments which makes a total of eighteen participants.
The group of HR-MT members was representative because all HR-MT members of IBS were involved. Also the group of managers seems to be representative because they agreed with each other for almost all questions. Besides that, for the conclusions, their answers were completed with information from the company documents. For the employees a lower percentage could be interviewed. Only six out of 1058 employees were involved in the interviews. However, for almost all questions the employees agreed with each other. Therefore, the answers seem representative enough to use as a source for the evaluation.
Seventeen participants were males, just one participant was female. This corresponds approximately to the male-female ratio (male: 952, female: 106) of the population of the organization. The seniority of the participants was in the range of one to thirty-three years, with a mean of ten years. Two participants joined the organization after the implementation of PMF. However, this has not caused any problems during the interviews.
3.2 Procedure
To decide which requirements should be tested with the help of company documentation, which with the help of interviews, and which with the combination of both sources, it was analyzed what knowledge was expected to be possessed by the participants, and what information should be derived from the company literature by scanning this.
Interviews. Two different interviews were developed, one for the HR-MT members and
managers, and one for the employees. The HR-MT members and the managers were combined because for them the same questions were appropriate. The HR-MT members were contacted to explain the purpose of this research, to ask them if they were able to participate in this research, and if they could provide names of another manager and an employee of the department which could be approached to participate. Subsequently, also these managers and employees were contacted. Eventually, all planned interviews could be conducted, each of which had a duration of approximately one hour, in a timeframe of one-and-a-half week.
Company documentation. For the information to be gathered from the company
documents, all information on the intranet was scanned for relevant data. Also employees who were expected to own the suitable knowledge were asked for useable documents.
3.3 Measures and data analysis
The content and order of the interview questions, the formulation of the questions, and possible answers have been registered as much as possible before the interviews were conducted to prevent influences of the interview and to gain as much comparable information as possible (Hutjes & Van Buuren, 1992). Therefore, most of the interview questions had the options to answer yes or no with always the possibility to elaborate on the answer. For some questions, which asked for an opinion, a five-point Likert scale was used. The reason to choose a five-point scale was that the rating scale the employees are used to work with is also a five-point scale.
The interview questions were all derived from the literature of the theory section. Only for the variable how employees judge PRP, part of section 2.1 Context – performance management characteristics, research of Silfhout (2000) was used.
To give a clear overview of the interview results, the answers are presented in a combined table in appendix C in the same order as the theory section. With the help of this table it was possible to analyze if the requirements were fulfilled or not. The results were not analyzed statistical because of the evaluative character of the research. The analysis was completed with the information from the company documentation.
4. RESULTS
4.1 Context
The first category of the framework, which is about the context, illustrated requirements concerning the context of the performance management perspective and the context of the organization. This section describes to what extent these requirements have been met.
Performance management perspective. The interview with the managers suggested that the
main perspective of the performance management of IBS is to integrate the management of organizational and employee performance. To obtain this objective, the organization works with a PMS, which consists of a cycle of three time points in a year. At the first time point, at the beginning of the year, the employee sets the objectives for the upcoming year. The individual objectives are derived from the organizational objectives by, first, a communication from IBS through the department managers, about the mission and objectives, down to each department. With this knowledge, each department develops its own objectives which are in accordance with the organizational objectives. Thereafter, the objectives of the department will be communicated back to IBS. IBS, however, does not change its objectives with this information; it is a top-down process. In accordance with the objectives of the department, the employees set their individual objectives in the form of competences. For each skill-level it is described in a competence matrix (Global Capability Model) which competences are obligatory. Besides those obligatory competences, an employee also has optional competences to choose and has the option, if desired, to add some self initiated competences. The employee fills his/her own balance scorecard with these competences, which has to be approved by the manager.
At the second time point, the mid-year evaluation, the objectives will be evaluated by the manager of the employee. In this evaluation the progress of achieving the objectives will be discussed. According to the company documentation, at this moment, it is possible to adapt the objectives to internal or external changes. According to the interviews, as well from the managers as the employees, the objectives are not being adapted at this moment. They mention that this information can only influence the objectives for the next year.
