Working title:
A framework for routines with microfoundations:
linking individual action to recurrent interaction patterns
MScBA specialization Organizational & Management Control
Student: F.B. Wolters
Address: Johan de Wittstraat 18, Groningen
Student number: 1542206
Tel.: 06-22113318
1 INDEX 1. INTRODUCTION... 2 1.1 Conceptual Model ... 3 1.2 Research questions ... 4 2. THEORY... 6 2.1 Routines... 6 2.2 Individual-level scripts ... 8 2.3 Influencing Factors... 11
2.4 Conceptual model...Error! Bookmark not defined. 3. METHODS... 15
3.1 Qualitative research... 15
3.2 Pilot interviews... 16
3.3 The case... 17
4. RESULTS AND ANALYSIS ... 19
4.1 Collective level factors influence on the script development on the individual level ... 19
4.2 Individual level factors influencing script development ... 23
4.3 Factors influencing the aggregation of individual action into the routine... 25
4.4 Framework for routines ... 26
CONCLUSION ... 27
DISCUSSION ... 28
REFERENCES... 29
2 1. INTRODUCTION
Nelson and Winter (1982) have proposed the concept of a routine as analytical lens for investigating organizational change, this idea has inspired much research (for an overview see Becker, 2004). The routine construct appears in this research both as a key independent variable in its own right, driving organizational performance, and as a building-block for higher level concepts such as organizational institutions (Burns & Scapens, 2000). Given the prominence of the routines construct in strategic management (Eisenhardt & Martin, 2000), organizational theory (Pentland, 1995; Levitt & March, 1988), technology management (Henderson & Cockburn, 1994), international business (Kilduff, 1992), evolutionary economics where it originated (Cohen et al., 1996; Nelson & Winter, 1982) and especially management accounting change (Becker, 2004; Feldman & Pentland, 2003; Perren & Grant, 2000; Hodgson & Knudsen, 2004), this research is focused at further developing the body of knowledge on the routines construct.
The routines construct is faced by difficulties when applied (Felin & Foss, 2009; Becker, 2005a). Becker (2005a) addressed difficulties with its multitude of definitions. However, a more important finding for framing this research is Felin and Foss (2009) elaboration on the missing microfoundation, individual action, in research on routines. The routine construct explanations remain at the collective level: phenomena or events at the collective level are explained in terms of other phenomena or events at the collective level. The explanations do not refer to what is happening at the underlying individual level. This is a significant deficiency of the explanations, Felin and Foss argue, for it is clear that individual agents and their actions play a crucial role. Group behavior is resultant of the actions of its individuals, how these actions of individuals aggregate into a routine is a missing link (Felin & Foss, 2009). The microfoundation of a routine is individual action, and it is exactly this foundation at the individual level of routines which is missing (Abell, Felin & Foss, 2008).
Figure 1 illustrates the problem of explaining collective level outcomes or events in terms of other phenomena or events at the collective level (Coleman, 1990). According to Coleman, arrow 4 explanation is not legitimate. The correct causal chain is an explanation through arrow 1, then arrow 2 and finally arrow 3.
3 This research is focused at the microfoundations of routines, as a collective level construct, in an intended change program. The research objective is to develop a framework which describes a mandated change in routines with account for its microfoundations. This is a first step in the conceptualization of the link between routines and the individual level of action. For developing the framework this research draws on theory as well as a case study of a company which is implementing a mandated change program, setting forth on the research agenda of Felin and Foss (2009).
The case company is a small sized production company called Cordial Adhesives BV located in Winschoten, the Netherlands. It produces liquid and powder adhesives for the international market. Cordial aims to formalize certain processes and to standardize work patterns. Besides that, it wants to make better use of its ERP-system which also enhances more structured and standard work patterns. This management accounting change is aimed for in a small, informal company where behavior is highly routinized. The company is headed by a director who manages seven heads of departments. In a meeting with the director and the heads of the departments, it is decided to follow through on a change program which resulted from an evaluation of the use of the ERP-system. The change program is supported by a consultant, who organizes meetings, individual training sessions and assists in problem solving. This support is mainly directed at the technical problems of the ERP-system. The influence of the consultant on the routinized behavior is expected to be fairly minimal, since the little time he spends at the company and his lack of operational influence and knowledge.
The unit of analysis of this study is the routinized behavior within an organization (Pentland & Feldman, 2005). This research will focus on behavior on an individual level, as well as the group level, resulting from intended change by the management of the organization. A routine is defined as recurrent interaction patterns (Becker, 2004). A routine involves multiple actors by definition, and is not to be confused with the performance of a task in a ‘routinized manner’ as in everyday speech. Further, the concept of routines has no limit of how many actors can engage in a routine. The group studied in this research is the middle management, consisting of seven heads of departments, responsible for making operational decisions.
1.1 Conceptual Model
4 aggregation of individual action, shaped by the developed scripts, is influenced by the factor mutual fit in task perceptions.
1.2 Research questions
The conceptual model draws on literature for its concepts and influencing factors. This model is tested empirically, therefore the research question is: what factors influence the mandated change of an individual’s action and its aggregation into the order confirmation routine at Cordial Adhesives BV? The sub questions which guide the case study are based on the assumed relations presented in the conceptual model.
- To what extent does the group acceptance of a new cue influence the script development on the individual level?
- To what extent do institutions influence the script development on the individual level?
- To what extent does work motivation influences the script development on the individual level? - Which contextual factors (Anderson, 1995) influence the work motivation of the individual during script development?
- To what extent does a mutual fit in task perceptions of the participants in the order confirmation routine influence the performance of the routine?
- To what extent does a mutual fit in mental models of the participants in the order confirmation routine influence the performance of the routine?
