Words: 9695
Management Accounting Research 19 (2008) 324–343 Operation of management control practices as a package—A case study on control system variety in a growth ? rm context Mikko Sandelin ? Helsinki School of Economics, Department of Accounting and Finance, P. O. Box 1210, FIN-00101 Helsinki, Finland Abstract This empirical case study examines the operation of management control practices as a package in a growth ? rm context by paying particular attention to the couplings among cultural, personnel, action and results controls.
The analysis focuses on two different management control packages in the face of similar contingencies at different points of time. The paper argues that the functionality of a control package depends on internal consistency, speci? cally on the reciprocal linkages of design and use between a primary mode of control and other control elements. Moreover, it argues that control package variety is driven by the way in which the management responds to functional demands. Two different control packages are considered equi? al to the extent of limited operational complexity, whereas an accounting-centric control package is also suf? cient in the face of increasing levels of operational complexity. © 2008 Elsevier Ltd. All rights reserved. Keywords: Management control package; Control system variety; Internal consistency; Functional demands; Equi? nality; Growth ? rm 1. Introduction This empirical case study examines the operation of management control practices as a package in a growth ? rm context. While it has been acknowledged that ‘soft’ and informal modes of control typically characterize small ? ms (e. g. Bruns and Waterhouse, 1975; Merchant, 1981; Flamholtz, 1983; Chenhall, 2003; Merchant and Van der Stede, 2007), an increasing body of literature now suggests that growth renders management control systems (MCS), especially management accounting systems (MAS), more formal (Granlund and Taipaleenmaki, 2005; Moores and Yuen, 2001) and that formal MAS facilitate the growth of a ? rm (Davila and Foster, 2005; Sandino, 2007; Greiner, 1998). In the growth ? rm context, the more comprehensive management control package has been investigated to a limited extent.
Lukka and Granlund (2003) have paid particular attention to organizational culture and Collier (2005) has focused on socialization practices as primary control mechanisms, whereas other studies have examined MAS and formal standard operating procedures (Granlund and Taipaleenmaki, 2005; Davila and Foster, 2005; Sandino, 2007). The main thrust here is that an equally good ? nal state can be achieved by various control system designs in the face of similar contingencies (Huikku, 2007; Merchant and Otley, 2007; Ferreira and Otley, 2005; Gerdin, 2005; Spekle, 2001; Otley, 1999; Chapman, 1997; Fisher, 1995).
The need for coordination and control can be met by several alternative ? Tel. : +358 9 4313 8457; fax: +358 9 4313 8678. E-mail address: mikko. sandelin@hse.?. 1044-5005/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10. 1016/j. mar. 2008. 08. 002 M. Sandelin / Management Accounting Research 19 (2008) 324–343 325 management control system designs (Gerdin, 2005). Contingency theory, as it builds on the assumption that variables are related to each other in a one-to-one manner, seeks optimal control system designs in speci? circumstances at the cost of system variety (Spekle, 2001; Gerdin, 2005; Merchant and Van der Stede, 2006). Moreover, a long-held view in organizational design literature suggests that multiple means of control do not only complement each other but may also operate as substitutes (Galbraith, 1973; Mintzberg, 1983; Fisher, 1995; Abernethy and Chua, 1996; Ferreira and Otley, 2005; Huikku, 2007). The potential for achieving the same ? nal state by various con? gurations of control elements and systems in the face of similar contingencies is referred to as equi? nality (Doty et al. , 1993; Gresov and Drazin, 1997).
In other words, organization designers have the latitude to decide about how to achieve organizational goals. This notion of equi? nality has received only marginal attention in empirical management accounting and control studies. Two recent studies, however, suggest that equal control of activities can be achieved either by an informal control practice (Huikku, 2007) or by different formal control systems (Gerdin, 2005). This study goes further and adopts a holistic approach to management control and seeks to increase our understanding of the simultaneous operation of multiple control practices at the ? m level and addresses the different, potentially equi? nal, control con? gurations that they form. The motivation for adopting a holistic perspective on management control stems from the fact that empirical management control studies have produced unclear ? ndings and con? icting results because too few components have been studied at the cost of more comprehensive and integrative approaches (e. g. Merchant and Otley, 2007; Ferreira and Otley, 2005; Covaleski et al. , 2003; Otley, 1999).
It is not uncommon for management control studies to focus narrowly on formal systems, and speci? cally on MAS alone. Frequently, they also implicitly rest on cybernetic tradition and assume that management control elements operate as a system of planning, measurement, evaluation, and feedback for corrective actions. In this tradition, it has been argued that the linkages between formal MCS components, particularly the strength and coherence of the couplings, explain the functionality of a control system (Ferreira and Otley, 2005).
However, our knowledge about the interplay between accounting-based and more subjectively constructed control elements is in its infancy (Brown, 2005), although it is increasingly acknowledged that formal MAS are only part of a broader set of organizational controls (Flamholtz, 1983; Flamholtz et al. , 1985; Abernethy and Chua, 1996; Alvesson and Karreman, 2004; Collier, 2005; Merchant and Van der Stede, 2007; Merchant and Otley, 2007). Formal control systems, including MAS, are important mechanisms for reducing uncertainty, although they may have a modest effect in controlling actual work (Karreman et al. 2002). On the other hand, formal control practices may signi? cantly shape social order and provide organizational culture with material signi? cance and meaning (Alvesson and Karreman, 2004; Ahrens and Mollina, 2007). While the control elements may play multiple roles, subtle linkages seem to make them operate as a ‘package’ (Abernethy and Chua, 1996). To date, however, the couplings between control elements have been exposed to limited academic scrutiny (Brown, 2005; Brown et al. , 2008). An examination of the couplings between
MAS and other control elements therefore has the potential to contribute to our understanding of the features explaining the functionality of control packages, but also of the need for formal control practices such as MAS. Many alternative conceptual frames are available for studying control elements as a package. Those by Otley (1999) and Ferreira and Otley (2005) are grounded in the systems view of organizations and are intended for examining combinations of formal control techniques, or at least of practices mediating information, thus ignoring more subtle ways of motivating and coordinating organizational behaviors.