TABLE 1
Description of the performance categories Category Description
0 Not able to evaluate 1 Not at level
2 Not yet at level
3 At level
4 Above level
5 More than above level
The next year, the whole process starts over again, influenced by the results of the past year. The degree in which the obligatory competences for the next year will change depends if the employee remains in the same skill-level or moves up to a higher skill-level. The degree in which the optional and self-added competences will change, depends on the employee.
Whether the results of the individual objectives go back up to the departmental purpose analysis, so that the department can adapt its mission in the light of enhanced performance levels achieved, varies in the answers from the interviews and appears to depend on the manager his or her vision about the importance of this feedback.
This yearly three-time point process is supported by a digital tool, the performance management framework (PMF). This PMF works with performance planning, assessment of performance, performance improvement, performance review, and actions via feedback (Performance brochure, 2006; performance management user guide, 2006; https://source.atosorigin.com/elearningpm/nl/portal.htm, 14-10-2009).
The HR-MT members, managers, and employees, made clear that besides the two formal moments (mid-year evaluation and end-year review), there are additional contact moments between the manager and the employee in which prominent issues can be discussed. In combination with these informal moments, the participants judge two formal moments a year as a sufficient amount.
performance evaluation score is connected to a percentage of salary increase. Besides that, every employee is scaled at a certain level in a reference scale which depends on the skill level.
Besides the connection of the performance evaluation with salary increase, the performance evaluation can also result in steps for development like training or workshops. However, this connection is indirectly related because it is not standard registered which evaluation outcome leads to which steps of development.
Organizational characteristics. The PMF Atos Origin works with, consists of three
HR-tools which together result in the evaluation. The first tool, the Global Capability Model (GCM), describes all functions of the organization with the related skills. From these skills, the objectives are derived which will be filled in the Balanced Score Card (BSC), which is the second tool. This BSC consists of four areas which are Customers, Financial, Internal organization processes, and People, education and development. The third tool is the Atos Origin Competency Model (AOCM). This model defines the behavioral characteristics, which are necessary for successful performance, and which are identified as competences. These competences are also connected with the skill-level. The score of the evaluation consists of a combination of the outcomes of the BSC (50%), the GCM (25%), and the AOCM (25%). In this way, the performance evaluation is composed of objectives, skills, and competences which is described in the 2008 Beloningssysteem Atos Origin Nederland, published by the Dutch HRM department. Also the interview results (both of the managers as the employees) show that both specific individual objectives and behavioral traits are part of the evaluation.
4.2 Performance management elements
This section will show the results of the second category of the framework, the performance management elements, in which the requirements of important topics in performance management and issues which can go wrong are addressed.
Important issues of performance management. According to the results of the interviews
with both the managers and the employees, the evaluation is based on input from the line-manager, colleague-managers of the line-line-manager, the employee (by filling PMF, this is the responsibility of the employee), and the customer by use of a customer evaluation form. The employees disagree with each other if these sources are sufficient. Half of them would like to be evaluated by more sources, like peers or subordinates.
The interviews also show that the evaluation outcome is presented with a score from one (not at level) to five (more than above level) which is a result of the BSC, GCM, and the AOCM (see table 1 for the definition of the five scales). Even though this score is shown with a number, the sources which it is based upon exist of subjective measures (described in competences and skills, not with the help of digits about facts) about the way the manager appraises the performance of the employee. Besides that these measures are subjective, the scale definition is negative (not at level) – positive (more than above level). Nine out of the twelve managers do agree with this negative-positive scale definition. Also all employees agree with it. The amount of five levels, is experienced as comfortable by eight of the twelve managers. Some managers mention that four levels would be better to make a clear distinction between under or above level and to avoid the ‘easy three’. Employees have different opinions about the five-point rating scale. Four out of six do agree with it, however, at the same time they call into question if it would be better to make a more precise distinction with more levels or make it more clear working with fewer scales.