+
+
+
+
Figure 2: Conceptual model
2 Collective level Individual level Routine Ostensive Performative Sense- making Cues Sense- making Scripts 1 3 Group acceptance of new cue Matching institutions Work motivation Contextual factors (Anderson, 1995)
6 2. THEORY
The central theme in this chapter is what concepts are needed to construct a conceptual model for routines with its microfoundations? Every concept described is operationalized because it is applied in the case study. The routine construct is the first concept elaborated on; artifacts, the ostensive part and the performative part of a routine are distinguished (Feldman & Pentland, 2003, 2005). The second concept is the notion of a script. It is argued that scripted behavior is the best theory of individual action to link with routines. Further, literature is reviewed for influencing factors to complete the conceptual model.
2.1 Routines
This section provides a better understanding of the construct of routines, what functions it has within an organization and what comprises a routine. To recoup, the definition of routines in this research is recurrent interaction patterns (Becker, 2004).
Based on Becker (2004), routines allow organizations to do four things. First and foremost, routines enable coordination. The capability of routines to enable coordination builds on the basis of a balance between the interests of the participants in the routine. Second, routines provide some degree of stability of behavior. Stability here is a relative term; it includes the potential change that is endogenous to the routine due to the agency of its participants (Feldman & Pentland, 2003). The stability of behavior has the implication that expectations about the behavior of others can be formed. Third, when tasks are routinized, these tasks can often be executed in the realm of the subconscious, thereby economizing on limited cognitive resources. Fourth, routines bind knowledge, including tacit knowledge. Next is a more detailed description of the different parts of a routine.
As unit of analysis, the routine construct has been developed by Feldman and Pentland (2003). They opened up the black box of a routine and distinguished two different aspects: the performative and ostensive. They consider artifacts as the physical manifestations of the organizational routine, see figure 2.
Figure 3: Routine
Routine
Ostensive Performative
7 The ostensive part of routines refers to the abstract, narrative description. Participants use it to guide, account for and refer to specific performances of the routine (Pentland & Feldman, 2005). For example, a common version of the ostensive aspect of the hiring routine involves attracting, screening and choosing applicants. The understanding of the abstract pattern may not be the same from person to person, from event to event or over time. Multiple and divergent understandings are probably more the norm than the exception. For these reasons, the ostensive aspect should not be conceptualized as a single, unified entity.
The performative part of routines consists of actual performances by specific people, at specific times, in specific places. Performances are carried out against a background of rules and expectations, but the particular courses of action we choose are always, to some extent, novel. In this sense, performances are inherently improvisatory (Pentland & Feldman, 2005). As with musical improvisation, the degree of divergence form the score may vary considerably and it involves listening to what others are playing, improvisations in organizational routines involves attending to the actions taken by relevant others and the details of the situation.
Pentland and Feldman (2005) refer to artifacts as the physical manifestations of the organizational routine. The range of artifacts that enable and constrain organizational routines is practically endless. The most obvious examples are those that deliberately attempt to capture or prescribe the routine, such as formal rules or standard operating procedures. More subtle examples include the physical layout of office space or seating. For example, the fact that an office includes a reception area facilitates the routine ‘intake of visitors’, but it does not directly prescribe who should be seen first.
Operationalization of a Routine
To analyze routines, first, a task must be selected in order to analyze how it is accomplished by the organization. Following from the definition of a routine, it should be a task that is recurrent, repeatedly accomplished in ‘the same’ way (Becker, 2005b). For purposes of observation, it would also be helpful if the frequency would be high rather than low so that numerous observations can be collected in a reasonable time frame. Furthermore, the task should be closely linked to the organization’s objectives in order to avoid sampling outliers such as activities that take place ‘at work’ but are not primarily concerned with accomplishing the organization’s objectives (Becker, Lazaric, Nelson & Winter, 2005).
8 group or team), not the individual level. Therefore, a task needs to be selected that require task-related interaction (Becker et al., 2005).
A task at the case company that complies with the criteria is the task of special order confirmation. An order confirmation consists of a price, date of delivery and other conditions which is replied after a customer places an order. Cordial has special and standard orders, special orders consist of about 30 percent of total incoming orders (see appendix A). Special orders are unpredictable, it can be an order of a test adhesive or a rush order. In case of a special order, the middle management engages in communication to come up with a decision for the special order confirmation, this is the performance of the routine analyzed in this case study. This task coordinates the demands of the sales department with the production department. Further, it shapes the conditions under which the other departments can work. The heads of logistics, production, sales agents, sales office staff, R&D, purchasing and finance engage in this routine. These people are supervised by a managing director who is not involved in operational decision making. Therefore, the group is able to make decision on a daily basis which they do by engaging in the routine.
2.2 Individual-level scripts
This research is about the microfoundations of routines; the individual action which aggregates into a routine (Felin & Foss, 2009). Therefore, a theory is needed of individual action which can be linked to the construct of a routine. The notion of a script, developed by Gioia and Poole (1984) is useful here. Scripts have a great deal of relevance for understanding organizational behavior on the individual level because they specify behavior or event sequences that are appropriate for specific situations (Choo, 1989). Gioia and Poole (1984) define a script as ‘a schematic knowledge structure held in memory that specifies behaviors or event sequences that are appropriate for specific situations’. First, it is described that a script has a lot in common when it comes to its functions and characteristics with routines. Second, the cues of scripts and sensemaking are shortly described. This section ends with a description of the development of scripts and different types of scripts.
9 A cue is the instigator of scripted behavior, it is defined as simple familiar structures that cause individuals to evoke scripts (Abelson, 1976), see figure 2. Cues can for example be a complaint from a customer which evokes a script for organizing a problem solving meeting. These types of cues lead to behavior and possibly event sequences which are part of an existing script. When a cue is unique, or unknown to the individual, this does not necessarily lead to scripted behavior. When a cue does not evoke an existing script, it leads to sensemaking.
The term sensemaking was first introduced by Weick (1979), and is defined as ‘understanding the situation’ or ‘getting the picture’ (Tillmann & Goddard, 2008). One of the functions of scripts is that they facilitate sense making (Ashforth & Fried, 1988). There are two ‘modes’ of sensemaking which need to be addressed here, sensemaking facilitated by scripts and sensemaking to understand a new situation. Sensemaking facilitated by scripts occurs if the perceived features of a situation match the features of an existing script, the individual can invoke the latter to explain the former (‘This is like the time that…’).