Also, Simons’ (1995) framework focuses on formal, information-based control practices and is primarily applicable to investigation of how managers use rather than design MCS in an attempt to control strategy. Moreover, the organizational control framework of Flamholtz (1983, 2005), Flamholtz et al. (1985) seems to lack speci? city although it is informative about the need to align organizational control elements. This study therefore draws primarily on Merchant’s (1998), Merchant and Van der Stede (2007) object-ofcontrol framework because it focuses on the spectrum of control practices, but still provides suf? ient rigidity with the speci? c control objects of culture, personnel, action, and results. Adopting a broad perspective on management control implies that organizational culture and structure are means rather than mere premises of control (cf. Flamholtz, 1983; Flamholtz et al. , 1985; Merchant and Van der Stede, 2007). While the mere existence of any organizational culture and/or structure affects organizational behaviors, they are also objects of deliberate managerial actions.
They are designed and changed to affect the behavior of organizational members in the attainment of organizational strategies and goals, irrespective of whether these strategies are ex ante planned or emergent (Mintzberg et al. , 1998). Hence, the distinction between culture and cultural control is that the 326 M. Sandelin / Management Accounting Research 19 (2008) 324–343 latter considers the means of shaping culture1 and the former the outcome that affects organizational behaviors. In addition, management control may nclude numerical measurement techniques, such as MAS, but also practices that are based on subjective judgments and evaluation (e. g. Alvesson and Karreman, 2004; Collier, 2005; Ahrens and Mollina, 2007). While Merchant’s object-of-control framework captures the richness of control practices, it lacks speci? city in explaining the couplings between control elements. As this study speci? cally seeks to examine these couplings, it only mobilizes Merchant’s framework as a medium for analysis. The remainder of the paper is structured as follows.
The next section describes the research site and method. The following section analyses empirical evidence by presenting two illustrative case examples of the management control package. These cases are grounded on the same ? rm at different periods of time. Each case description is followed by a short analysis of the couplings between control elements. The subsequent section discusses empirical ? ndings on the basis of comparative case analysis and the ? nal section draws the conclusions. 2.
Research method and site There is little prior theory of the linkages between formal and informal control elements. As the research objective is explorative in this sense, it favors the case study method at the cost of the less in-depth methods of survey studies. On the other hand, ? ndings about less orthodox forms of management accounting and control in the relatively small and entrepreneurial ? rms (e. g. Perren and Grant, 2000; Greenhalgh, 2000; Collier, 2005) encourage adoption of a sensitive, deep-probing research method that also enables the tracing of subjectively constructed forms of control.
Thus, a case study appears to be the most promising mode of enquiry because it makes possible a comprehensive approach to the study of controls in use (Otley and Berry, 1994; Ahrens and Dent, 1998). However, as this study examines two distinct periods of ? rm evolution, it relies on a comparative logic of analysis, which is likely to reveal subtle similarities and differences between cases and lead to more sophisticated understanding—to categories and concepts unanticipated by the investigator (Eisenhardt, 1989).
Thus, this mode of analysis has the potential to re? ne management control theory about the functionality and appropriateness of management control packages (Lukka, 2005; Keating, 1995; Scapens, 1990). Theme interviews with organizational participants and stakeholders are the primary data source. Data triangulation was sought throughout the data collection process. On the one hand, people were interviewed from various organizational levels and functions, and on the other, interviewees consisted of current employees and those who had already left the company.
Altogether, 22 semi-structured interviews were conducted among the key informants before the data became saturated. The data amounted to 34 h of tape-recorded interviews that were subsequently transcribed. While individuals tend to impose order retrospectively on phenomena, an attempt was made to eliminate this type of bias by letting informants discuss freely, by seeking facts rather than opinions in retrospect, and by re? ecting interviews on archival data2 . Moreover, con? rmation on individual ‘accounts’ was sought by posing similar questions to multiple informants (cf.
Cardinal et al. , 2004; Miller et al. , 1997; Abernethy and Chua, 1996). Only the general aspects of the study design were determined in advance and data collection was undertaken prior to the ? nal de? nition of research question because of the relatively unknown research phenomena (Yin, 1993). Hence, the data analysis was an iterative process between theory and data, in which patterns emerged and theoretical insights crystallized after several readings, particularly around three distinct periods along the evolution of the ? rm.
At the ends of the continuum control practices appeared as the most promising ones in the theoretical sense because the contingencies were similar but the designs of control packages varied to a large extent. In explaining this variation, the study relies on the interpretive mode of theorizing (Lukka, 2005). The research site of the study, Mobitel Ltd. 3 , was chosen because it was an entrepreneurial ? rm seeking growth on immature ‘high-technology’ markets. Inherent ambiguity and uncertainty of this type of ? rm (Granlund and Taipaleenmaki, 2005), as well as growth, were expected to characterize management control dynamic phenomena. Culture can be shaped e. g. by codes of conduct, group-based rewards, intra-organizational transfers, physical arrangements and tone at the top (Merchant and Van der Stede, 2007). 2 Archival data included IPO prospectus, investor presentations, interim and annual reports, second hand interview material in public press and the ? rm’s internal documentation and presentations. 3 The name is pseudonym in order to secure anonymity of the case ? rm. M. Sandelin / Management Accounting Research 19 (2008) 324–343 327 On the other hand, the relatively small size, although the ? m was growing, was expected to make the linkages between various forms of control more visible and pronounced than in a large and complex organization in which multiple levels of management hierarchy and bureaucratic processes tend to hide informal practices (Mitchell and Reid, 2000; Collier, 2005). The major events, contingencies, and performance data during the ? rm evolution are presented in the appendix. 3. Case study 3. 1. Control practices as an enabling platform—Case 1 The ? rm was established at the beginning of 1998 in the merger of three entrepreneurial ? rms.
