The influence of client-interaction on managing professionals in Professional Service Firms

(1)

Faculty of Economics and Business

The influence of client-interaction on managing professionals in

Professional Service Firms

Name: Kim Beltran Borrero

Student number: 11102780

Thesis supervisor: Dr. Sander van Triest

Date: 25 June 2018

Word count: 14.386

MSc Accountancy and Control, specialization Control

(2)

Statement of Originality

This document is written by student Kim Beltran Borrero who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

Professional Service Firms are considered a very specific type of business due to the nature of the product. The product is intangible and inseparable from the production that requires a high degree of customization and high degree of interaction between the client and professionals. Consequently, these professionals have a desire for autonomy and freedom to make decisions (Von Nordenflycht, 2010) and would like to be managed based on their performance and less on their actions, especially because the professionals’ performance is dependent on the client verifiability and integrativity. Thus, the design of Management Control Systems in Professional Service Firms is subject on one side to the ability to measure results on performance and the knowledge which actions and behaviours are desirable from professionals. On the other side Professional Service Firms is subject to the role of risk aversion.

This research uses the survey method to investigate the influence of client interaction on the design of Management Control Systems (MCS) in PSF. The MCS used in this research are the Merchant and van der Stede (2012) controls, known as result controls, actions controls and personnel/cultural controls. The effect of client interaction is exanimated on each of these types of controls and the interaction on the sectors accountancy, consultancy and engineering on result controls and action controls. 199 questionnaires were administrated to Professional Service Firm’s managers and 79.8% of these were used in this study. Using multiple regression analysis, the results indicated that PSF that deal with high client interaction are positive related to result controls and action controls and accountancy firms are positive related to personnel/cultural controls. Finally, the results indicated that client interaction in the sectors accountancy, consultancy and engineering does not have a significant effect on result controls and action controls.

(4)

3.5.6 Robust test ...35 3.5.7 Control variables ...36 3.6 Descriptive statistics ...36 3.6.1 Correlations ...37 3.7 Research model ...39 4 Regression ...40 4.1 Hypothesis 1 ...40 4.1.1 Hypothesis 1a ...40 4.1.2 Hypothesis 1b ...41 4.2 Hypothesis 2 ...42 4.2.1 Hypothesis 2a ...42 4.2.2 Hypothesis 2b ...43 4.3 Hypothesis 3 ...44

5 Discussion and conclusions ...46

5.1 Research limitations ...47

References ...49

Appendices ...52

Appendix A: Survey Professional Service Firms Project...52

Appendix B: Dataset characteristics ...53

Appendix C: Factor analysis Client interaction...54

Appendix D: Factor analysis Management Control Systems...54

(6)

(7)

7

1 Introduction

1.1 Motivation

Over the past decades the Professional Service Firm (PSF) sector has performed as one of the most rapidly growing, profitable, and significant sectors of the global economy. In 2013 the PSF sector generated around US$ 3.6 trillion in revenue, including accountancy, management consulting, legal, architectural, engineering services and advertising sectors (Empson et al, 2015). The service sector is a very specific type of business due to the nature of it’s product (i.e. intangible and complex). Some professional service requires holding a professional license such as architects, auditors, engineers, doctors and lawyers. Others involve providing specialist business support to all types of organizations in all sectors; this can be management consultancies in tax, financial or IT. Some literature has provided different classifications to be able to understand and standardise the PSF types (Silvestro et al 1992). However, there has been a limited amount of research done on PSFs. Van Nordenflycht (2010) suggests this is due to the PSF’s distinctiveness, the ambiguity about the PSF term and the difficulty to gain up-to-date information about the scale of the sector.

PSFs generate intangible experiential services in the form of knowledge-rich and time-sensitive advice that is customized to specific client’s problems or needs (Von Nordenflycht, 2010), this implies a much higher degree of relational embeddedness and context sensitivity compared to other kinds of businesses. PSFs limit the scope for traditional strategies of standardization, commoditization (Empson et al, 2015) and manifest different type of management and organizational challenges (Von Nordenflycht, 2010). Such as managing the professionals employed, client expectations and external image (i.e. reputation). Moreover, due to the reason that professionals and client relationships are the main PSF assets (Greenwood et al, 2005); dependence on these highly mobile and highly portable assets creates important complexities in terms of how PSFs approach their client relationship and how PSFs manage their professionals (Empson et al, 2015). In regard to complexities as high demanding clients, quality of product (this being subject to client expectation and satisfaction) and professional motivation.

As mentioned before, one of the most important assets of PSFs are professionals, these bring complex managerial challenges. This human capital is composed of high-educated professionals, with a high level of knowledge and a high desire for autonomy. For instance, power in PSFs tend to be highly dispersed between autonomous professionals who retain a significant amount of information about how their work is organized (Empson et al, 2015). It

(8)

can be presumed that there is conflict between professionals and bureaucratic principles of control (Von Nordenflycht, 2010), such as hierarchical structure, authority, formal procedures, rules and regulations. Empson and Langley (2015) emphasize that PSFs present distinctive leadership challenges, given professionals’ expectation of autonomy from organizational constraints. Others argue that PSFs apply more loose management, where the management is more consensual and mindful of individual preferences (Empson et al, 2015).

Furthermore, the service provided by professionals is based on experience and knowledge (e.g. engineering, financial advice) and tailored to the client’s needs (Greenwood et al, 2005). Therefore, the client plays an active role in the production of the service. It requires coordinated efforts of professionals and clients; close contact and on going communication between the two firms are all necessary for the delivery of service (Broschak, 2015). Clients provide input that is critical for the production of the service. The quality of this input is important, so professionals are able to provide a good service and satisfy the client’s demand. For the PSFs it is therefore important to have a good client relationship to be successful (Broschak, 2015).

Management Control Systems (from now on refer as MCS) refer to the mechanisms a firm uses to make sure that objectives are achieved, such as targets, incentives and procedures. Having MCS means that management can be reasonably confident that no unpleasant surprises will occur. How do these characteristics of PSFs, autonomy professionals and client interaction, impact the MCS design? Regarding MCS in PSFs, Auzair and Langfield-Smith (2005) concluded that PSFs normally use less bureaucratic MCS compared to mass service. Less bureaucratic MCS are mostly result, informal, loose, flexible, and interpersonal controls.