When a cue does not trigger an existing script, sensemaking can be the instigator of change as sensemaking occurs when an individual has no specified behavior to deal with a new situation. Sensemaking to understand a new situation is evoked by a cue and occurs when individuals give meaning to the new experience. Opposite to scripted actions which are often performed ‘unconscious’, sensemaking is a conscious activity. Sensemaking can result in emotional discomfort when the individuals are not able to process the cue with a script (Gioia & Poole, 1984). Further, sensemaking is recognized when an individual is eager for information about the situation which needs to be understood (Tillmann & Goddard, 2008). This type of sensemaking is referred to in the remainder of this article.
When a cue leads to scripted behavior, there are various forms of scripts. These are described with a continuum of script development by Gioia and Poole (1984). On the one hand there are novel situations and at the other stereotypical situations. Choo (1989) argues that novel situations require intensive conscious processing to decide appropriate events and behaviors. Such action involves little or no script processing because no script for behavior exists (Gioia & Poole, 1984). This situation
Script
Cues
10 creates conscious efforts of employees that are directed at searching the appropriate script and behavior. In stereotypical situations little or no conscious processing is required. It can be characterized as automatic script processing. Between these two extremes are events and behaviors requiring progressively less active processing as situations become increasingly conventional, repetitive, and stereotypical, see figure 3.
For the conscious efforts of employees that are directed at searching the appropriate script and behavior the term sensemaking is used, relating to novel situations. As a situation becomes more familiar, a weak script is developed. Weak scripts apply to situations where employees know what is going to happen in general, but they cannot predict a specific order (Abelson, 1981). Examples of such situations are problem solving meetings in management accounting or solving a potential conflict of interests.
Strong scripts contain expectations not only for the occurrence of events, but also for the progressive sequence of the events (Abelson, 1981; Choo, 1989). They apply to stereotypical and ritualistic occasions when employees know what will happen as well as the order in which it will happen. Choo (1989) argues that strong scripts are distinctive because of the relevance of learned associations between prior and consequent events.
Operationalization of scripts
To investigate the existence, development and influence of scripts for understanding behavior, it is helpful to define some operational distinctions involving scripts and script processing. A cognitive script is a mental representation of behaviors and behavior sequences appropriate for given contexts and retained in memory. A behavioral script is the performance of the observable stream of behaviors retained in an activated cognitive script (Gioia & Poole, 1984).
11 Behavioral scripts can be distinguished further into ‘performative’ and ‘inferred’ scripts, as a function of perspective on the behaviors enacted. The actor’s perception of his or her own behaviors is referred to as the performative script. The observer’s perception of, and inferences about, these same behaviors is called an inferred script. The study of behavioral scripts provides a vehicle for making inferences about underlying cognitive structures and processes on the basis of observed behavior. Observational methods, employing interpretive analyses, offer routes to inferring organizational scripts (Gioia & Poole, 1984).
Because the case study deals with a mandated, intended change program it is expected that the individuals engaging in the routine are confronted with cues which lead to sensemaking. On an individual level, the scripts and development of scripts by cues and sensemaking necessary to perform the order confirmation task is the unit of analysis.
2.3 Influencing Factors
The processes highlighted in the preliminary model are influenced by several factors. Next, these factors are explained. As mentioned before, it is important to consider how and when a factors has influence on the process. Considering timing and causality is crucial in adopting factors in a process-model.
Collective level factors influencing individual script development
In order to construe a framework of routines with its microfoundations, individual action, one needs to address the influence of group level dynamics, such as routines, on individual action, because individual behavior and cognition is in part shaped by the group (Harris, 1994). Since this research deals with a mandated change program which demands new behavior, a specific focus on the influence of the group level on the development of new behavior is required.
12 Since the concept of a routine has become prominent in research on organizational change, much empirical work has tried to explain inertia in change efforts. One of the relationships researched is that between institutions and the success of a change effort. An institution is defined by Burns and Scapens (2000) as ‘the shared taken-for-granted assumptions which identify categories of human actors and their appropriate activities and relationships’. Further, they argue that the existing routines embody the prevailing institutional principles. So it seems that in order to bring about a change in routines, the institutions need to change as well. In accordance with this, Feldman (2003) argues that it is difficult to change organizational routines when the change is inconsistent with broader understandings about how the organization operates as these understandings are produced and reproduced by other performances in the organization. Burns, Ezzamel and Scapens (2003) have concluded that a change in routines is possible as long as employees can in part adhere to their existing institutions. As such, institutions as a dynamic of the group are a factor which is expected to influence the script development of individuals.
Individual level factors influencing individual script development
It is important to consider factors which influence the process of developing new scripts, because they can influence the way or mode in which individual action is aggregating into a routine. There is abundant literature on factors to consider in change efforts, this literature will be reviewed.
There is a lot of literature which adopted a factor approach towards management accounting change. This research initially resulted in a few factors which were associated with successful change programs, often making use of statistical analysis, for example covariance (Shields, 1995; Anderson 1995). As this stream of research progressed, the list of factors associated to change increased up to 22 factors, see table 1 (Anderson, 1995). Research emerged which argued that in practice, the list of factors to consider were not much help, and the field of management accounting change needed a process view (Burns & Scapens, 2000). The argument was that factors cannot deal with the timing in a change process, often have a problem with causality and lack a behavioral focus. Besides that, a review of organizational change literature in the 1990s indicated that although lists of factors were carefully considered, this was no guarantee for a successful organizational change, instead, inertia was often the case (Armenakis & Bedeian, 1999). This resulted in researchers looking for other ways to approach organizational change, with the routines literature being one of them.