At the same time, a new CEO with a managerial background in a large telecommunications company was hired and the CEOs of the merged ? rms took positions in the senior management. The ? rm had a steady cash-? ow business from Internet access services. As the ? rm grew, it hired a CFO at the beginning of 1999. Before the entry of the CFO, who had earlier worked for years in a pro? table family business and subsequently as a management consultant, an external agency took care of bookkeeping. The ? rm has been venture-capital backed since its foundation and it has sought to compete with the lowest price. . 1. 1. Cultural controls and culture From the beginning, the personnel were highly committed to the ? rm, as each employee owned a relatively signi? cant share of the ? rm. The interests of the organizational members and the venture capitalist coincided—each party wanted the ? rm to increase the shareholder value. The venture capitalist described the managerial freedom as follows: “We were not so interested in knowing what was going on within the ? rm. We were interested in entry and exit prices. ” In fact, the senior management had freedom to decide how to develop the ? rm.
The choice was made for growth instead of pro? tability, as a senior manager commented: “Growth was our priority but we also discussed pro? tability. ” As the ? rm grew and it hired new managers, they were made shareholders. New employees at the grass root level had a base salary, but they were also committed to the long-term development of the ? rm with share options. By means of broad ownership and group rewards, the senior management maintained an ‘entrepreneurial’ culture. The personnel were extremely committed to the ? rm. There was no real need to monitor working time. Personnel were willing to work even seven days per week and they were ready to cancel holidays if any business challenges occurred,” a line manager commented. The ownership-based entrepreneurial culture provided solid ground for running and developing the business because “everybody tried to seek the best for the ? rm in order to maximize the value of their personal ownership,” as one business unit manager stated. Shared values and beliefs culminated in technological advancement which guided managerial behavior, as another business unit manager recalled: “We made decisions on the basis of technology.
Once something was technologically possible, we developed a product rapidly and made it available to customers [in the portal]. We did not think in advance whether there was demand for such a product. ” Technological opportunities dominated managerial thinking. Financial and demand concerns were secondary, as the Chief Technical Of? cer (CTO) explained: “We developed and built things [products] which were extremely dif? cult to ? gure out how they could have generated cash [in-] ? ows. ” Managerial actions focused primarily on discovering new applications for existing technologies and developing new technological solutions.
On some occasions, however, the management planned technology development. In these exceptional cases, development efforts were driven by economic concerns in contrast to more general patterns of simply thinking and acting in terms of technological possibilities. The CFO illustrated a planned project: 328 M. Sandelin / Management Accounting Research 19 (2008) 324–343 “The propeller-heads [entrepreneurial managers] had an idea to develop a ‘service of all time’ with which customers could download e-mails onto mobile phone. There was to be a monthly charge for the service or a usage-based charge. . . . ] They developed ‘sms-gateway’ technology on that basis. ” Once the e-mail service was ready, only some pioneering customers ordered it. The service was too sophisticated at the time, despite of the weak demand signals detected by the ? rm. The sms-gateway technology seemed to be merely a costly project without commercial use. However, the information engineering staff had drawn logos and transmitted them via sms-gateway technology to their mobile phones during the technology development process. A line manager noticed these logos and presented them to the senior managers.
Soon the logos were available in the portal at a price of less than one euro per download. The CFO commented on the subsequent events as follows: “We were too slow to advertise it. Suddenly, without any advertisement expenses, rumor about the logos spread among customers and we sold almost a million logos in a month. ” The demand prospects for e-mail service and related ex ante planning of activities had little signi? cance in guiding ultimate behavior and ? nancial success. Instead, keen interest in technologies across the organization facilitated the emergence of logos.
And trying out logos as a commercial product was driven by the freedom of action that originated from the ownership. The culture guided action throughout the organization. In addition, action itself was a characteristic of the culture, as a senior manager described: “We were doers rather than thinkers. [. . . ] Some people are good at planning and making theory, but it does not help too much because they are unable to arrive at any conclusion. I mean that it is better to have little theory and to get things done than get nothing done. ” Action orientation was also re? cted in how the management decided to capitalize upon the logo business. The CEO explained the early steps in taking the logo innovation to the international markets at the end of 1999: “Instead of planning, we just ? ew over to check in which countries ? rms were technologically ready, where there was willingness, and where good guys were ready to proceed rapidly. ” Formal plans, budgets, and standard operating procedures were replaced by rapid and intuitive actions. By taking bold action, the senior management acted as role models for line managers and employees.
They shared responsibility and provided support to each other. All of this was based on the ownership culture, as the CEO explained: “Flexibility came from the ownership. Once you had own money in question, you were ready to be ? exible. Our backbone was like spaghetti bent to meet the situation at hand. ” Although organizational culture was the primary mode of control and it was deliberately shaped by means of ownership, incentives and the ‘role model’-like behavior of senior managers, the senior management also employed other control practices in an attempt to facilitate the pro? able growth of the ? rm. 3. 1. 2. Control practices complementing cultural control orientation Personnel controls were deployed from the beginning because growth constantly required more staff. As the ? rm lacked a human resource manager, the senior managers or the heads of the support functions recruited new employees. A senior manager commented on the evaluation of the candidates as follows: “People were mainly hired but ? rst they needed to be whipped and kicked a little, just like in a gang, as you’ve seen in Italian ma? a ? lms. That was the school.
It did not matter whether an employee had an academic degree or not. ” The CEO further clari? ed the criteria applied in the personnel selection: “I learned quite rapidly that track record and work experience were not appropriate criteria because those people were not able to adapt to our chaotic culture. ” The managers evaluated the candidates on the basis of subjective judgment. Although technological skills and knowledge were sought, the ultimate evaluation was conducted against the relatively chaotic organizational atmosphere where continuity and stability were replaced by bold and intuitive actions. The ‘old economy gurus’ were unable to M. Sandelin / Management Accounting Research 19 (2008) 324–343 329 adapt to these settings—perhaps the ? uidity stressed them,” a senior manager commented. Moreover, the management did not hesitate to ? re an employee if s/he did not ? t to the organizational culture. There was not, however, a need for signi? cant lay-offs. The systematic though informal deployment of cultural criteria in personnel selection and share-based incentives granted to employees homogenized the atmosphere within the ? rm.