There has been no specific research done to the client-interaction and it’s impact on MCS of PSFs. Also, many studies indicate the limited research made to PSF and the design of their MCS. Prior research to PSF and MCS has mostly been case studies and are focused on organizations in a single industry, e.g. health care or communication industry (Auzair & Langfield-Smith, 2005). Taking into account the distinctiveness of PSFs and the important role of the client for PSFs this study examines MCS design in PSF, looking in a closer loop to the effect of client-interaction in action, results and personnel/cultural controls in PSFs. Focusing on the following question, ‘How does client-interaction influence the Management Control Systems of Professional Service Firms?’

(9)

9

The methodology of this research is a survey and can be classified as a quantitative study. Quantitative research examines statistical relationships on the basis of predetermined hypotheses.

1.2 Academic relevance

The purpose of this research is to provide logical and structured research about the design of management control systems in professional service firms specific in regard of client-interaction. Furthermore, the aim is to make an empirical contribution to current limited theoretical studies about PSF. This study uses PSF theory to go through the research and explain outcomes.

1.3 Outline

The rest of the paper is organized as follows. In the next section, the theoretical framework provides the relevant theories and explanation of terms relevant for this research. This is followed by the development of specific research hypotheses to be empirically tested. After reviewing the research method used in this research, the results of the empirical tests were reported. The last section includes the discussion, conclusions of the research along with the study’s limitations.

(10)

2 Theoretical framework

2.1 Professional Services Firms (PSF)

PSFs do not know a clear definition (Von Nordenflycht, 2010), due to their distinctiveness amongst PSFs themselves. As there are many services that fall into the category of the PSF term; examples are management consultancies (i.e. auditors, financial advisors, etc.), law firms and corporate banking. Despite this, Silvestro et al (1992) defines professional services as:

“Organizations with relatively few transactions, highly customized, process-oriented, with relatively long contract time, with most value added in the front-office and where considerable judgement is applied in meeting customer needs.” (Page 73)

As well, several researchers have pointed out that direct application of theories and techniques developed from the manufacturing area to the service area are inadequate (Thomas, 1978; Anderson et al., 1989; Wemmerlov, 1990 as cited in Kellogg & Nie, 1995).

Therefore, many studies identify different characteristics of PSFs distinctiveness. One important characteristic is the PSF’s governance. The governance in organizations is based on the premise that managers and employees do not have identical goals. As a consequence, employees need to be monitored and controlled through rules, practices and processes. Governance should balance the interests of an organization’s stakeholders, such as shareholders, employees, management, customers, suppliers, financiers, government and the community (Shapira, 2000).

Empson et al (2015) indicate that extensive individual autonomy and contingent managerial authority characterize PSF’s governance; where core producers, i.e. professionals, own or control core assets (such as procedures and production of the service). Therefore, a well-known governance type in PSF is the partnership form (Empson et al, 2015; Auzair and Langfield-Smith, 2005), this is the opposite to the corporate form (Boussebba & Morgan, 2015). The corporate model enabled the organizations to access capital for investment purposes and insured that owners could separate their personal wealth from their investment in any particular enterprise (Boussebba & Morgan, 2015). The partnership form relates to two or more partners, which induces partners to monitor and pressure each other to provide a quality service; since each are at risk for any actions of the others which expose the firm to financial or legal liability (Greenwood & Empson, 2003 as cited in Von Nordenflycht, 2010).

(11)

11

Another identified characteristic is the nature of the PSF’s product. Firstly, the service is intangible. This means that firms cannot hold products in inventory. Secondly, the inseparability of production from consumption. This relates to the complexity of the co-production of the service, where clients are also involved in the co-production of the service. Their involvement relates to the input of information required to be able to provide the service; thus, the customization of product. Thirdly, the perishability of services and the heterogeneity in service products (Auzair & Langfield-Smith, 2005), where services that are not consumed are lost and the services provided by the same person may differ between clients or differ at different times. These items are specific and time related. The advice provided by the PSF is customized to certain needs and problems of the client in a particular moment in time. For example, a client has a new sales office in another country and needs some advice on how to deal with their transfer pricing. This advice is subjected to the current legal situation of the firm and the location of the new sales office at that certain time. Additionally, knowledge is another PSF characteristic. This relates to the fact that knowledge is carried by a highly educated and mobile workforce, i.e. professionals in a PSF. (Auzair & Langfield-Smith, 2005; Campbell & Verbreke, 1994). The professional is the expert in the matter.

Consequently, Von Nordenflycht (2010) identified and concluded three characteristics of PSFs; as following, knowledge intensity, low capital intensity and professionalized workforce. Knowledge intensity indicates that production of a firm’s output relies on a substantial amount of complex knowledge. Low capital intensity indicates that a firm’s production does not involve significant amounts of nonhuman and tangible assets, such as inventory, factories and equipment, and even intangible nonhuman assets like patents and copyrights. Thirdly, professionalized workforce relates to the highly autonomous professionals that work within a PSF.

As a PSF characteristic, the client plays an important role in the PSF delivery of the service, making the provider-consumer interaction an integrated part of the process (Boussebba & Morgan, 2015). Chase (1981, as cited in Kellogg & Nie, 1995) identified customer contact as the most important characteristic of service operations. Time, knowledge and human resource are key in this interaction; as the outputs are intangible services encoded with complex knowledge (Greenwood et al, 2005).

2.1.1 Classification of PSF

In the past few decades the service literature has been expanded and different authors have provided many classifications based on different dimensions (Silvestro et al, 1992). This research

(12)

focuses on the two-by-two dimensional Maister and Lovelock (1982) classification model and the six-dimensional Silvestro et al (1992) classification model to allow a better understanding of service types and the customer interaction. Maister and Lovelock (1982) provide a two-by-two-dimension classification of service based on the customization of activities and customer contact. Customized activities involve framing a service package for each customer. Customer contact relates to the length of time the customer has an interaction with the service provider. Maister and Lovelock (1982) place PSFs with high service customization and high customer contact. See below figure.