13 Individual characteristics Organizational factors Technological factors Task characteristics External Environment Intrinsic reward in change
Centralization Complexity for users Uncertainty Heterogeneity of demands Education Specialization vs. Multi-disciplinary Compatibility with other systems Variety Competition
Job tenure Internal
Communications Relative improvement over existing system Worker autonomy Environmental uncertainty
Role involvement Extrinsic reward systems Relevance to managers’ decisions Worker responsibility External communications
Informal support Training investments
Table 1: Factors related to management accounting change (Anderson, 1995).
It is plausible that the factors of Anderson (1995) influence the individual through a mediating factor, work motivation (Vithessonthi & Schwaninger, 2008). Work motivation is expected to be critical in the adoption of mandated new behavior. When employees have high work motivation, they are likely to show better adaptive responses and support to change in the organization. The data of the case study can be conclusive on this matter.
Factors influencing the aggregation of individual action into a routine
As discussed in the introduction, Felin and Foss (2008) elaborate on the missing microfoundations of the routine construct in research. Fortunately, this attention for microfoundations has instigated more research on the link of the individual and routines. In 2009, a conference was held in Rotterdam with the title: “Towards the micro-level origins of organizational routines and capabilities”. The conference too inspired research which is promising, however, so recent that there are no publications yet. Therefore, a working-paper is reviewed for factors influencing the aggregation of individual action (Witt, 2010).
14 very different from institutions previously described, because the perception of how the organization work and ones one role in it is institutionalized (Burns & Scapens, 2000). The factor mutual fit in mental models is not adopted due to the results of a pilot to test the data collection, see 3.2.
15 3. METHODS
This chapter addresses the methods used in the case study. First, characteristics of qualitative research will be described after which the data collection method of interviewing and observation will be addressed. Further, the chapter elaborates on the interpretation strategy. Second, the result of two pilot interviews are discussed. Third, the case is introduced extensively in order to contribute to the
comprehensiveness of the next chapter, results and analysis. 3.1 Qualitative research
A case study is chosen to study phenomena in an explorative way. Explorative research aims at understanding certain phenomena in practice, therefore qualitative data gathering is preferable. A case study is an effective study to gather qualitative data on phenomena in a particular situation. This research is conducted at one particular change program, therefore the single case study research strategy is appropriate (Flick, 2006). Other aspects of this case that invoke this research strategy are the real-life context of the case and the fact that the boundaries between context and phenomena are not clearly evident (Yin, 2003). The change program studied covers a change agenda of two months, which is extended upon success. During this two months, two methods of data gathering in case study are preferred, interviewing and observation.
Interviews
In qualitative research, methods for collecting data preferably enable the researcher to anticipate on the situation of the research. Interviewing as a method for gathering verbal data suffices for this (Emans, 2002). The interviews will take place with all participants of the routine in the change program at different stages in the process. Furthermore, in a case study research ‘guided conversations’ are more appearing and are an essential source of case study information (Yin, 2003). Therefore, the method used for collecting the verbal data is an open ended interview, more specifically, semi-standardized focused interviews (Flick, 2006). That means that the interviewee is free to contribute whatever he feels is necessary. This flexibility is needed in order to identify factors, a part from factors identified in literature. Fully structured interviews are not flexible, and therefore do not suit this research.
For structure, a number of questions are formulated to guide the interview. These questions are based on a topic list, this list contains several subjects per influencing factor and accompanying questions. After an interview, the transcript is submitted to the interviewee. The interviewee can indicate whether he agrees with the description of their opinions and if necessary, can complete or qualify their
16
Observations
Verschuren (2003) stated that observation reveals behavior, but not motives for that behavior and that interviewing reveals motives, but no behavior. Therefore, triangulating observation and interviewing is preferable when performing case study research. Observations will be done frequently, and are
documented for further analysis. Observations will be on an individual basis but also during group meetings and problem solving meetings between participants of the routine. To increase the validity of the observations, they are both performed hidden as well as open. Hidden observations means
observing people without notice, the open observations is an observation were the subject is informed about the observation. It can happen that a question is added to an observation, this is done to discover the motives of observed behavior.
Once every two weeks, a scheduled meeting with all participants is observed as well as the ad-hoc meetings with multiple participants. Observations are done on a daily basis, these can be triggered by an event or can be performed on a scheduled basis.
Interpretation strategy
The interviews and observations of individual and group behavior will be turned into transcripts for further analysis. This data will be interpreted by categorizing and coding the information. Every sub question deals with one particular factor. As such, all factors are categorized with the relevant pieces of information coded in the category. Besides that, it is possible to create an emergent category dependent on the data.
With this strategy, every phrase from an interview or observed behavior relevant for understanding the influence of a factor is categorized. This interpretation strategy is consistent with the qualitative content analysis (Flick, 2006). After categorization, it is possible to see patterns and connections both within and between the categories. This interpretation strategy allows for further analysis on an individual or chronological basis. As such, it is expected that the collected data can be used to create new insights and knowledge. Before presented the results and analysis, the case is thoroughly introduced.
3.2 Pilot interviews
17 First, the pilot showed the limitation of the data collection method of interviews. The processes in the framework contain mostly behavior. Behavior, in the sense of action, can't be studied through
interviews, only the emotions, motives and thoughts that accompany that behavior can. Therefore, the interviews should go hand in hand with the observations. The results from the interviews were more valuable when referred to observed behavior. This has led to the planning of interviews after certain events.
Second, the pilot has indicated that the factor ‘mutual fit in mental models’ is not relevant in the aggregation of individual action. As explained before, these mental models have a lot in common with the factor institutions. Institutions are expected to influence the development of individual action, not its integration. Although institutions are similar, not finding evidence for the factor mutual fit in mental models in the pilot is the main reason for not adopting it in the conceptual model. 3.3 The case
Cordial Adhesives BV is a functionally organized organization. With the management heading the organization and seven heads of departments. In total, the organization has about 35 employees with a few temporary employees. The managing director has very loose control over its heads of
departments, not interfering with operational decision making. In general, rules and procedures play a minor role at the company, decision making and communication has an informal character.