A business unit manager described the personnel as follows: “The most positive character of our personnel was the willingness to work crazily. They had intrinsic motivation to develop new things [technologies and products]. [. . . ] This was the ? rst challenge for many of our employees. However, you were able to make them to exceed themselves time after time. ” Although the average age of the staff was about 27 years, the management considered it a strength rather than a weakness. The young personnel had no reference points about how matters should be organized in a ? m. Thus, they were apt to adapt to high velocity. The culture-driven mode of personnel control, subjective evaluation, was also mobilized in assigning tasks on the basis of job performance, as a senior manager commented: “In practice they ultimately showed what they were capable of. [. . . ] Of course we did not assign overall responsibilities to a guy who was better at details. ” The relatively young and technologically motivated personnel formed a coherent resource for the ? rm. However, the open-plan of? ce further encouraged culture-driven behavior.
Besides ef? cient information dissemination between organizational members, this particular workplace design facilitated the adoption of cultural values by providing a natural platform for socializing with other employees. On the other hand, the open-plan of? ce induced peer-control, as individual actions were transparent to others. Work was conducted in teams by which the management sought to enhance the sense of individual accountability to other members of the team. In the absence of team rewards, the accountability within teams was quite modest.
However, motivation and especially commitment were not particularly problematic, as the CTO described: “Even though the content of tasks changed signi? cantly, the employee turnover was low. ” The share-based modes of compensation and the challenging content of tasks provided suf? cient incentives for the personnel. Action controls operated upon the open-plan of? ce. The ? rm had been divided into four business units according to customers and services. Each unit had a responsible manager authorized to make operative decisions.
In these units centralized decision-making was sought by means of standard operating procedures, such as approval policies and pre-action reviews. These modes of control were informal, conducted as part of daily discussions in the open-plan of? ce. Moreover, the most important form of action control was personal supervision. The management, by ‘walking the talk’, assigned tasks and provided decision-aid to the teams, as a line manager described: “We had quite an army of hackers. . . You only needed to say to them ‘? nd out if this or that makes any sense. ‘ Then these guys worked on it for half a day or a day.
Then they said ‘this makes sense and it takes a week to copy it. ‘ After a short discussion I said, ‘ok, copy it. ‘ Then we [managers] started to build up a business plan on it. ” Operative planning and scheduling of activities prior to action were untypical for the ? rm. Neither were speci? c job instructions prepared. These control practices were seen as secondary because events unfolded at a rapid pace inside and outside the ? rm. The operative management let employees to take the initiative and coached them about how to proceed as new information was available. The line manager continued: “We had 20–30 projects going on . . Many of them proceeded so that we had worked on a project for about three weeks until we suddenly observed a better solution. We terminated the project and decided to copy an alternative solution. ” Line managers challenged teams to consider technological advancement, but placed less emphasis on the ? nancial consequences related to the new ideas. Business planning on technological ideas was a managerial responsibility. Financial evaluation of the new ideas and their ? t with other businesses was conducted in discussions between the line managers and the business unit management. Consequently, technological and ? ancial criteria for appropriate action were separated from each other at the operative level. On the other hand, while the senior management sought pro? table growth, it considered new product development to be crucial for the ? rm’s success. This critical function 330 M. Sandelin / Management Accounting Research 19 (2008) 324–343 was controlled, beyond personnel and cultural modes of control, by ensuring that there were enough employees and time available for developing technologies. The CFO commented on his perception as follows: “Somewhat similar issues were considered in multiple teams and that’s why we had so many employees. . . . ] Never had it [the number of employees] been considered primarily from the ef? ciency perspective. ” Besides new product and technology development, the actions related to business processes of a routine type were controlled by more formal means and automated whenever possible. However, irrespective of how formal action controls were, managers conducted subjective performance evaluation and provided constructive feedback, occasionally rewarding exceptionally good performance. On the other hand, at the top of the ? m, the management group formed an important body for coordinating and controlling actions, as the CEO described: “We shared and updated information regularly on each Monday morning. We had management group meetings for two hours and we discussed what was going on. ” The weekly management group meeting consisted of the business unit managers, the CEO, the CFO, and the heads of technology and operations. The action-reviews and ideas presented by the business unit managers were evaluated and discussed more broadly. Besides directly controlling business unit initiatives and actions, the management group had another control role.
The CFO described the management group as follows: “The size and composition of the management group varied constantly. ” A senior manager further clari? ed the rationale underlying the dynamic management group: “. . . in a larger scope [of the whole ? rm], one man would have not been able to control it. In this sense, the management model of constantly changing the composition of the management group was reasonable. ” To obtain a comprehensive understanding of actions within the ? rm, the senior management invited the line managers to attend the meetings. Knowledge was thereby circulated within the ? m. Due to possible personal limitations at the lower organizational levels, important pieces of information could have been missed despite the open-plan of? ce. Results controls were introduced by the CFO. The operative non-? nancial indicators, such as website downloads, time spent on website, and spoken gsm and ? xed line minutes per customer were monitored on a daily basis. These data, however, were mobilized primarily for learning purposes, as the CFO explained: “. . . we monitored the number of subscriptions, gross and net, in order to understand how to keep customers happy and loyal. The frequent monitoring of non-? nancial results directed managerial attention to the quality of services and the viability of new products and thus, it provided decision support for terminating or revising a product concept. Non? nancial indicators were not, however, standardized performance targets tied to compensation. Instead they were used in forecasting turnover. Nevertheless, the non-? nancial information was problematic in budgeting, as the CFO pointed out: “The general problem with them [non-? nancial indicators] was poor linkages with book keeping. It was damn dif? ult to ? gure out the ? nancial implications and linkages. ” It was unclear how the many products developed on the basis of technology push would generate revenues. The non-? nancial information of the product or service volumes (e. g. in the portal) was not directly associated with revenues and some products were based on termination fees that depended on changes in the telecommunication regulation. A particular challenge in budgeting was the assignment of overhead costs to product or service lines because the ‘true’ cost drivers were based on bytes (e. g. rented network capacity).