Figure 1: Maister and Lovelock’s Classification of service (Maister & Lovelock, 1982)

Silvestro et al (1992) provides an extended literature analysis concluding with a new classification of PSFs composed by three types, Professional Services, Service Shop and Mass service. Their classification knows six dimensions; (1) Equipment/people, this relates to the core element of the product, (2) Contact time per transaction, (3) Customization of activities, (4) Discretion, which is the degree of management involvement by providing the service, (5) Front-office is the proportion of customer contact staff to the total company staff, (6) Product/process focus, a product-oriented service is where the focus is on what the customer buys and a process-oriented service is where the focus in on how the service is provided to the client (Silvestro et al, 1992). See figure 2, on the next page.

Client-relationship is an important factor in the classification of the service types. As mentioned previously the following dimensions emphasise the client relationship and PSF: client contact time, customization, front office and product. This can be explained by the fact that the client takes part of the service production.

Concerning the management of the professionals Silvestro et al (1992) include the dimension discretion. PSF’s have a high degree of discretion, meaning that professionals that have direct contact with the client can exercise judgement in amending the service or process

(13)

13

without consulting a superior/manager (Silvestro et al, 1992). This concerns not only the desire for autonomy by professionals but also the ability to manage highly independent professionals.

Figure 2: Service types by Silvestro et al (page 71, 1992)

2.1.2 Human Capital

The professionals constitute a critical asset of the PSF because it embodies, operates, and translates the knowledge inherent in the firm’s output. These professionals are also referred to as human capital of a PSF. Empson et al (2015) argue that the professionals are the ones investing their time and effort in developing their knowledge (i.e. firm and client specific) and building their client relationship, which is mostly based on trust and commitment from both parties, eventually enabling them to apply their specialist technical expertise appropriately (Finchem, 1999 as cited in Empson et al, 2015).

Professionals expect comprehensive levels of autonomy, legitimated by the requirement for professionals to preserve the right to make choices about how best to apply their specialist knowledge to the delivery of the tailored service (Empson et al, 2015) and about how they schedule their own daily activities. Autonomous professionals normally do not like to share information about how they organize their work or schedule their time. Their desire of autonomy is strong. In Silvestro et al (1992) classification it is remarkable that discretion is so high, meaning that professionals use their own judgment to solve client’s problems without management intervention. Autonomy in PSF is associated with relatively low levels of managerial intervention.

Besides, it is critical for PSFs to make sure professionals are aligned with and harnessed to the firm’s goals, which raises issues in relation to the motivation of professionals, the sharing

(14)

of economic rents, the MCS and the overall governance of the firm (Somaya & Mawdsley, 2015). Maister (as cited in Somaya & Mawdsley, 2015) explains strategic management of PSF as following:

“Management of a professional firm requires a delicate balancing act between the demands of the client marketplace, the realities of the people marketplace… and the firm’s economic ambitions” (Page 2)

Maister (as cited in Somaya & Mawdsley, 2015) emphasizes the complex relation between client, professional and PSF. From the PSF perspective it is important that management give enough strategic direction to professionals to make sure the firm and professional objectives are in line. Von Nordenflycht (2010) refers to this challenge, the management of professionals, as ‘cat herding’. The challenge arises from an intellectually skilled workforce to retaining and directing those skilled professionals. Additionally, Somaya and Mawdsley (2015) mention that employee’s productivity can have more of a positive influence when the firm’s goals and the goals of employees are aligned.

2.2 Client interaction

2.2.1 Client relationship

Client relationship is an important factor in the classification of the PSF and classified as one of the main assets of PSFs. As mentioned before the following dimensions emphasise, the relationship between client and PSF, client contact time, customization, front office and product. As the client takes part in the service production, it is expected that there is a high degree of contact with professionals, especially with the fact that professionals and client relationships are the PSF’s main assets (Greenwood et al, 2005). In several studies the degree to which the client is part of the service process is postulated to have the most profound impact on service operations issues (Chase, 1981; Mills and Margulies, 1980; Chase and Tansik, 1983; Lovelock, 1983; Chase and Bowen, 1989; Wemmerlov, 1990 as cited in Kellogg & Nie, 1995). The professionals and client work together to define, produce and deliver the service (Kellogg & Nie, 1995), considering that the client must help define service requirements.

PSFs have different characteristics that can be related to client relationship, such as: the nature of the product (i.e. service is intangible), the inseparability of production from consumption (e.g. complexity), the client’s involvement in the production of services (i.e. input knowledge) and the perishability of services (i.e. specific time related product). Broschak (2015) argues that it is necessary to get a better understanding of the nature and dynamics of client

(15)

15

relationship to understand how the PSF is organized, managed and it performs. PSF’s client relationship is a critical strategy asset and is an indicator of the market strategy of the PSF (Broschak, 2015). For instance, firms need to be more flexible and prepared for the challenging and demanding client expectations, as this will ensure a good company performance.

Therefore, some studies argue that PSF’s client relationships are more relational compared to customer relationships of manufacturing companies. Relational client relationships tend to have a long term focus, involve regular and extensive interaction between professionals and client, and are based on concepts such as attraction, trust and commitment (Halinen, 1997 as cited in Broschak, 2015). It is suggested that trust in inter-firm relationships includes a set of expectations between the partners regarding each other’s behaviour and each partner’s fulfilment of its perceived obligations in light of such anticipation (Madhok, 1995 and Thorelli, 1986 as cited in Aulakh et al, 1996). The relational relationship of PSFs and clients is based on the fact that the output of a PSF is produced through coordinated efforts of the professionals and the client. This implies close contact and ongoing cooperation between these two parties to be able to deliver the service. Taking into account the ongoing cooperation and close contact can be suggested that the client and PSF relationship needs to be relational (Broschak, 2015). Broschak (2015) suggests that the relationship between the client and professional is characterized by information asymmetry and uncertainty.

Asymmetry of information between the firm and its clients makes both parties dependent on each other (Greenwood et al, 2005), in this specific case the professional is dependent on information provided by the client to be able to deliver the service and the client is dependent on the service provided by the professional. Broschak (2015) refers to information asymmetry in the client relationship,

“as how informed parties are about each other in a contractual relationship, to the differing amount of information clients and professionals have that must be shared for the professional service to be performed, and to the fact that clients and professionals may have different understandings of information that needs to be shared” (page 4)

Professional services are produced in close interaction with clients requiring customized solutions by offering a high-customized product and with high customer contact, as classified by Silvestro et al (1992). In the production process of service information asymmetry takes place between the firm and its clients, making both parties dependent on each other, the client from the professional and the professional of the client (Greenwood et al, 2005). Information asymmetry can lead to opportunism from both parties, before and during the production process

(16)

and should be managed to minimalize it (Homburg & Stebel, 2009). Notwithstanding, that both parties, the PSF, providing the service, and the client, receiving the service output, would profit from the output of the service.