The company was founded little over ten years ago, and experiences rapid growth over last years. With that growth, the administrative burden and complexity of doing business has increased. Cordial’s customers are industrial firms active in the paper and cardboard market. A part of the customers have very unpredictable production runs when it comes to timing and material. This leads to orders at Cordial for products which need adjustments or completely new development on a short notice. Further, the prices of these products fluctuate constantly and delivery dates are on a very short timeframe. Because of the interdependency between Cordial and such customers, the sales in house and sales agents departments have a close relationship with those customers.
18 The routine of order confirmation start with a telephone call, fax or email from the customer ordering a delivery and ends with the confirmation. Between these two activities, the heads of departments need to communicate and create a consensus on the specifics of the order. They do so by engaging in a routine, with lots of face-to-face communication and telephone calls. The consensus on the specifics of the order shapes the condition under which the departments perform their work. Therefore, the routine enables coordination by balancing the interests of the heads.
The confirmation of orders can be very time consuming, since a lot of communication is required. The date of order confirmation is often very close to the date of delivery, therefore a lot of activities have to be performed in a small timeframe. Furthermore, the longer it takes for the customer to receive an order confirmation the more nervous the customer gets if it can follow through with its planned production runs. These problems have led to a change program, which tries to use time more efficient by making better use of the ERP-system.
This change program is support by an ICT-specialist who has evaluated the use of the ERP-system. He proposed two changes which intended to speed up the process of order confirmation. The first change is automatic order confirmation. Upon entering the data of the order, the system sends a confirmation to the client. This saves time at the sales in house department for composing the confirmation. The second change is the planning of the head of production. The system can indicate available production capacity, as the labor and machine hours per batch are entered into the system, with a planning
19 4. RESULTS AND ANALYSIS
The order confirmation routine at Cordial has been researched for over a month, this chapter reports the findings of this research. After presenting the most relevant data per sub question, the influence of the factor in the framework is analyzed. As stated before, in the analysis it is carefully considered how and when a factor has influence on the relevant process. First, the results of collective level factors that influence the script development are presented and analyzed. Second, the influence of the individual level factor work motivation on the script development process are discussed and analyzed. Third, the results of the factors which influence the performance of the routine are presented after which they are analyzed. The chapter concludes with recapitulation of the main findings in terms of the conceptual model.
4.1 Collective level factors influence on the script development on the individual level
The collective level factors influence the script development on the individual level. The most striking results will be discussed here. From the interviews, several quotes are presented and the most
important observed behavior is described as well. This way, the presented results will show both motivation and action. First, the factor matching institutions is analyzed, after which the group acceptance of a new cue will be analyzed.
4.1.1. Matching institutions Quotes matching institutions
Head of sales in house "We always go for the deadline, since we don't want to lose the customer. If you know that in 95% of the cases you go for the deadline, why waste precious time talking about the deadline?"
Head of sales agents "We have a name of being flexible to our customers, we need to protect that. So if it is possible to deliver goods, we should always at least try to make a deal."
Head of production "The system can't tell us how to do our jobs. When it comes to planning, you always need human judgment. I've been here for years, and when you have a breakdown of a machine, you can't ask the system how to deal with it." Head of financial
administration
"If the sale price is little above the material costs, we make the sale. We always make the sale, the agent is just looking for extra weight in his quest for an on time delivery. But don't get me wrong, I know we need to make a lot of sales."
20 Drawing on the results of the observations, the origin of these institutions is that the company has grown as a trading company, doing only purchasing and sales. As such, when you can make a sale at a price above the cost of purchase you always sell. Therefore, choosing not to sell or only under a certain set of conditions is not regarded as an option. Furthermore, distrust in information systems is due to a negative experience with information systems of the head of production in a previous company and the lack of working with computers as ever since two years ago by the logistics and production heads.
The initial cue was a get together of all participants in the routine and the message was clear: the change program is necessary to better serve our customers by making better use of information systems. After a vivid discussion, the meeting moved on to the details of the change program. Below are some descriptions of behavior during that discussion.
Behavior concerning matching institutions
Head of sales in house Enthusiastic, supportive for goals of change. Questions concerning training with new system functionalities.
Head of sales agents Expressed relief, commented that finally action was taken for this problem. Wanted guarantee that customers would be better off.
Head of production Very skeptical, showed irritation concerning customers wishes. Raised his voice on several occasions. Made the impression to feel 'threatened'. Stated he would fully cooperate as long as the functioning of the company is not at risk. Head of finance Skeptical, emphasized communication must not suffer from more use of
information system.
Head of logistics Angry, wanted to know who did something wrong. Was not interested in information from consultant. Supported the program 'as long as everybody knows what they are doing'. Left the meeting early.
Head of purchasing Supportive, did not asked questions. Stated that he will cooperate but did not want to get 'in the middle'. Appeared very interested and said he hoped things will run smooth now.
Head of R&D Supportive, expressed that his department would benefit from an effort of the entire group.
These results show that the participants in the routine who have the institution of always make a sale react differently to the cue than the participants who do not have this institution and distrust
information systems.
21
Analysis
The results show that it is possible that a group has certain institutions and understandings about how the organization works which can differ among group members. The heads of department whose institutions conflicted with the initial cue showed emotional discomfort. This discomfort shows that they were in the mode of sensemaking. Therefore is it found that cues which do not match with prevailing institutions are strong in the sense that they lead to sensemaking.
Although cues which conflict with institutions are strong enough to lead to sensemaking, this sensemaking is not leading to an effort of developing new scripts. Sensemaking leads to the
understanding that the company will not benefit from proposed changes, and therefore inertia towards the change effort. Sensemaking of cues that conflict with existing routines which lead to script development is difficult. These cues generally lead to a strong lapse into existing routines. The heads of the other departments all reacted positive towards the change program. However, the head of sales in house was the only one really eager for information. This showed he was making sense of the new cues. The other heads were interested but did not show emotional discomfort or eagerness for more information. It was only later in the change process that they discovered the consequences for their own functioning although it was made clear that the changes would affect all. Therefore it is stated that a cue which matches institutions, or are easily linked to existing institutions, can be experienced as weak and as such do not lead to sensemaking. If the cue is strong enough and matches institutions, sensemaking probably leads to an effort of behavioral change (new scripts). 4.1.2. Group acceptance of a new cue
As described in paragraph 2.3, the acceptance of a new cue is an important group dynamic for the development of individual scripts. Several results are presented as quotes below.