However, through trial and error the CFO learned to approximate the ‘hype’ in budgeted sales and underestimates in variable costs and also to allocate overheads to business units at the aggregate level. However, the budgets had little signi? cance in guiding or restricting behavior at the operational level, particularly in controlling costs, as the CFO explained: “Cost control, well . . . I had to take disciplinary action quite often, to yell loud enough. It worked for a while but then, I saw it so clearly, quite rapidly there were this and that, invoices not directly related to business . . coffee machines, new computers, software licenses and upgrades, extremely high traveling expenses, etc. ” The budgets brought ? nancial issues to the attention of senior management only, to whom the CFO also reported invoicing data, the number of new customers, and the cash balance on a weekly basis. This information emphasized the liquidity of the ? rm and set boundaries for growth. Although the senior management was exposed to ? nancial data M. Sandelin / Management Accounting Research 19 (2008) 324–343 331 and the CFO coached them about ? ancial planning, the budgets were perceived more as a means to shape culture than to control activities, as a senior manager described: “Budgets had two purposes: to nail down pro? t center [business unit] targets and hence, to enhance the internal spirit in terms of making the impossible possible. ” The budgets provided direction by setting targets, but had little signi? cance as control mechanism at the operative level because the degree of participation in budgeting was very low within the business units. Consequently, the line managers in the business units were loosely committed to the budgeted targets. However, udgets operated as a formal control system in the sales function in which ? nancial performance targets were set for the salesmen. Their performance was evaluated on a monthly basis and the achievements were linked to the formal reward system—extra cash bonuses. The CEO summarized the signi? cance of the formal control practices in managing the ? rm as follows: “My principle was not to rule out anything. [. . . ] You never knew what’s going to happen . . . you just sat and waited. [. . . ] We intentionally tried to avoid rules and formal procedures [in decision-making], but if they emerged anyway, we got rid of them. Many transactions, e. g. establishing the logo business in Norway, had proven to be a success though the plans and calculations were not formal—they were only in the minds of the senior managers. These implicit plans were realized through bold and intuitive actions. However, the ? nancial numbers were not totally insigni? cant—they were mobilized as rough guidelines and targets which fuelled growth aspirations, as the CTO recalled: “He [the CEO] said that we are not allowed to do things [businesses] from which the expected turnover is less than [x. x] million euros. Financial controls were employed at the aggregate level as mediums of directing attention and motivating managers, whereas operational activities within the business units were controlled by mutually reinforcing forms of cultural, personnel, and action control. 3. 2. Analysis of the Case 1 Scant ? nancial resources limited the growth of the ? rm in the beginning. The lowest prices and innovative products, however, resulted in high volumes and cash ? ows enabling growth—the headcount grew from one to two hundred. Soon after establishment, an external board of directors took over, but controlled the ? m loosely. The problem of developing and controlling the ? rm in the relatively complex environment was left to the discretion of senior management. The coherent organizational culture was largely based on the fact that ownership and control were not separated. The senior management created an ‘ownership culture’ (Merchant and Van der Stede, 2007) by means of direct share ownership and share options and thus sought to establish and commit employees to self-control. The culture of technological intra-entrepreneurship replaced speci? c task descriptions and comprehensive pre-action planning.
The role model provided by senior management further shaped the organizational culture by emphasizing bold and intuitive action. In addition to enabling relatively autonomous behavior, the organizational culture guided search and development actions and provided management with criteria for selecting personnel. Personnel controls, however, also reinforced the culture, as employees with an intrinsic technological motivation and ability to adapt to ‘unstructured’ organizational settings were hired. The relationship between these modes of control was reciprocal, the culture, however, having primacy.
The ‘hybrid’ controls, such as the open-plan of? ce, complemented cultural, personnel, and action controls. In the open-plan of? ce, daily operative actions were conducted in teams. These structures introduced action accountability, but they were essentially complemented by personal supervision. The need for personal supervision stemmed from the personnel selection and placement methods, but it also enabled the existence of these practices. In line with the organizational culture, the standard operating procedures, such as the approval policies, were part of the informal daily interaction.
Moreover, the coordination of managerial activities relied heavily on personal supervision and mutual adjustment in discussions, which also took place in the senior management group. The cultural, personnel and action controls had reciprocal linkages, which made them harmonious to a large extent. These controls enabled technologydriven new product development, which was considered critical to pro? table growth and thus, its performance was ensured and controlled by allowing several teams to conduct similar type of tasks. 332 M. Sandelin / Management Accounting Research 19 (2008) 324–343
Moreover, all except the results controls were based on broad ownership, which created an incentive for contributing to ? rm-wide performance. Financial results were accordingly measured and budgets prepared only at the aggregate level. On the other hand, while the ? rm primarily sought growth, it also monitored pro? tability at the senior management level. The ? nancial concerns were not deliberately extended to the operative level, as the aim was to preserve innovativeness. To sum up, non-? nancial results controls were only loosely linked with action controls as they provided decision support for product portfolio decisions, whereas ? ancial results were separated from other control elements in order to maintain the focus on technological advancement. However, the ? nancial performance indicators (see appendix) do suggest that the ? rm was successful in attaining its goals with this speci? c management control package. 3. 3. Prologue for Case 2 The ? rm went public in the late spring of 2000. With the IPO money, it continued to grow with increasing emphasis on international markets. By the end of 2000, it had established businesses in more than 10 European countries primarily by means of acquisitions that were mainly funded by share exchange.
At the end of 2000, the domestic business was pro? table; turnover was D 27 million, EBITDA was D 4. 5 million, and the operating pro? t was D 1. 7 million, although the ? rm had faced problems in taking over the acquired companies. As the events unfolded, it became evident that rapid growth in turnover was insuf? cient to recover takeover costs. The ? rm drifted into ? nancial troubles in the late summer of 2001. The senior management was replaced by a crisis management team. The crisis-time CEO and CFO sought to keep the ? rm a? oat by negotiating additional funding.
Finally these efforts culminated in voluntary debt-restructuring. The crisis-time CEO and CFO left the ? rm as their primary task of securing its survival had been accomplished. A new CEO and CFO were hired, both with many years of experience in management consulting. The ? rm continued to offer products and services at the lowest prices. The cash-in? ows were modest—the ? rm had lost reputation and customers during the crisis. In general, the organization was in a state of ? ux, as the CEO described: “. . . the organization was in a state of great uncertainty: from the perspective of management, responsibilities, and atmosphere. In other words, at the outset the new CEO and CFO faced a ? rm resembling that was presented in the ? rst case—there was little administrative infrastructure of any kind in place. The new management withdrew the debt-restructuring plan after a few months, as new venture capitalists invested in the ? rm. The following section focuses on the subsequent two-year period. 3. 4. Control package as a system of accountability—Case 2 During the acute crisis, business development had been less than modest. Various ‘ad hoc’ activities had been undertaken in an attempt to secure ? nancial survival.