Further, Broschak (2015) refers to uncertainty,

“as the number of behavioural contingencies that are available to professionals in the conduct of their service and to the inability if clients to predict those behaviours and their consequences in advance.” (Page 4)

For instance, uncertainty arises before the contractual relation starts. The client is usually not aware of the capabilities of the PSF beforehand and mostly starts a relationship with a well-known PSF (Empson et al, 2015). Consequently, the client cannot predict the behaviours of the professionals and the outcome of the service. These characteristics of PSF’s client relationship should be mitigated by having a relational relationship, based on more long-term focus with extensive interaction and with a high level of trust and commitment. Furthermore, PSFs have to secure their reputation by making sure that the employed professionals are high quality experts, which are qualified, experienced and well interpersonal skilled (Kellogg & Nie, 1995).

2.2.2 Client interaction

The client being part of the service process has been addressed in different studies with different terminology, as customer contact, customer interaction and customer participation. What underlies these terminologies is that client presence, interaction and/or participation, in some way influences the service process. Homburg and Stebel (2009) describe client-interaction as the degree of integration of the service. Thus, the more input is needed from the client to realize the service, the more interactive the relationship between service provider (i.e. professionals) and customer is. This study is based on Homburg and Stebel (2009) definition of client interaction. Normally, the client will have an influence on both the cost of the service process and its output (Homburg & Stebel, 2009). In their research they defined two constructs to defined client-interaction, verifiability and integrativity (Homburg & Stebel, 2009).

It is often difficult to define verifiable performance measures for evaluating the service output, due to the characteristics previously mentioned, intangibility and complexity (Homburg & Stebel, 2009). As verifiability of the service output is mostly subjected to the client’s opinion and expectation, notably due to the fact that they have a high involvement in the production of the service. The client provides the basis of the input to the professionals for the productions of the service, such as explanation of the needs and problems, details of the requirements and

(17)

17

situation. Continual feedback from the client that determines selection, sequence and timing of services operations and delivery (Kellogg & Nie, 1995). Von Nordenflycht (2010) refers to ‘opaque quality’, to situations where the quality of a professional’s output is hard for no experts (i.e. clients) to evaluate, even after the output is produced and delivered.

Further, Homburg and Stebel (2009) refer to integrativity in their research, as the level of customer involvement. Integrativity means that the agreed service cannot be delivered without the integration of external factors, known as the client. PSFs, such as management consultancies, are characterized by a high level of client interaction (Homburg & Stebel, 2009). The client would define the what, how and when of the service delivery (Kellogg & Nie, 1995) or even disrupt the process of the service. Due to the integrativity of the service, the efforts of both parties (i.e. client and professional) are assumed to be complementary and necessary for the production of the service.

Concluding, that clients are highly dependent on the PSF because of asymmetry of information. Professionals are highly dependent on the client’s input. And, PSFs are dependent on their professionals because this critical asset is mobile (Greenwood et al, 2005). Some literature argues that due to the complexity and intangibility of the service, clients become dependent of professionals (Greenwood et al, 2005). The professionals have an advantage, due to the fact that professionals have expertise that the client does not have.

2.3 Management Control Systems (MCS)

Management control is about making sure that organizations reach their objectives. It is important for an organization to have ‘good’ MCS, to avoid thefts, fraud, unintentional errors, misalignment of goals and maintaining motivated employees. Keeping employee motivated and aligned with the firm’s goals are important issues for PSF, as their service is an intangible product based on human resource.

MCS know many definitions and there have been some discussions about it (Chenhall, 2003, p. 129). The following definition has a clear focus on ‘control’:

“A combination of control mechanisms designed and implemented by management to increase the probability that organizational actors will behave in ways consistent with the objectives of the dominant organizational coalition” (Abernethy and Chua, 1996, p. 573)

According to Merchant and van der Stede (2012), management control is used to steer behaviour throughout organizations towards organizations goals. MCS are thereby the tools by which this is achieved (Merchant & van der Stede, 2012). It can be expected that MCS can be adapted or

(18)

influence to better suit the steering of behaviour under uncertainty, in this study the output uncertainty of professionals.

There are different MCS models available in the accounting literature on how MCS design can differ. So, has Simon (Simon as cited in Ferreira & Otley, 2009) developed a framework known as the Levers of Control. Simon’s framework is based on the following concepts core values, risks to be avoided, critical performance variables and strategic uncertainties; these concepts are translated into four management controls, known as belief system, boundary system, diagnostic control system and interactive control. Alternately, Merchant and van der Stede (2012) have also defined other MCS. These controls are result control, action control, personnel control and cultural control. This research is based on Merchant and van der Stede terminology, as they believe management control is used to steer behaviour and believe people are the main driver in making things happen in organizations, in this case refers to the professionals of PSF.

Agency theory considers the optimal contract form for that control relationship in which one person, the principal, delegates work to another, the agent (Eisenhardt, 1985). Considering the relationship between employee and employer, the employee acts as an agent and the employer as principal. The agency problem is to determine the measurement and reward structures for the employees such that they appropriately balance their behaviour and company goals. It is a concern that problems can exist in agency relationships due to unaligned goals or different aversion levels to risk. Agency theory maintains that firms can avoid high monitoring costs by ensuring appropriate incentives are in place to motivate and align employees (Somaya & Mawdsley, 2015). Eisenhardt (1985) argues that agency theory suggests two underlying strategies of control; behaviour based, and outcome based. Both strategies rely on performance evaluation. Additionally, agency theory has two key insights. Firstly, the role of uncertainty in the choice of control strategy, through its impact on risk sharing cost. In the case outcome based control is in place, this would mean that the agent, i.e. employee, will also carry some risk, as there are uncertainties that come along. Secondly the role of information systems in maintaining behaviour control as an alternative to outcome control when information about the agent’s behaviour is otherwise incomplete (Eisenhardt, 1985).