performance, and what help is needed from the manager. During the whole process of the Performance Management cycle, individual career ambitions, good performed issues, and issues which have to be improved will be discussed. Employees can set some actions which are helpful to reach the objectives. Therefore, every employee at Atos Origin needs to have an Individual Development Plan (IDP). This is an instrument with the objective to stimulate and plan the personal and professional development (Notenboom & Schouwenaar, 2002; Kamphuis, 2005). The final part of the PMF cycle is the performance review, which occurs in the end-year review. With the results of the end-year review, the performance will be rated. Furthermore, the information will be used for setting new objectives for the next year.
Errors in rating employees. Like mentioned above, multiple observations are used to
evaluate the employees. The behaviors which have to be observed are, according to all managers, relevant to the job and are first evaluated separately. After these separate evaluations, the behavior evaluations are all taken together for a total end score. For this result, the different behaviors all have different weights. The manager interviews show that some workshops and trainings about evaluating employees are provided to all managers. However, these are optional. According to the manager interviews, besides rating, also ranking the employees against each other is used in the organization. The groups in which the ranking takes place, have a size from two to one hundred sixty-five employees. This differs per department. In some departments, managers rank all their employees against each other. In other departments, managers only rank employees with the same function or skill-level. For the ranking, Atos Origin makes use of a forced distribution. The norm for this forced distribution and the forecast for the end-year review of 2009 are shown in Table 2.
TABLE 2
Forced distribution: norm vs. forecast.
Evaluation category 0 1, 2 3 4 5
Norm: 3% 6 - 10% 60 - 63% 23% 4%
Forecast 2009 5,3% 4,1% 64% 23,7% 2,9%
(source: R. van der Hoek, HR department)
According to all managers, the goals and specific evaluation criteria are communicated clearly to them. Most employees agree with each other that they are informed well about how they are getting evaluated and to which standards they have to adhere. Eleven of the twelve managers agree that the ranking process merges with other HR-metrics in the organization. They argue that, for the evaluation, multiple issues are considered with the help of the BSC.
Top performers are rewarded by a higher percentage of salary increase or with promotion to a higher skill-level at the end of the year. Some other employees will receive a bonus when they have reached their target at the end of the year. Employees express that now, due to the current economy, the salary increase is very low.
4.3 Incidental circumstances
The third category of the framework, of which the results will be described here, was about important requirements when implementing changes and the importance of taking care of legitimacy and prevent discrimination.
Implementing changes. In the company documentation it was explained that, with the
implementation of PMF, the explanation of the evaluation categories has changed. Especially the meaning of the ‘2’ has changed. Before, receiving a ‘2’ meant that an employee was in a dangerous zone, that dismissal was close. There was no salary increase matched with a ‘2’. Now, the meaning of the new ‘2’ should be more positive, it now means that someone just has to develop a little bit more. With the new ‘2’, it is possible to match it with a salary increase by choice of the manager.
Legitimacy and discrimination. Nine of twelve managers feel that the PMS is derived from
the norms and values because the vision and culture are visible in it; competences are derived from the norms and values. The employees experience the evaluation system as honest and as a complete overview of all aspects of their performance. Besides these opinions of the employees, also the PMF makes that the system is derived from the norms and values because, with this tool, employees fulfill their objectives by them self and, in this way, influence it.
According to the managers, performance evaluation standards are communicated clearly to the employees. Employees can read all areas of concern in their PMF. The employees almost all agree with this. They know what is expected because of the competences and objectives in PMF. Also per skill-level every employee knows what is expected. Each employee is evaluated two times a year with a formal review meeting between the manager and the employee. According to the managers this is often enough because, when necessary, additional informal meetings are inserted. Employees say that in the evaluation the result is explained very well but that it is not discussed, that they do not have any input in the meeting.