The biggest individual shareholder of the ? rm, the venture capitalist who occupied the board with other investors, described the expectations imposed on the new CEO as follows: “Prospects for growth were promising, although competition was intensifying. We did not want to let the ? rm go bankrupt because we had invested so much money in it. We just wanted it to continue in some ways. ” Indeed, the venture capitalists did not set tight restrictions for the CEO. They were primarily interested in increasing the share price. Whether it would happen by means of growth or pro? ability or both was left to the CEO’s discretion. 3. 4. 1. Controlling by results The CEO and CFO, in co-operation with the business unit managers, analyzed the cash ? ows. While no ? nancial calculations were readily available, they estimated cash ? ows at the aggregate level and perceived the internet and teleoperator products and services as the most promising in ? nancial terms. The senior management perceived the core of the ? rm as an ‘invoicing machine’. Instead of emphasizing new product development, they focused on ef? cient delivery of current products that would secure a steady cash ? w. Among the secondary businesses was the more innovative mobile entertainment, which formed a separate business unit with promising European-wide operations. The rest of the businesses formed the third unit, which the management wanted to divest or terminate as soon as possible in order to reduce uncertainty and thereby establish simple premises for systematic management, as the CEO remarked: M. Sandelin / Management Accounting Research 19 (2008) 324–343 333 “The ? rm was organizationally unsteady [as the boundaries between businesses were fuzzy].
A clear separation of businesses from each other enabled some sort of management: assignment of managerial responsibility and personnel. ” Clear boundaries between business units established a foundation for development of a new control system. Because of the CEO’s preference for numbers, the control system was to be based on accounting practices. The idea was to establish a broad ? nancial control system by which ? nancial performance targets would be extended not only to business unit managers, but also to operative managers. All the existing cost centers were discarded, new cost and pro? centers were formed, responsible managers were appointed, and the accounting manual was fully renewed. While the accounting function developed ? nancial control mechanisms, the CEO made decisions on the basis of rough ? nancial estimates, as he pointed out: “It was a case by case evaluation. We couldn’t do it in relative terms, from the perspective of the whole ? rm. I mean we could not think that now we have achieved this much toward our overall target with this project. We did not have such a comprehensive frame. Decision-making was fact-based, but each case was evaluated separately. The absence of a ? nancial control system and decision uncertainty was not a reason for the CEO to abandon results-focused control. Once the accounting function had accomplished the ? nancial reporting structure, the CEO immediately sought to allocate ? nancial responsibility to the business units. The CFO issued budgeting instructions and the business units started to calculate their revenues and costs. The budgeted operating pro? t set performance targets for the businesses. The monthly performance evaluation, however, revealed that the budgets had been highly optimistic.
The CEO recalled subsequent events as follows: “. . . it soon became evident that revenues were much less than forecasted. Having seen the [actual] numbers, I started dismissals in the business unit [Internet and teleoperator]. ” For the CEO, the variance between ? nancial targets and actual results was the primary criteria for evaluating the businesses. Although the business unit managers explained unpleasant variances as merely a consequence of external surprises, the CEO regarded such explanations as secondary. Instead, he sought managerial accountability.
If he considered a variance signi? cant enough, he did not hesitate to take corrective action—to lay-off personnel. Even though the ? rm adjusted its operations in the Internet and teleoperator business, the actual numbers indicated that pro? tability was less than expected two months later. The business had not grown as budgeted—the environment was unpredictable because of increasing customer mobility due to New Communications Market Act. Consequently, a few dozen of employees were ? red—more than ever before in the ? rm’s history. The CEO also decided to facilitate control and ef? iency by physically centralizing all the domestic operations and assets into the same premises. He thereby sought to increase his capacity to control employees and other assets more directly and increase integrity within the ? rm. Along with the centralization efforts, the ? nancial results indicated that the mobile entertainment business, even though growing, was lagging behind its ? nancial performance targets. The CEO dismissed more employees in three waves that followed each other at intervals of about three months. Some international subsidiaries and pro? centers were closed as they were burning cash rather than generating revenues. In addition, the CEO terminated development programs that seemed to occupy too much management attention and time. He thus sought to focus management efforts on improving demand prospects for the current products and services. The CEO and CFO primarily applied the budget-actual-variance framework in exercising control over business unit performance. Although pro? tability was primarily sought by means of operational ef? ciency, the ? rm also invested in growth. The investment decisions were simply based on payback times, thus supporting the pro? ability criterion of securing liquidity, as the CEO described: “Besides pro? tability, I emphasized fast liquidity. I did not accept a long delay between costs and revenues. This type of thinking drove our investment decisions. Probably we rejected pro? table investments because the payback period was too long. ” While pro? tability targets set the direction and liquidity set the boundaries for business development, the ? rm also monitored operative results carefully. The operative information systems produced valuable non-? nancial feedback in a timely manner to the operative units.
The CFO described the non-? nancial indicators as follows: 334 M. Sandelin / Management Accounting Research 19 (2008) 324–343 “We monitored spoken minutes in ? xed line and gsm calls, net of subscriptions as regarded almost all of our products, and order-delivery period in Internet access services. We calculated averages, absolute changes, and other ratios on a weekly basis in order to be proactive. In this sense the non-? nancials were extremely important. ” These non-? nancial indicators were not performance targets. Weekly analysis of the ratios aimed at learning about ? ancial results because the linkages between non-? nancial data and ? nancial results had remained blurry as earlier. Digitized technologies impeded accurate allocation of overheads and thus exact product margins were unknown. The non-? nancials provided a proxy for estimating revenues and tracing customer satisfaction. With the passage of time, the ? rm learned more about its cost-structure and the ratios provided important feedback for adjusting actions in a proactive way. 3. 4. 2. The control practices complementing the results controls Action controls were manifested in administrative structures and constraints.