Properly designed MCS influences employee behaviour in desirable ways and consequently increases the probability that the organization will achieve its goals (Merchant & Stede, 2012). Management control focuses on execution that involves the behaviour of employees. It can be argued that MCS design can differ across organizations and how each type

(19)

19

of control is applied. Also, the ownership form has also an influence in the MCS design of a company. Von Nordenflycht (2010) suggests that firms with high share of equity allocated to outside investors, require firms to be more centralized and formalized in order to maintain greater control. In other words, the MCS design could be more formally organized. On the other hand, an absence of outside investment allows management to allocate a firm’s equity or the compensation of it to employees with the aim to align each other’s goals. More importantly, this allows firms to adopt more autonomy and informality to better satisfy employee preference (von Nordenflycht, 2010). For example, accountancy firms make use of the ‘partnership’ control mechanism, where partners organize ownership of the organization; they have an active participation in the government of the firm and take part of the organization’s profits (Greenwood et al, 2015).

2.3.1 Result controls

Merchant and van der Stede (2012) argue that pay-for-performance is a powerful way to influence behaviour in organizations and called it an effective motivator. The pay-for performance can be seen as a result control because it rewards employees for delivering good results; good results can be referring to results that are in line with the organization’s goals or create company value. Result controls are based on performance measurement and rewarding employees for the delivered work. The rewards of employees are not just monetary compensation; it can be job security, promotions, autonomy and recognition (Merchant & van der Stede, 2012).

Result controls do not prescribe the actions or decisions employees should take; instead these should empower employees to take actions or decisions they believe will be produce the best result for the organization. For instance, the rewards linked to results inform employees of which results are important and motivate them to work on results that the organization rewards (Merchant & van der Stede, 2012). Thus, actions are indirectly influenced and give employee enough autonomy. Result controls applied in organizations can be evaluation based on performance (i.e. sales volume, percentage of satisfied customers, gross margin), appraisal based on performance outcome, utilization rate performance, efficiency and count of hours. Merchant and van der Steve (2012) mention that the organization goals should be known before hand and be well defined. Also, the performance measurement should be objective and accurate.

In the law industry there is the up-or-out promotion control used (Malhotra et al, 2010), this can be categorized as a result control as it is based on performance. The outcome of the performance of a certain time is measured and assessed. If the output is good, then the

(20)

employee would be rewarded with a promotion. On the other hand, in the case of the output not being good enough or not in line with the organization goals and/or budget then the employee would be sacked. In other sectors such as engineering consultancy, this up-or-out promotion management control is not practised.

2.3.2 Action Controls

Another type of MCS used is action controls that ensure employees perform in a way known to be beneficial to the organization. This type of control is only effective when management can define the actions that employees should follow and have the ability to ensure that the desirable actions occur (Merchant & van der Stede, 2012). This type of control can be seen as the most direct form of management control because they involve steps to influence the behaviour and actions of employees directly for the best interest of the organization. To be able to monitor, assess and reward employees, action controls consist of standard procedures, (production) processes, budgets, formal templates, instructions, rules and routines. This type of management control can be seen as a formal control as there is prescribed procedures beforehand.

2.3.3 Personnel and cultural controls

Personnel controls are designed to make it more likely that employees will perform the desired tasks satisfactorily on their own. And cultural controls are to shape organizational behavioural norms and to encourage employees to monitor and influence each other’s behaviour (Merchant & van der Stede, 2012). Other literature often refers to these types of controls as input controls, clan controls and soft controls (Ouchi, 1979). These management controls are based on self-control and social behaviour. Also known as social self-controls. Concrete personnel and cultural controls can be employee selection, shared norms and values and training. Making sure that the ‘right fit’ person is employed would make sure they are in line with the organization goal and have the required skills to represent the organization. Further training would help employees to develop skills that are required to generate the desired behaviour and actions for the organization’s best interest, e.g. think about certifications for engineers or accountants.

2.4 Hypotheses formulation

Considering the above theoretical framework this research studies the influence of client interaction on managing professionals in PSF. As explained before MCS are tools used to influence the employee’s behaviour to make sure the organization goals are achieved. The type of MCS differs between organizations depending on the available information and management. It

(21)

21

is argued that the nature and characteristics of the product, i.e. services, and the client relationship do influence the MCS choice of an organization.

The production of the service depends on the client’s experience, expectation and input. Normally in PSF, the client provides significant amount of necessary information, i.e. input knowledge regarding the organization. Besides, they can influence the cost of the product, the quality, quantity and the process. Von Nordenflycht et al (2015) emphasize in their literature review the power of the client to exert pressure on the professional in order to get the results they want. The client should not just be considered as a source of information but also part of the process of production. Due to the integrativity of the service the efforts of both parties (i.e. client and professional) are assumed to be complementary and necessary for the production of the service. Considering the client involvement in the production of the service gives a degree of complexity to management. A high integrativity of service implies that the client involvement is high. It is argued that this high integrativity will complicate the management of professionals in PSF. As professionals would be more concerned to deliver the service and maintain the client satisfaction. And not be focused on the realization of the PSF’s goals.

Furthermore, services are intangible outputs encoded with complex knowledge. This gives an uncertainty of the output. Normally it is challenging to define the output before the contractual relation starts and during the production of the service. As well, it is difficult to observe the professional’s behaviour. Therefore, it is often not possible for clients to evaluate or measure the output of a professional (Broschak, 2015). Thus, it is difficult to define verifiable performance measures for evaluating the service process and output as it is notably subject to the client’s specific requirements. Eventually, high client interaction influences the verifiability of the service output, as they are in constant involvement with the service production and they are the only part that can evaluate or be satisfied with the output of the professional.

Organizations with high client interaction will require MCS that give professionals a high degree of control and freedom over their work. First, the high demand and requirements from the client. Second, as professionals embody the firm’s knowledge and control the client relationships; resulting in a desire of considerable autonomy in their work. The professionals must have flexibility to adapt to an individual client’s varying needs and facilitate the customer influence. The professional must request client ideas, interact with the client dynamically, analyse the client’s inputs and finally use critical thinking skills to make judgment. The consultant acts as the expert and seldom are two consulting clients identical; the service is identified as a highly customized one of-a-kind service which is created with a great deal of input from the client.