Half of the managers answer that they apply the performance standards in the same way to all employees. The other half, which do not apply the standards for all employees the same, argues that this is because personal circumstances are taken into consideration. They also argue that it is impossible to apply it for every employee exactly the same because there is still a lot of subjectivity. Also the employees mention this issue of subjectivity as something which makes it hard to be evaluated in the same way as other employees.
4.4 Performance management outcomes
Finally, the results of the fourth category of the framework will be shown. This category was about the requirements which have to be met for motivating people and about how to deal with low performers.
Possibilities to motivate performance. The rewards which are used in the organization are
above the average responsibilities of the employee). According to the managers, besides salary also public recognition is used in the form of diner coupons, presents, and gratification. The intrinsic rewards which are experienced as very important according to the employees, are the enjoyment, freedom, and responsibility they have in their job and chances to learn from their job. The employees all do experience these forms of intrinsic rewarding in their job in a sufficient way.
According to the managers, the level of rewards depends on the performance which constitutes of competences and skills. The 2008 Beloningssysteem Atos Origin Nederland, published by the Dutch HRM department also explores some more sources the rewards depend on. First, there is a relationship between the weight of the job and the place in the range Atos Origin works with. Here it does not depend on the person who performs the job, but on the weight of the duties and responsibilities which are part of the function. The second source is the level in which an employee contributes to the company result. The third source in this document is about the competences, desired behavior of employees which determines the employability that is valuable for the company. The fourth source is about the market value. Not all functions are available on the market to the same extent. As a result of scarcity some functions can (temporary) receive a higher salary than other functions. The final source described in the document is the result of the company. The level of the remuneration has to move with the company results.
Four of the six employees perceive the way they are rewarded as fair. They notice however, that now, with the current economy, the percentage of the salary increase is very low. Two of the six employees set their individual objectives in order to achieve attractive extrinsic rewards. Most of the employees experience the intrinsic rewards as more important for setting the individual objectives.
According to the managers, they provide facilities to the employees to ensure that they possess high self-efficacy for performing their tasks. However, the employees are responsible to ask for it. Half of the employees experience the support from the managers as good, the other half is less positive. They say that support is only provided when they ask for it.
Managing low performance. When an employee does not perform well, is evaluated with a
‘2’, this will not immediately result in firing. From the interviews with the managers it seems that now, with the new ‘2’, there are two different interpretations of it. An employee who has just started in a new function, may not have had enough time to develop him- or herself to the expected level yet. In this situation the employee will not be judged on it but, nevertheless, will be evaluated with a ‘2’ because he or she is not yet at level. In this situation the manager can choose to combine the ‘2’ with a salary increase, further nothing will happen with the employee. The situation is different when an employee underperforms while working on the function for a long time. In this case, the employee will also receive a ‘2’ and the manager will not choose to reward this employee with a salary increase. A performance development plan will be written by the employee with the help of the manager. All five actions which are needed according to Armstrong (2006), are taken in this plan. Without any progress of this plan, eventually the employee will be fired and has to leave the organization.
5. DISCUSSION
To answer the research question ‘Which elements of the performance management process of the Atos Origin department IBS should be improved?’ in this research, first two sub questions needed to be answered which were:
• what are the requirements for a good performance management process, and; • to what extent does Atos Origin – IBS meet these requirements?
The first sub question was answered with the help of the framework, developed in the theory section and showed in appendix B. To what extent these requirements are met by Atos Origin – IBS, the second sub question, was showed in the right column of appendix B. This section will discuss the interesting results of each of the four categories, followed by a discussion of the research limitations, and an overall conclusion.
5.1 Context
Performance management perspective. Atos Origin works with a global PMS, the PMF.
Because this PMF consists of performance planning, assessment of performance, performance improvement, performance review, and actions via feedback the way of using the PMF fits with its perspective of integrating the management of organizational and employee performance.
First, the feedback loop, which should be made from the departments to IBS about the set objectives, is partly absent. The departments do communicate their objectives back to IBS but subsequently IBS does not adapt its mission and objectives. However, because the objectives are also derived from the skill-level, the function, and the personal desired objectives, this should, in my opinion, not lead to problems.