The top management group consisted of the business unit managers, the CEO and CFO, and the heads of the support functions. Its role was “to maintain a shared information base and assess strategic options,” as the CEO explained. Operative decision-making authority was allocated instead to the business unit management groups. In the beginning, the CEO worked with business unit managers in striving to create action plans for the year at hand. He insisted that the managers draw up three alternative plans: best, basic and worst case scenarios. The chosen plan de? ed performance targets for each unit, although the scenarios also provided guidance for adjusting operations in the case of unexpected changes in the complex and uncertain environment. On the other hand, to establish the sense of accountability and hence to control actions in the business units, the CEO required formal minutes from management meetings at the business unit level. He commented on the rationale: “Without documentation these business items would have been forgotten. ” Many of the managers had experienced rapid growth in the ? rm’s history and were used to taking the initiative without clear responsibility.
They had less experience of the disciplined development and management of projects. Thus, the requirement for the formal documentation enhanced the sense of responsibility and accountability for the new ideas. A controller clari? ed the atmosphere within the business units as follows: “It was quite common to throw business ideas into air [in the business units]. Nobody really carried responsibility for these ideas. In this sense, documentation forced them to put someone in charge and also to draw up plans, at least roughly. ” The CFO continued about the commitment to the ? ancial plans: “The team that produced the forecast was also responsible for achieving the numbers in real life. We wanted them to understand this. The sense of accountability in the business units was not as strong as it should have been. ” As the operative managers were unfamiliar with results controls, namely the operative budgets, controllers were hired for the business units. Their primary task was to conduct training and thus to enhance participative budgeting and commitment to the performance targets. The CEO’s cost-cutting actions had made it clear to the business unit managers that the performance targets had to be achieved.
Consequently, they allocated ? nancial accountability within their units by setting more detailed monthly and quarterly ? nancial performance targets for the cost and pro? t centers. Moreover, the CEO called for speci? c actions in the business units, especially follow-up for working time, in order to decrease the emergence of loose ideas and to keep the focus on planned and budgeted actions. The CFO and controllers had namely observed that despite the ef? ciency claims and speci? c performance targets, many employees and managers readily engaged in ‘high-? ying’ discussions.
Despite the CEO’s requirement for working time monitoring, the results were unsatisfactory as new initiatives kept appearing constantly. The cultural tradition of unconstrained working time still affected behaviors. Stronger actions were taken as the CEO described: “I wanted the employees dedicated to operative tasks to be separated from those doing development tasks. As long as they were mixed, the operative staff was disturbed by the development staff. There was too much inef? ciency. ” Although the CEO did not totally deny growth initiatives, he prioritized pro? tability.
Particularly his perception of a reasonable path for growth was based on systematic planning and organic business development rather than on radical M. Sandelin / Management Accounting Research 19 (2008) 324–343 335 innovations. Hence, he decided to separate developmental and operational tasks structurally. The CEO’s requirements forced the business unit managers to adopt new modes of thinking and to focus on economic factors in decision-making. Their primary role was not to coach and inspire the search for innovation but to manage ? nancial performance. Financial results gained signi? ance and gradually they became routine in managerial working, particularly in planning, as a line manager described the plan to establish business in a new country: “We calculated initial investment costs including mandatory payments and fees, as well as marketing expenses. Then we kept the product price constant and calculated required sales volumes in order to ? gure out pro? tability. On that basis we evaluated the [business] potential of each country [prior to ? nal decision]. ” In line with the preparation of alternative scenarios, the business unit and line managers included sensitivity analyses in their action plans.
Financial planning induced action accountability, which was enhanced by communicating speci? c task descriptions and expectations to the employees. Moreover, given the importance of ? nancial results, the senior management decided to nourish action accountability by implementing organization-pervasive approval policies, as a controller described: “We hired a person to take care of the development of policies, project proposals, approval practices, and so on. [. . . ] We had to get approval from our superiors, a recommendation and signature for any type of purchase. ” The acquisition of this speci? resource highlights the senior management’s strong desire to control not only results, but also the actions underlying them. Personnel controls were relatively informal, although action and results controls were based on formal modes. The latter two were re? ected in task assignment and recruitment practices. There was no explicit compensation for exceeding the performance target. However, if the target was not achieved, the senior management concluded, in the CFO’s words that “competencies did not match task requirements. ” Indeed, the senior management was con? ent of making decisions on the basis of results controls because they had hired and appointed business controllers for the business units. Consequently, it was reasonable to assume that unsatisfactory results were due to managerial competencies rather than poor ? nancial understanding. Once there was a need for hiring employees, the management sought to match competencies with tasks requirements. Competences were systematically evaluated on the basis of a candidate’s track record. There was no time and other resources for detailed hands-on training; new employees were expected to have the competence for the task in question.
The CFO clari? ed personnel selection and placement: “We had a lot of process development work in the ? rm. Practically it meant that we had to hire many people with academic degrees. [. . . ] These people had better quali? cations to work in a bigger entity, organize and build up processes, and in a way, to structure activities. ” Academic education was expected to re? ect familiarity with process thinking and thus, capability to develop the ? rm infrastructure. Employees were placed strictly according to the competencies, as a senior manager described: “The functional placement of employees and tasks changed radically.
For example, many development-oriented people were laid off while staff were simultaneously hired for customer service. ” As the comment hints, the senior management emphasized customer service because they saw customer satisfaction as the only solid basis for organic growth. In the digital business, the costs of switching between service providers were minimal and thus the reputation of the ? rm was highly signi? cant. The senior management allocated more human resources to customer service in an attempt to improve customer satisfaction. Hence, they sought to ensure that tasks were accomplished in this critical function.