(22)

Without the inputs and feedback from the clients, it is hard to achieve the delivery of the service. A high level of customer interaction brings great variability and unpredictability into the service process (Kellogg & Nie, 1995). Hence high client interaction has a high integrativity that makes verifiability of the service difficult.

One example of MCS commonly used to manage professionals in accountancy is the partnership, where the ownership of the firm is vested in a relatively large group of professionals (i.e. partners) who assume responsibility of governance and share the firm’s profit (Greenwood et al, 2005). The need to design MCS within a PSF are extremely important to make sure the professionals are managed, motivated and making sure the firm’s goals are achieved. This is mostly due to the complexity of the product, the influence of the client and the professional desire for autonomy. This study excludes the ownership management control mechanism and focuses only in the options that Merchant and van der Stede provide.

Considering the high uncertainty of the professional output, the professional desire of autonomy and the client involvement in the production of the service, it is suggested that the optimal MCS design should be less bureaucratic with little emphasis on formal instruments of control. For instance, MCS should be focused on result outcome, informal, loose, flexible and interpersonal controls. As with the uncertainty of the service variability and the client integrativity it is not accurate to define actions and result before the production starts. In contracts a strict MCS such as more focus on procedures, non-flexibility in budget and hierarchy would have a bad influence on professionals, as their perception would be that they don’t have the right of autonomy and freedom to organize and manage their work. In sum, taking into account the high client interaction and autonomous professionals, the overall prediction is as follows: given the uncertainty of the output, the high involvement of the client and the desire of autonomy of the professionals the optimal MCS design is focus on result, less in procedures and create a base of interpersonal culture in PSF.

As mentioned previously, it is argued that managing professionals is more difficult due to the high client interaction. As professionals are constantly interacting with the client and the quality of their output depends on client satisfaction. It makes it difficult for the management of PSFs to measure the performance of a professional as all verifiability lies on the client. In this case it is more difficult to define the actions that professionals should take to act in the best interest of the organization and add further value. It can be argued that the output can be quantified in the financial result of the service. Quantifying results, such as count of hours, revenue or gross margin per client, would say something about the performance of the

(23)

23

professional and financial result. Having result controls in place, will give professionals the desire autonomy and freedom to organize their work and be only concerned with the result. Taking into account the high client interaction and the autonomous professionals in PSFs it can be argued that the MCS of PSFs are positive related to result controls. Based on these observations the hypothesis is:

Hypothesis 1: High client-interaction is positive related with result controls in PSF

Based on the nature of the product (i.e. service is intangible), the inseparability of production from consumption (e.g. complexity), clients are also involved in the production of services (i.e. input knowledge) and the perishability of services (i.e. specific time related product). It is considered that each sector within PSFs is different and requires a different MCS. Consultancy and engineering firms are organizations where desired behaviour is more difficult to capture due to the considerable customization of the product. For example, a solution is customized to the specific client requirements, where the actions taken to achieve the results are more dependent to the client input and participation. In contrast, it can be argued that accountancy firms desired behaviour or/and the required procedure is already prescribed. For example, accountants have a prescribed list of the required input needed from the client. For this reason, the management of accountancy firms is less focused on result controls and expects that professionals carry less risk over the client involvement in their outcome. The following sub hypotheses ensues based on the previous observations:

Hypothesis 1a: The relationship between client-interaction and result controls is more positive for consultancy and engineering firms

Hypothesis 1b: The relationship between client-interaction and result controls is less positive for accountancy firms

The expected interactions of these hypotheses are illustrated in figure 3.

(24)

Secondly, autonomy in PSF is associated with relatively low levels of managerial intervention. In Silvestro et al (1992) classification it is also remarkable that the discretion is high, meaning that professionals use their own judgment to solve client’s problems. The high integrativity, where the client has a high involvement in the process of the service delivery makes it difficult to define the desirable actions of the professionals. Also, due to the desire of autonomy by professionals it is logical that professionals are not willing to follow prescribed procedure. Based on these observations the proposition is that high client-interaction in PSF is negative related to actions controls due to the autonomy desire of professionals and the high integrativity.

Hypothesis 2: High client-interaction is negative related with actions controls in PSF

Based on the nature of the product (i.e. service is intangible), the inseparability of production from consumption (e.g. complexity), clients are also involved in the production of services (i.e. input knowledge) and the perishability of services (i.e. specific time related product). It is considered that each sector within PSFs is different and require different MCS. There would be a discrepancy in the choice of MCS between different industries in PSFs. For instance, accountancy firms always follow the same protocol and always require the same information from their client for an audit. Accountancy firms would have service operations standardized and mostly routinized, such as financial audits, procedures, policies, checklist and defined process. The client interaction remains high, as they are the ones responsible for the input, such as with balance sheet specifications and contracts. Also, some accountancy firms (also consultancy) have created modules, which the consultant merely assembles as they are needed for client necessities (Kellogg & Nie, 1995). Besides, autonomous professionals appreciate the structure provided by the action controls. Professionals can experience action controls as a helping tool, as the desirable actions and procedure are defined.

On the contrary, in consultancy and engineering firms the product is more complex and customized, requiring high integrativity in the service and higher autonomy from the professionals; where actions controls are experienced as limiting and boundary in their work. Based on this observation it is expected that high client interaction is negative related to actions controls. The following sub-hypotheses are based on the previously observations:

Hypothesis 2a: The relationship between client-interaction and action controls is more negative for

consultancy and engineering firms

Hypothesis 2b: The relationship between client-interaction and action controls is less negative for accountancy firms

(25)

25

The expected interactions of these hypotheses are illustrated in figure 4.

Figure 4: Interaction of hypothesis 2 illustrated

Some researchers have pointed out the importance of normative controls rather than bureaucratic controls as the basis of control of professionals (Greenwood et al, 2005). Personnel controls focus on the types of people employed and their shared norms and values, as the success and performance of the PSFs relies on the expertise and experience of the professionals. Therefore, special attention needs to be paid to hiring, training and retention of employees. Formal recruitment processes to employ new professionals should find the professionals that ‘fit’ in the organizational goals, culture and match jobs with their skills (Kellog & Nie, 1995), and know how to deal with high client interaction. Consequently, as the desired actions are not easily observable and defined due to the high client interaction, high integrativity and difficulty in variability of the output, PSF have strong personnel and cultural controls in place. A PSF aim is to hire the correct/best professional that represents and sells their service and make sure their professionals are certificated and trained. Through personal and cultural controls professionals understand their work and its importance through the enculturation received when becoming a professional (Kellogg & Nie, 1995) and part of the organization. The following hypothesis is based on the previous observation.