Quotes about group acceptance of a new cue
Head of logistics "I need the transport order for the goods in order to do my work, so I request them as I've always done. Besides, it is a little effort for sales to provide me one."
Head of production "We had a little difficulty with the system, so I call sales every now and then if there is something up."
Head of sales in house "Some of the production guys would phone that they wanted to generate a production order. That's impossible without the belonging sales order, so I had to enter the data. I figured that if they needed a production order they would make the order in time, so I confirmed the order."
22 The results make clear that when a routine changes, the actions of participants are cued in different ways then before. In this particular change program some important cues are provided by the system instead of verbal. From the interviews, it can be drawn that the heads who did not accept the new cues were slower in changing their individual action than the heads who accepted the cue.
Observations showed that the heads of logistics and production tried to work around the new cues, this hampered their development of new scripts in a major way. If the ‘old cue’ is maintained or often referred to, it triggers the ‘old behavior’ of the routine which is unconsciously performed. In the first stages of the change effort, the changes were the hardest, because they needed to be performed fully conscious. This requires a lot more effort to perform a task and that’s why it is appealing to resort to past behavior.
Old cues are generally stronger then the new cues, but after two weeks working with the new cues, this balance changed. Old cues resulted in confusion of the individual towards which action is appropriate. This confusion was sometimes similar to a state of sensemaking.
The quote of the head of sales agents is remarkable, after not engaging in the new routine the first week, he realized that the group had to step it up. Observations showed that the head of sales agents deliberately passed around the vibe of ‘new’ when talking about a special order confirmation. At first, other participants in the routine reacted reserved but quickly picked up the jokes and passed it on.
Analysis
The results show that after a cue each individual reacts different towards that cue. Some are inert and ignore the cue, they put no effort in accepting the cue and react with existing scripts. The heads of production and logistics immediately demanded that old forms of communication were persisted. They claimed that the organization could not function without them. This is a good example of inert
behavior. They hampered the progress of script development of the heads that supported the change because the new behavior was repeated less often.
The individuals who accepted the new cue had to put in effort to develop new script in order for the new routine to work. When a group member referred to the new cues when specifics about the special was communicated, the level of effort was shared. This made the development of new script easier for the individuals. When referred to old cues, the basic reaction of these individuals was to cooperate accordingly because the old cue triggered unconscious behavior.
23 Willing individuals got confused and resorted to a state of sensemaking, showing discomfort with the multitude of expectations about their functioning within the organization.
The heads who supported the change effort experienced a big change when the head of sales agents started his campaign of ‘new special is coming up’. He really required an effort of the group because of the ‘newness’. It was a lot easier to perform the newly developed script in a conscious way for the individuals involved. Therefore, it is found that when engaging in the routine, new ways of working are addressed as being new and accepted by the group, early stages of script development are speeded up. This is because it leads to more consciousness which is crucial in developing new behavior that later on will become less conscious.
4.2 Individual level factors influencing script development
The only individual level factor studied and analyzed in this research is work motivation. Work motivation is an important determinant in the ease of learning new behavior. It is expected that being motivated helps to overcome discomfort in developing new scripts.
4.2.1. Work motivation and contextual factors influence on individual script development
During the interviews, most participants in the routine stated to be fairly motivated. Results indicate that only when referred to a setback or emotional discomforting moment that motivation showed changes. Therefore, in the presented quotes the context of experienced motivation is relevant.
Quotes related to work motivation Head of sales
in house
Conflict with head of production
"This is no way of changing the business. Now it is totally up to me to meet the demands of these customers. I'd rather go back to the old situation, this stress is not part of my job."
After first training session
"Perfect! This is what I needed. It saves more than an hour a day, further, I have to debate less than usual. I think it is a big step forward to finally commit to the system."
Head of sales agents
After claim of client
"Terrible, at times like these you think why are people so damn stubborn. A part from financial loss, this is very bad for the team spirit. We just need to communicate better and make it work, because I know we can do this, we’ll make it a success." Head of production After first training session
"I do what is expected. Everybody thinks this thing will deliver, but nothing changes really. I just have to type in a load of numbers but it's business as usual."
24 The people who started to develop new script right after sensemaking, also were highly motivated. The best example of this is the head of sales in house, he wanted to start the first training session directly after the kick-off meeting. But to the contrary, he also experienced very low motivation. This made him to display almost inert behaviour, however, he performed the new tasks without even one complaint or discomfort.
The results further show that the head of sales in house did experience a major drop in motivation, resulting from a conflict with the head of production. He stated that he did not mind aborting the new way of working. Although this motivation drop off, he performed the ‘new’ task as was expected. He repeated the new task often enough to become a more or less unconscious effort. The performance of the task therefore required less effort, so it did not need to be backed up by a high motivation. 4.2.2. Contextual factors of Anderson (1995)
It is important to note that this research did not aimed to research which of these factors (Anderson, 1995) influenced the process exactly. For this research, it is more important to show how and when this category of factors have their influence.
The result indicate that when making sense of the initial cue, the heads of production and logistics concluded fairly quickly that the change effort was not in the best interest of the organization. The heads of other departments showed on several occasions an eagerness for information, which is besides emotional discomfort the most present during sensemaking. This eagerness was fully directed at the contextual factors of Anderson (1995).
Analysis
Work motivation changed during cues, setbacks, conflicts and system failures. Besides these moments or incidents, where the individual experience a state of sesnemaking, motivation seemed to be fairly stable. Therefore, sensemaking is a situation in which the level of work motivation is determined. During sensemaking, the individual is making a conscious effort to understand the situation.