Second, the literature describes that the performance evaluation should be differential with performance related payments (PRP) or with steps for development. For IBS the performance evaluation is differential with PRP but not with steps for development possibilities. Because the requirement is that just one of the two should be applied, also this remark is not a problem. However, it should be valuable to do more research if it would have a positive effect on employees if the performance evaluation would also be differential with development steps.
The third comment is about the antinomy between the company documentation and the results of the interviews about the mid-year evaluation. At this moment, according to the documentation, the objectives should be adapted when this is needed. This does, as shown by the interviews, not really happen, only at the end of the year the objectives can be adapted. In the theory-section the ratchet-effect was mentioned as a possible negative and demotivating influence on employees when frequently renewing objectives. However, for the system Atos Origin works with, the obligatory objectives (in the form of competences) only need to change when an employee is motivated to move to a higher skill-level. In this way, renewing objectives has a motivating character. When an employee has decided to stay on the same skill-level, these objectives do not need to change. If optional and self-added competences will change, also depends on the employee. In this way, the ratchet-effect will hardly occur. Because the documentation of the organization states that renewing the objectives should also happen with the mid-year evaluation and the literature in the theory part states that this is a process which should happen during the year, it is worthwhile to make clear to the managers that they are supposed to adapt the objectives when this is needed.
The fourth and final comment for these requirements is that, at the end of the cycle, there should be a feedback loop from the departments to IBS about the results of the performance evaluations. This occurs at some departments and depends on the manager, depending on whether he or she considers this step as important. It would be helpful if all managers should feedback to IBS the results, then IBS can adapt the mission and objectives at this point.
Organizational characteristics. As a consequence of using three different instruments
of the professional bureaucracy and the division structure/adhocracy, fits with the performance management of the organization. This requirement is fulfilled as a result of some characteristics. First, the departments have leeway in how to apply the PMS, which is worldwide. In this way the departments can build a strong identity to attract and bind employees and the dichotomy of the division structure/adhocracy is fulfilled. Another characteristic is that, according to the results, employees are a strategic factor.
5.2 Performance management elements
Important issues of performance management. For a more comprehensive evaluation it
would be valuable to use a 360-degree appraisal which is an approach that gathers behavioral observations from many layers within the organization (Hoffman, 1995). To reach a 360-degree appraisal, evaluation sources which can be added to the sources already in use are peers and subordinates of the employee.
The used scale definition is negative – positive. Although the literature argues that it is better to use just positive levels, all employees experience this negative – positive scale as acceptable. Therefore, this needs no change.
Most managers and employees agree with the amount of five levels, however, according to the literature it is much better to use four or six levels. Besides that, it still seems that most managers evaluate with the ‘easy three’ which is a form of central tendency. Therefore, it would be better to work with a four- or point scale. According to Armstrong, it is better to use a six-point scale instead of a five-six-point scale because it has a higher motivational effect and it gives a more clear overview of performance (Armstrong, 2006, p. 111). However, whether it is more suitable to use a four- or six-point scale, should be investigated in the organization. Though, the organization has to take into account that, in this case, again a new implementation has to be communicated which can cause commotion among the employees.
Errors in rating employees. To maintain high quality of the evaluation it is important that
bias and errors are controlled. Some requirements to control this have been met, like the use of multiple sources and evaluation of relevant behaviors in a separate way with a different weight.
their function. This way it can be prevented that managers do not have sufficient knowledge of evaluating employees when they start being a manager.