Hypothesis 3: High client-interaction is positive related with personnel and cultural controls in PSF It is considered that all PSFs should have strong personnel and cultural controls in place as they deal with professionals. As all PSFs should have clear organization norms and values; training in provided for professionals to develop their skills. In this case no interaction is expected through the sectors of PSFs.

(26)

2.4.1 Conceptual Framework

Below conceptual framework displays the interactions and relations expected in this study. Client Interaction is the independent variable and the three different MCS types, the result controls, action controls and personnel/cultural controls, are the dependent variables. There are some interactions expected along the sectors of PSF and for this reason the sector accountancy and consultancy are added as moderated variable. See section 3.5 for a detailed explanation of the constructs.

Figure 5: Conceptual framework

Result controls Action controls Personnel/ Cultural controls

Professional Service Firm + (H1) - (H2) + (H3)

Consultancy + (H1a) - (H2a)

Accountancy - (H1b) + (H2b)

Note:

(27)

27

3 Methodology

This research studied the influence of client interaction on MCS of PSFs using survey data collected during a period of one year. The study is classified as a quantitative study. Quantitative research examines statistical relationships on the basis of predetermined hypotheses. The hypotheses within a quantitative study are derived from a theory and are tested against the research data. An in-depth explanation of the empirical approach will be provided by the following paragraphs.

3.1 Survey

The Accounting section of the University of Amsterdam is investigating the management of teams of professionals in PSFs. The PSF Practice Management Survey Project investigates the trade-offs of PSF, such as dealing with client demand and managing highly educated professionals. This research is based on a survey conducted online via Qualtrics by UVA to experienced managers of professionals.

The survey contains the following constructs: innovation, client relationships, knowledge sharing, team learning behaviour, task characteristics, client interactions, personnel control, autonomy, decentralization and performance assessment (see Appendix A for the complete questionnaire). Also, it contains control variables (contingency variables), e.g. age, management experience, size and country. These elements allow investigation of a variety of issues of PSFs that cannot be tackled by publicly available data. The questionnaire is developed based on relevant literature by Dr S.P. van Triest, as project coordinator.

To be part of the project the participating researcher needs to provide at least six completed questionnaires from suitable respondents by a certain period. This requirement is to make sure to obtain a substantial amount of research data. Upon doing so, the extended response set is made available to the student.

3.1.1 Validity test

The constructs are based on existing constructs from other literature and researchers. This is made to guarantee the validity of the survey. Further to this, the PSF Practice Management Survey Project is currently running a validity test. This validity test is performed by a number of professionals that are managed by the same manager. It is expected that the compared answers of the professionals are consistent, as each member should have had the same experience of the MCS. The outcome of this validity test is still unknown, due to the immaturity of the project.

(28)

3.2 Respondents

The respondents consist of managers of teams of professionals. A suitable respondent has experience in managing a practice, with responsibility for professionals and projects within the practice. The settings have a high level of interaction with the client, such as management consulting, IT consulting, engineering or financial services (van Triest, 2018).

3.3 Dataset

The complete response set, which was gathered by all participants in the project, contained 199 respondents; this set is called the ‘raw data’. From those 199 respondents, 36 respondents did not complete the questionnaire. These respondents were deleted from the raw data. This led to a first reduction of the sample size to 163 respondents. Furthermore, the input of question

‘managers experience’ 1_{should be filled in numbers. However, there were some respondents filled it}

in letters. These errors in the input were corrected to numbers. This does not have any impact on the count of the sample and the reason behind is to make sure the data is filled in the correct format, so it is recognisable for the analysis.

The questionnaire covered different PSF’s sectors2_{, of the 30 options included Other}

and the option to fill in manually what type of sector the practise is. From the 163 respondents, 15 were filled Other and 11 of these had filled in the type of sector. These 11 respondents could be allocated to one of sectors provided, for example respondents that it stated Audit were allocated to Accounting. And others that stated Corporate service and Service Management were allocated to Consulting Management/Strategic. Of these 15 Other sector, 4 were not able to be allocated to a certain sector and were deleted from the sample. The reason behind is the complex and specific nature of PSF, where a group of ‘other’, that cannot be allocated, is included in the analysis and might not belong to PSF. This led to a final reduction of the sample size to 159 respondents. This sample data has been used for the analysis. The three biggest sectors in the sample data are Consulting Management/Strategy, Accounting and Consulting IT, the count of respondents 43, 37 and 16, respectively. The remaining 63 respondents, 39,6% of the sample

1_{Question formulated in questionnaire as following: ‘How many years have you managed within your current}

practice? Please enter the number of years.’

2 _{The 30 options were: Accounting; Actuarial services; Advertising; Architecture; Biotechnology; Consulting}

Engineering; Consulting IT; Consulting HR; Consulting Management/Strategic; Consulting Technology; Engineering; Fashion design; Financial advising; Graphic design; Insurance brokerage; Investment banking; Investment management (hedge funds, VC, mutual funds); Law; Marketing/public relations; Media production (film, TV, music); Medicine/Physician practices; Pharmaceutical; Project management; Real estate; Recruiting - executive; Research/R&D; Risk management services; Software development; Talent management/agency; Other.

(29)

29

data, are spread over 19 PSF’s sectors. The distribution of the sample data among the sectors is displayed in below table.