Sensemaking can lead to a high or low motivation, in order to understand this relation, the role of the contextual factors of Anderson (1995) need to be explained first.
25 As the scripts were more or less developed and performed less conscious, the influence of motivation and the role of the contextual factors began to level off. It was only during setbacks that several heads referred to contextual factors for example a lousy organization of the project but usually motivation did not suffer radically from it since it was already high.
4.3 Factors influencing the aggregation of individual action into the routine
The conceptual model has only included the factor mutual fit in task perceptions, however, during the research the factor power relations showed to influence the aggregation of individual action into the routine. Therefore, this factor is presented as well.
4.3.1. Mutual fit in task perceptions
The following quotes are related to the fit or misfit in task perceptions. These quotes are selected to indicate how the head of the department perceives its job in relation to that of an other head.
Quotes about task perception
Head of sales in house "My job is to take stock of the customers wishes and demands. If the sales agent involved considers those demands as correct, I communicate them to departments involved. The head of production only needs to asses if the demands are possible on a reasonable timescale. If so, business as usual, otherwise we need to figure out what way to approach the problem."
Head of sales agents "As sales agents we promote the business, we acquire customers who have their particular demands. We need to satisfy them, otherwise they'll be with our competitor any time soon. We make sure that doesn't happen, and therefore are constantly offering solutions for their problems. The R&D department needs to use its brains as well and provide us with the necessary know-how, the rest is simple." Head of production "At production we know best what's possible and reasonable for clients. Those
salesmen don't know the basics of making glue. We have our limits and I tell you were pretty close."
Head of finance "I support the decision making from a financial perspective, assist with coming up with the numbers and try to keep everyone in line. The sales agents act like wild dogs at the moment, trying to sell glue to everything that breaths. If you keep spoiling the customers, their demands will get more and more."
Head of logistics "The real decision making is here, close to the fire. We know the stocks, the transport planning, what production is up to and so on. A lot of phone calls end up here from the office if they need additional information. People don't realise what we make happen every week. Apart from me and the head of production, it's all paperwork and the 'affair of the week'."
Head of purchasing "My responsibility is to keep production in stock of raw material. Off course I need to keep the prices as low as possible, but most of them are based on the market. Sales and production haven't figured out what way to go here, so as a safety, I keep stocks on a pretty high level."
26 Other results from the interviews indicate that the heads of the sales departments go head to head with the heads of logistics and production. One side thinks that the production department is the driving force behind the company, the other side think that sales is leading in the decision making. In a functionally organized organization this is not so surprising, however, this ongoing debate results in very little knowledge about the complexities in tasks of other departments.
The result indicate that the other heads of department feel ‘stuck in the middle’. When they interfere with the territory of sales or production, they feel like choosing a side in the decision making process. The change program has led to a less active role of these heads in the order confirmation routine. The main reason for this is that they think that production and sales need to sort it out before they interfere with the order.
The observations have shown an inward focus of the heads of department after most meetings. The individual changes were so individually changed, that they were heavily influenced by ones ideal process of order confirmation. When shortly asked why they performed a task in that particular manner, the answer always reflected upon the self-interest of the head. It was exceptional that a head stated that another department will profit from the way he performed his task.
Analysis
The aggregation of changed individual action led to a new routine which performance was no real improvement over the old routine. Every head of department altered the new routine to fit his personal task perception. The image of how a special order should be confirmed, the ostensive part of the routine, was leading in the aggregation of individual action.
When the knowledge of the task of other participant in the routine is low, it means that the ostensive part of the routine is very different among participants in the routine. The production department and sales departments show that were the biggest difference is, the most problems occur.
The individual action was, especially in the first week, in accordance with the change program. The problems with aggregating these actions are cause by the fine grained contrasting ideas about what is most important in the order confirmation. A mutual fit in task perception can alleviate these
contrasting ideas.
4.4 Framework for routines
27 repetition for the participants was easier. Persisting in old scripts, demanding the old cues for the task and rejecting new cues hampers the development of individual scripts. Once a participant in the routine falls back on its old existing scripts, it becomes harder to engage in the script development again. Therefore, group acceptance has a positive effect on individual script development, especially in the first stages of development.
Matching institutions has a positive impact on the individual script development. This match is not between participants, but between the cues to change and the institutions of the individual. When they match, it likely that the cue will result in the development of new scripts because the institutions support the sense one can make of the cue. Unless the cue is not strong enough, a cue which matching the institutions of the individual will induce an effort the change. It is stated specifically that it is about the institutions of the individual, because it is found that within a group institutions can differ.
During sensemaking, the contextual factors of Anderson (1995) are important. The participants of the routine who can make sense of the cue, are eager for information. This eagerness is for a big part directed at the contextual factors. When pleased with the specifics of the factors, the individual has a higher work motivation when developing the new scripts. This is because the state of sensemaking is the only time when an individual is rational. The factors of Anderson (1995) are based on economical models with the assumption that a person is fully rational, so during sensemaking the contextual factors come in to play.
A mutual fit in task perceptions influences the aggregation of individual action and thereby the
performance of the routine. A misfit in task perception prevents coordinated individual action, because the action is aimed at contributing to the ostensive image of the ideal routine the individual has, instead of the best interest of the group. As such, the fit in task perception smoothens the aggregation of individual action and therefore results in better performance of the routine.
28 CONCLUSION DISCUSSION + + + + + Routine Individual level Collective level Cue Sense-making Script development Action Ostensive Performative
Mutual fit in task perceptions Matching institutions Group acceptence of new cue Contextual factors Work motivation
29 REFERENCES
Abell, P., Felin, T., & Foss, N.J. 2008. Building Micro-foundations for the Routines, Capabilities, and Performance Links, Managerial and Decision Economics, 29:489-502.