Moreover, the high employee turnover, as a result of the structural arrangements and strict accountability, resulted in more fact-based thinking within the ? rm, as a business controller reported: “Once we ? red old managers and hired new ones, we got fresh blood here. [. . . ] At the operative level people started to talk about facts and to look at reality instead of dreams and beliefs. ” Despite the severe ? nancial crisis, one third of the initial employees still worked for the ? rm. These employees had visions and beliefs about how to perform tasks and occasionally these traditions blurred the advancing accountability culture. 36 M. Sandelin / Management Accounting Research 19 (2008) 324–343 Cultural controls were not the primary mechanisms by which the senior management tried to affect and guide employee behavior. The ownership culture had lost its signi? cance due to employee turnover4 . Organizational culture was not, however, considered insigni? cant. The senior management had recognized that some employees were not willing or able to accept fact-based management. Consciously or not, these employees kept up old patterns of action, manifested in innovative, ad-hoc decisions and actions.
To overcome these somewhat dysfunctional but deeply rooted traditions, the senior management sought to shape organizational culture by sharing ? nancial information, as a controller described: “Once the numbers for a quarter were ready, we had a ? rm-wide meeting. The CEO represented the actual ? gures, ratios, and plans. We had a clear aim to share ? nancial information for the broader audience rather than just for the management. ” As the controller continued, the senior management succeeded in this to some extent: “In these meetings some people almost fell asleep.
However, I was often surprised about how interested the staff from business unit A was in the performance of business unit B. They had detailed questions and they wanted to know the underlying reasons. ” Moreover, while interest in ? nancial thinking increased throughout the organization, cultural change was even more evident at the management level. A business unit manager who had been involved in the ? rm operations from the beginning described the changes in thinking as follows: “We no longer talked about ‘? rst-mover-advantage’ but about ‘? t-advantage’.
The second step taker was able to further develop the product, launch it at a lower price and even to introduce a technologically more sophisticated product than the ? rst mover. ” The result oriented performance management model had made managers consider economic factors and ? nancial pros and cons more carefully prior to action. 3. 5. Analysis of the Case 2 In the beginning, the ? rm struggled with cash suf? ciency. The senior management sought, by making operations as ef? cient as possible, to secure liquidity and pro? tability, which were not, however, externally imposed imperatives.
The external board was interested in monitoring share price. Hence, the new CEO and CFO had the freedom to decide about how to develop the ? rm and organize the control function. The headcount decreased from 250 to 220 in two years. On the basis of digital technologies the ? rm sought to compete with lowest price in a complex and uncertain though gradually stabilizing environment. Management control was primarily built around ? nancial results. As the endeavor was to make the employees accountable for their actions by setting speci? c performance targets, the structural arrangements were a necessary precondition for the budgets.
Besides the budget-actual-variance framework, result orientation was facilitated by non? nancial indicators and payback time calculations. Here the payback calculations and action plans operated as pre-action reviews, and thus as action controls. It is not, however, obvious whether such controls should be understood as results or action controls, particularly if they include ? gures (cf. Merchant and Van der Stede, 2007). Other action controls focused on increasing accountability. Formal documentation and minutes shifted attention of business unit managers from innovating to carrying responsibility for ef? ciency.
By means of approval policies they sought to control ef? cient and orchestrated actions at the lower organizational levels. Although these control practices were not linked as directly as action plans with the results controls, they complemented the hierarchical structure of result accountability. On the other hand, working time monitoring and structural separation of development and operational staff within business units supported the attainment of ef? ciency in operations and creativity in development. However, these practices were loosely coupled with other control elements although they evidently facilitated ef? iency pursuits by focusing attention. 4 The company had earlier (in late 2000) issued new shares in order to fund acquisitions and thus the value of the shares held by the initial employees had been diluted. In addition, many employees had sold shares in the IPO. More recently, employee turnover had eroded the ownership culture. M. Sandelin / Management Accounting Research 19 (2008) 324–343 337 Personnel selection and placement methods were affected by results and action controls. Task performance was evaluated on the basis of results and sanctioned accordingly.
As this procedure increased results accountability, action planning pointed out the needs for speci? c competencies and thus, set the criteria for personnel selection and placement. Moreover, ? nancial training in the business units complemented the result orientation although it also enhanced cultural change. For the same purposes of establishing an ‘accounting culture’, the senior management arranged a company wide meeting around interim reports in order to increase employees’ interest in ? nancial results. The design of the control package was based on result orientation.
The predominance of results controls determined the design of the most signi? cant action and personnel controls. These controls complemented the result orientation by increasing accountability and also contributed independently to goal achievement by shaping organizational culture. In fact, only action controls were reciprocally linked with results controls, whereas the presence or absence of personnel controls had little impact on results or action controls. This package of controls seemed functional—the ? rm grew and improved pro? tability as manifested by the ? ancial results (see appendix). 4. Comparative analysis of the cases and discussion This study has examined operation of management control practices as a package by presenting two case descriptions from the same ? rm at different time periods. The empirical analysis was based on Merchant’s (1998), Merchant and Van der Stede (2007) object-of-control framework about the diverse forms of management controls. Hence the approach went beyond the controls with information content (Simons, 1995) and the formal MAS on which the prior studies have focused in a growth-? m context (Granlund and Taipaleenmaki, 2005; Davila and Foster, 2005; Sandino, 2007). This section builds on the comparative case analysis and discusses features explaining the functionality of the control packages, the extent to which they seem to be equi? nal, and functional demands as explanations for variety in management control packages. The comparative analysis revealed that contingent factors were quite similar between the cases. Although these factors varied, the degree of variation was modest. Hence, the contingent explanation for control package variety was not obvious.
Instead, the cases were interpreted as being similar in terms of contingencies and an alternative explanation was sought from the notion of equi? nality, which suggests that control systems may look fundamentally different but result in an equally good ? nal state even in the face of similar contingencies. If there is a high degree of con? ict in functional demands, two forms of equi? nality are possible – suboptimal and con? gurational5 – depending on the degree of latitude with respect to the structural, here the control package, arrangements (Gresov and Drazin, 1997).
In both cases, the senior management had somewhat unconstrained latitude to implement and organize control practices, as is typical for newly founded and family businesses. The venture capitalist(s) were more interested in share prices than in management methods. The management also faced con? icting functional demands—whether to emphasize new product development and innovation, or ef? ciency and development of ? rm infrastructure. Therefore, the cases seem to represent con? gurational equi? nality—different combinations of control elements with which the ? rm achieved both growth and pro? ability6 . None of the functional demands