Table 2: Sector distribution of sample data (159 respondents)

3.3.1 Dataset characteristics

As shown in table 3, 40,9% of the respondents are in the age between 25 -34 and 81,1% of the respondents are male. Further, respondents have average 14,9 years of working experience and 6,5 years of management experience. The study is conducted from the Netherlands and United Kingdom; however, some participants have gathered respondents from other nationalities. 54,7% of the respondents are from the Netherlands and 13,2% from United Kingdom of GB & NI. Count Percentage Consulting Management/Strategic * 43 27,0% Accounting ** 37 23,3% Consulting IT * 16 10,1% Consulting Technology * 7 4,4% Project management 7 4,4% Financial advising 7 4,4% Software development 6 3,8% Law 5 3,1%

Risk management services 5 3,1%

Marketing/public relations 4 2,5%

Engineering * 3 1,9%

Real estate 3 1,9%

Consulting Engineering * 3 1,9%

Media production (film, TV, music) 2 1,3%

Consulting HR * 2 1,3% Recruiting - executive 2 1,3% Investment banking 2 1,3% Actuarial services 1 0,6% Pharmaceutical 1 0,6% Talent management/agency 1 0,6% Research/R&D 1 0,6% Advertising 1 0,6% Notes

* Combined in one group and referred as 'Consultancy sector' ** Referred as 'Accountancy sector'

(30)

Table 3: Respondent characteristics

Additionally, 49,7% of the organizations has more than 5000 employees working and 42,8% have more than 100 clients. As mentioned before the study was conducted in the

Respondent characteristics

Gender Count Percentage %

Female 30 18,9%

Male 129 81,1%

Total 100

n 159 159

Age Count Percentage %

< 25 1 0,6% 25 - 34 65 40,9% 35 - 44 48 30,2% 45 - 54 31 19,5% 55 - 64 12 7,5% 65+ 2 1,3% Total 100 n 159 159 Mean SD* Median

Years working experience 14,9 9,7 13

n 159

Years management current practice 6,5 6,2 4

n 159

Nationality Count Percentage %

Netherlands 87 54,7%

United Kingdom of GB & NI 21 13,2%

Germany 9 5,7%

Canada 7 4,4%

India 4 2,5%

Portugal 3 1,9%

Spain 3 1,9%

United States of America 3 1,9%

China, People’s Republic 2 1,3%

Russia 2 1,3% Cyprus 2 1,3% Australia 2 1,3% Austria 1 0,6% Japan 1 0,6% Czech Republic 1 0,6% Colombia 1 0,6% New Zealand 1 0,6% Saudi Arabia 1 0,6% Vietnam 1 0,6%

China, Hong Kong 1 0,6%

South Africa 1 0,6% Greece 1 0,6% Thailand 1 0,6% Brazil 1 0,6% Italy 1 0,6% Georgia 1 0,6% Total 100 n 159 159 *SD: Standard Deviation

(31)

31

Netherlands. The headquarters of these organizations are mostly in the Netherlands, by 31,4% and followed by 27% of United States of America, see appendix B.

3.4 Data analysis

In this research, client interaction will be used as the independent variable and MCS as dependent variable. Also, the sector of service organizations, such as management consultancy, field service and bank corporates among others will be used as the moderator control variable to test the contingency of MCS. The constructs are subject to a reliability and validity analysis in SPSS. Reliability means that a questionnaire should consistently reflect the construct that it is measuring (Field, 2009). These analyses assess the quality of the constructs based on Cronbach's Alpha and factor analysis. Cronbach’s alpha checks the mutual correlation of the items of a construct. The value for a Cronbach's Alpha is between 0 and 1. A strong correlation suggests a very reliable construct. The rule for a reliable alpha is an alpha of at least 0.5 or preferably above 0.7 (Field, 2009).

Subsequently, a Principal Component Analysis (PCA) factor analysis shows the underlying patterns and correlations between the different parts. Using factor analysis assesses the convergent and discriminant validity. Thus, it assesses the communality in item variance from the viewpoint that this variance is in part caused by an unmeasured variable that has an effect on all the items. The factor has a value between 0 and 1 and a factor loading of at least 0.7 is required. The main underlying factors show how much variation the factors explain. The number of factors is determined based on the Eigenvalue. All the constructs have been individually tested for convergent and discriminant validity. And combining all the constructs, a varimax principal component analysis was conducted to assess the discriminant validity.

Finally, before the regression model analysis was made, a robust test was done to see if the reliability of constructs would improve. The choice was to trim some data from a certain extreme. One requirement of the respondents was to have some experience of managing professionals. It was decided that respondents with less than 2 year management experience should be excluded from the data as they would not be fully aware of how the MCS of the organization are applied.

(32)

3.5 Operationalization and measurement of the constructs

The survey has different constructs as mentioned previously in section 3.1, for the complete

questionnaire see appendix A. Most of the questions are designed in a 7-point Likert scale3_{; from}

strongly disagree to strongly agree. A few other questions have a 6-point Likert scale4_which

measure the contact moments with the client, i.e. from hourly to monthly contacts. This research makes use of survey questions 42 till 61. Client interaction is the independent variable that has an influence on the dependent variables MCS, result, action and personnel/cultural controls, in PSF. Sector is used as the moderator variable as it is believed that due to the complexity and speciality of the service product the MCS are different for each sector of PSF. Further, some control variables are added to the analysis and this is further explained in this chapter.

3.5.1 Construct: Client Interaction

The construct client interaction deals with the nature of the service provided by the PSF. This relates to the participation of the client in the production of the service. Some constructs are based on already developed research of Homburg and Stebel (2009), such as verifiability and integrativity, and other are self-developed constructs by Dr. Sander van Triest, project coordinator. As explained in section 2.2.2, verifiability is the degree of the client participation to verify the output of professionals, and integrativity is the level of customer involvement in the service production. This construct consists of seven questions (42 till 48), where six questions are 7-point Likert scale and one uses a 6- point Likert scale, see table 4 for detailed information of the construct.

The client interaction construct had low reliabilities, Cronbach’s alpha of 0.159. This could be related to the manner of how client interaction is measured and interpreted. As client interaction is complex and depends on the nature of the service, this can be interpreted differently on each specific situation. This Cronbach’s alpha is improved to 0.357 by deleting items 43 and 47, see appendix C for specific items. Item 43 is about the resources that the client uses to evaluate and monitor the PSF work; it can be argued that the PSF cannot foresee and does not have enough knowledge on how the client is organized.

The convergent of client interaction construct including all of the items, 42 till 48, was tested and it reflects a factor loading on three factors. This also reflects the low reliability of the construct and the low Cronbach’s alpha of 0,159. By excluding items 43 and 47 the construct

3_{7-point Likert scale: (1) Strongly disagree; (2) Disagree; (3) Somewhat disagree; (4) Neither agree nor disagree; (5)}

Somewhat agree; (6) Agree; (7) Strongly agree.

The influence of client-interaction on managing professionals in Professional Service Firms