Abelson, R.1976. Script processing in attitude formation and decision making. Cognition and Social Behavior, Eds. Carroll, J. S. & Payne, J. W. Erlbaum, Hillsdale, NJ, 1976.
Abelson, R. 1981. Psychological status of the script concept. American Psychologist, 36: 715-729. Anderson, S.W., 1995. A framework for assessing cost management system changes: The case of activity based costing implementation at General Motors, 1986-1993. Journal of Management
Accounting Research, 7: 1-51.
Armenakis, A.A., & Bedeian, A.G. 1999. Organizational change: A review of theory and research in the 1990s. Journal of Management, 25(3):293-315.
Ashforth, B.E., & Fried, Y. 1988. The mindlessness of organizational behaviors. Human Relations, 41(4): 305-329.
Becker, M.C. 2004. Organizational routines: a review of the literature. Industiral & Corporate
Change, 13(4): 643-678.
Becker, M.C. 2005a. The concept of routines: some clarifications. Cambridge journal of economics, 29: 249-262.
Becker, M.C. 2005b. A framework for applying organizational routines in empirical research: linking antecedents, characteristics and performance outcomes of recurrent interaction patterns. Industrial
and corporate change, 14(5): 817-846.
Becker, M.C., Lazaric, N., Nelson, R.R., & Winter, S.G. 2005. Applying organizational routines in understanding organizational change. Industrial and corporate change, 14(5): 775-791.
Burns, J., & Scapens, R.W. 2000. Conceptualizing management accounting change: an institutional framework, Management Accounting Research, 11(1): 3-23.
Burns, J., Ezzamel & Scapens, R.W. 2003. The challenge of management accounting change: behavioural and cultural aspects of change management, Oxford, Elsevier, 51p.
Choo, F. 1989. Cognitive scripts in auditing and accounting behaviour. Accounting, Organization and
Society, 14: 481-494.
Cohen, M., Burkhart, R., Dosi, G., Egidi, M., Marengo, L., Warglien, M. 1996. Routines and other recurrent action patterns of organizations: Contemporary research issues. Industrial and Corporate
Change, 5: 653-698.
Coleman, J.S. 1990. Foundations of Social Theory. The Belknap Press of Harvard University Press: Cambridge, MA/London.
Eisenhardt, K., & Martin, J. 2000. Dynamic capabilities: What are they? Strategic Management
Journal, 21: 1105-1121.
Emans, B. 2002. Interviewen: theorie, techniek en training. Groningen: Stenfert Kroese
30 Feldman, M.S. 2003. A performative perspective on stability and change in organizational routines.
Industrial and Corporate Change, 12(4): 727-752.
Feldman, M.S., & Pentland, B. 2003. Reconceptualising organization routines as a source of flexibility and change. Administrative Science Quarterly, 48: 94-118.
Flick, U. 2006. An introduction to Qualitative Research (3rd ed.). London: Sage publications
Gioia, D.A., & Poole, P.P. 1984. Scripts in Organizational Behavior. Academy of Management
Review, 9(3): 449-460.
Gioia, D.A. 1992. Pinto Fires and Personal Ethics: A Script Analysis of Missed Opportunities.
Journal of Business Ethics, 11(5): 379-389.
Harris, S.G. 1994. Organizational culture and individual sensemaking: A schema-based perspective.
Organization Science, 5(3): 309-321.
Henderson, R., & Cockburn, I. 1994. Measuring competence? Exploring firm effects in pharmaceutical research. Strategic Management Journal, 15: 63-84.
Hodgson, M.G., & Knudsen, T. 2004. “The firm as an interactor: firms as vehicles for habits and routines”. Journal of Evolutionary Economics, 14(3): 281.
Kilduff, M. 1992. Performance and interaction routines in multinational corporations. Journal of
International Business Studies, 23: 133-145.
Levitt, B., & March, J. 1988. Organizational learning. Annual Review of Sociology, 14: 319-340. Nelson, R., & Winter, S. 1982. An evolutionary theory of economic change. Cambridge, MA: The Belknap Press.
Pentland, B.T., & Feldman, M.S. 2005. Organizational routines as a unit of analysis. Industrial and
corporate change, 14(5): 793-815.
Pentland, B., & Reuter, H. 1994. Organizational routines as grammars of action. Administrative
Schience Quarterly, 39: 484-510.
Perren, L., & Grant, P. 2000. The evolution of management accounting routines in small businesses: a social construction perspective. Management Accounting Research, 11: 391-411.
Shields, M.D. 1995. An empirical analsis of firms’implementation experiences with activity-based costing. Journal of Management Accounting Research, 7: 148-167.
Tillmann, K., & Goddard, A. 2008. Strategic management accounting and sense-making in a multinational company. Management Accounting Research, 19: 80-102.
Van der Steen, M.P. 2009. Inertia and management accounting change, The role of ambiguity and contradiction between formal rules and routines. Accounting, Auditing & Accountability Journal, 22(5): 736-761.
Verschuren, P. 2003. Case study as a research strategy: some ambiguities and opportunities.
International journal of social research methodology, 6 (2): 121-139
Vithessonthi, C., & Schwaninger, M. 2008. Job motivation and self-confidence for learning and development as predictors of support for change. Journal of Organisational Transformation and
31 Weick, K. E. 1979. The social psychology of organizing. 2nd ed. Reading: Wesley Publishing.
Winter, S. 1986. The research program of the behavioral theory of the firm: Orthodox critique and evolutionary perspective. Handbook of behavioral microeconomics. Greenwich: JAI Press.
32 APPENDIX A
Result of a sample of 133 orders. These orders are analyzed on the percentage of special orders coming in at Cordial Adhesives BV. The results are also calculated in terms of volume.
Number of orders
Orders Special orders Percentage
Holland 53 17 32%
Germany 41 12 29%
Rest of Europe 39 8 21%
Total 133 37 28%
Volume
Kilo's Kilo's special Percentage
Holland 878.601 253.599 29%
Germany 643.069 191.070 30%
Rest of Europe 545.538 63.388 12%