Second, the way in which employees were ranked was not valid. The composition of the group of employees to compare with each other was different and not always large enough; some managers ranked all employees of their department with each other, some only compared employees of the same skill-level or function. Forced distribution has, as shown in the theory part, many advantages and disadvantages. It is important to consider which impact these advantages and disadvantages have for the organization and if the organization should work with this forced distribution when so many managers do not agree with it and do not follow the rules. If the organization decides to use it, then it is important to be stringent about achieving the desired norm and all managers have to apply the ranking for the same sort of group composition. Now, the forecast for the end-year evaluation of 2009 looks as is showed in Table 2 in the results section. This table shows that the forecast comes close to the norm but there still needs to be evaluated in a more stringent way.
The final requirement which needs some attention is the rewarding of top performance. High performance is rewarded. However, because of the current economy, now the difference in salary increase for a higher rating scale is very small. This makes it a less stimulating factor for high performance.
5.3 Incidental circumstances
Implementing changes. The phased approach of Conger et al. (1999), described in the
theory section, was not used while implementing the PMF. There was a large communication process but still a lot of the employees are not aware of the new category description, they still associate it with the old description. A reason for this is that the categories still look the same with numbers from one to five. It is not enough just to communicate the new description, it is important to tackle the problem where it comes from, to change the vision of the employees. Like mentioned before, this can be done by changing the amount of levels of the evaluation system.
Legitimacy and discrimination. To prevent resistance against the evaluation process, it is
Discrimination is prevented sufficiently. Performance evaluation standards are communicated clearly with the personal information of PMF. Further, the frequency of formal meetings is experienced as sufficient by the managers as well as the employees. Nevertheless, employees would find it more desirable if the results can also be discussed. Now, the results are just communicated to the employees, who experience they have no influence. The managers refute this argument with the fact that employees do influence their performance evaluation by filling out the PMF before the evaluation meeting. In this way, it is their responsibility to influence the evaluation.
It is not easy to apply the evaluation standards on every person in exactly the same way. Although the PMF tries to make measures as clear as possible, it is still partly subjective. This is acknowledged by the employees. In my opinion, evaluation will always be subjective somehow and has to be managed as correctly as possible.
5.4 Performance management outcomes
Possibilities to motivate performance. It is important that the rewards are perceived as fair.
Most of the employees feel the manner in which they are rewarded as fair but due to the economic situation the difference of salary increase between the different rating scales has become very small. This makes the salary increase a less motivating aspect to perform. Intrinsic rewards are experienced by the employees as the most important for setting individual objectives. With this second aspect, an antinomy arises. The fact that employees experience intrinsic rewards as more important than extrinsic rewards, should make it more acceptable for employees that the salary increase has become very small with the current economy. However, they are almost all bothered about the low salary. With this information it can be concluded that, with this general increase of the salary, PRP does not have the effect as it should have.
For fair rewarding, it is important that managers can observe or record the target behavior of the employees and that they can assess the value of it. The managers all agree that they can observe their employees well and assess this value. When employees work for a client, it is not always possible to observe the performance of the employee in a direct way. However, this fits with the sort of work area the organization is in (secondment).
Managing low performance. The results made clear that without any progress of the
performance development plan, which is performed in accordance with the five actions described by Armstrong (2006), an employee has to leave the organization. Atos Origin has to be aware of the cannibalization described in section 2.2 ‘Errors in rating employees’. When the bottom layer has to leave every year, because there always needs to be a certain amount of ‘2-performers’, this can have a demotivating effect on the employees.
5.5 Research limitations
Some remarks have to be made about the way the research has been conducted. First it is important to make notice of the representativeness of the research sample. Like mentioned in the methods section, for the interviews with the employees, only six out of 1058 were involved. However, this was experienced as representative because the employees almost all agreed with each other. For further research, involving more employees by additional questionnaires can be used to give a stronger establishment for the research findings. Though, this will not change the outcomes.
The second limitation which has to be mentioned is about the seniority of the participants. Two of the participants started working in the organization only after the implementation of PMF. Because of this, they are unaware of the situation of the organization before. However, those participants were able to answer all questions. For further research, it would be helpful to take notice of the seniority of the participants.
The third and final remark which has to be made is the current economic situation which possibly influenced the answers of the participants negatively.
5.6 Conclusion
