DISCUSSION 6.1

Organizational design approaches in management consulting

Klaasjan Visscher and J. Irene. A. Visscher-Voerman University of Twente, Enschede, The Netherlands

Abstract

Purpose – The purpose of this paper is to map the variety in organizational design approaches, to clarify their differences, and to find out what constitutes good designing in practice.

Design/methodology/approach – A series of in-depth interviews with experienced, high-reputation consultants is used to reconstruct organizational design practice.

Findings – The paper presents a typology with three organizational design approaches, stemming from different theoretical traditions. The paper demonstrates that the three approaches comprise both traditional design activities – analysis, design, implementation, and evaluation – experimental activities, and political activities, but that the emphasis, elaboration, rationale and order of these activities are very different for each approach.

Research limitations/implications – The study provides a framework for further research on the contextualization of design processes, investigating which design approaches work under which conditions.

Practical implications – Practitioners can use the results of this study to clarify, improve and enrich their own approach of organizational designing.

Originality/value – Management literature contains many models of organization design processes, but empirical studies of these processes are rare, and not yet existing in the context of management consulting. The paper fills some of the gaps.

Keywords Organizational design, Problem solving, Organizational politics

Paper type Research paper

Introduction Organizational designing – the shaping of organizational forms – is an important issue in management thought and practice. Designing and redesigning organizations is a core activity of managers and management consultants and has been a focal point of management research in the past (Simon, 1945; Khandwalla, 1977; Mintzberg, 1979; Van den Ven and Joyce, 1981) as it is in the present (Romme, 2003; Van Aken, 2004; Boland and Collopy, 2004; Dunbar and Starbuck, 2006). In management literature, most attention has been given to the content of the design (Haberstroh, 1965; Mintzberg, 1979; Harris and Raviv, 2002), but in recent literature, the process of designing is also gaining attention (Weick, 1993; Yoo et al., 2006). It is recognized that the quality of a design and its success after implementation is dependent on the quality of the design process. Designing is not just the proper assemblage of blueprint designs in specific situations, but a creative process, in which the form and functions of a design are being wrought (Schön, 1983; Akin, 1994).

In the literature about designing, in organization studies as well as in other disciplines, the design process is mostly presented in the form of a phase-model. These models contain the subsequent steps a designer has to carry out to create a good design (Simon, 1969; Pahl and Beitz, 1996; Cross, 2000; Van Aken, 2004). Models may be more or less elaborate, but they have in common that they contain four basic steps: the

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0025-1747.htm

Organizational design

approaches

713

Management Decision Vol. 48 No. 5, 2010

pp. 713-731 q Emerald Group Publishing Limited

0025-1747 DOI 10.1108/00251741011043894

analysis of the design problem, the design of a solution, the implementation of the solution, and the evaluation of the solution in the light of the original problem (Lang et al., 1978; Van Strien, 1997). In short, we will call these models ADIE-models (Analysis, Design, Implementation, Evaluation). In classical design literature, these ADIE-models were presented as the best way to design. In a good design process, all stages should be carried out, and in the prescribed order. This is called the phase theorem (Witte, 1972; Lipshitz and Bar-Ilan, 1996). In more recent literature, some flexibility is allowed in the elaboration and the order of the steps (Roozenburg and Eekels, 1995; Van Aken et al., 2007). However, in-depth studies have shown that in practice, design processes are much more complex than even sophisticated ADIE-models assume. Some authors even turned skeptical. “To anyone interested in process, these diagrams shed very little light on how design acts are actually carried out” (Bucciarelli, 1994, p. 112). The bottom line of these studies is that designing is situated action (Suchman, 1987), which means that the course of a design process is influenced by the characteristics of the specific context and by the actual events during the process. Thus, it can be expected that variety exists in the ways in which practitioners design organizations, and that the activities of analysis, design, implementation and evaluation are carried out and sequenced differently in different contexts.

In organization design literature, hardly any attention has been paid to this variety in design processes. When variety is being thematized, it is mostly done on a general level, by addressing the limits of a certain design approach (e.g. Burnes, 1996; Van Aken et al., 2007). More thorough insight into the limits and variety of organization design approaches is important, though, to find a middle road between naı̈ve adherence to the phase theorem and total methodical skepticism. The purpose of this paper is to map the variety in design approaches, to clarify their differences, and to find out what constitutes good organizational designing. The ADIE activities will be used as a guideline, but the value of the phase theorem will be bracketed.

For the purpose of this study, we take a configurational approach, which attempts to bring order in variety by developing so-called ideal types, configurations that are internally consistent and that fit with multiple contingencies (Doty et al., 1993; Meyer et al., 1993; Short et al., 2008; Grandori and Furnari, 2008). We will develop a typology of design approaches – a rational, a dialogical and a pragmatic approach – which are consistent in the sense that the different activities that constitute the approach are aligned with each other and which fit within their context.

The typology has its basis in literature. For the elaboration we conducted empirical research. In particular, we focused on one group of practitioners: management consultants. Consultants play a prominent role in organizational design, both as providers of new organizational forms, and as change agents in the implementation of these in organizations (Benders et al., 1998; Clark and Fincham, 2002; Faust, 2002; Heusinkveld and Benders, 2005; Heusinkveld and Reijers, 2009). Consultants work on organization designs in different organizations and thus experience the success and failure of their ways of working in different settings, which gives them an opportunity to develop mature approaches for diverse situations (Greiner and Metzger, 1983). Therefore, experienced consultants form a rich source of practical knowledge about organizational designing.

MD 48,5

714

The structure of this paper is as follows. We will first develop the typology. Second, we will present the research design of our empirical study. Then the different typical design approaches will be elaborated on the basis of the empirical study. The paper concludes with a discussion of the results and suggestions for further research.

A typology of design approaches In the literature about organizational designing, three different perspectives on the process of designing have been developed: a rational, a dialogical, and a pragmatic perspective. These different approaches encompass different views of what a design is and what good designing is, and have different ideas about the importance and elaboration of different design activities.

Within the rational approach, organizational designing is seen as the construction of blueprints for the structural characteristics of organizations. Traditionally, these structural characteristics concerned the formal structure of the organization, i.e. the division of labor into functions, the allocation of tasks, responsibilities, and authority of these functions, and the creation of hierarchical and lateral mechanisms to coordinate and integrate them (Mintzberg, 1979; Harris and Raviv, 2002). In current textbooks, this is still considered the core of organizational designing, but other “tangibles” such as strategic plans (Ansoff, 1980) and business processes (Hammer and Champy, 1993) are also included. The design process is seen as a rational problem solving process (Simon, 1969), of which analysis, design, implementation and evaluation are the basic stages (Lang et al., 1978; Lipshitz and Bar-Ilan, 1996; Van Strien, 1997; Van Aken, 2004). Rational design models can be elaborate and contain one or more cycles, but the ADIE stages, in their original order, form the basis. In this, the first stage – analysis – is considered crucial, and it should receive much time and attention (Greiner and Metzger, 1983; Cross, 2000). In the analysis, the problem is identified, and the goals and objectives are determined. In complex situations, where the problem is not clear right away, decomposition is prescribed (Simon, 1969), i.e. the division of a complex problem into sub-problems, until the level at which the problems are manageable. In the second stage – design – solutions are created to meet the pre-established objectives. Creativity can play a role in order to find these solutions, but selecting the best solution is a rational process (Simon, 1969; Zeleny, 1982). The best solution is the one that scores the highest on a set of criteria that are derived from the problem statement, and possibly from a variety of financial, structural, or other constraints. The third stage – implementation – should be separated from design, logically and in time. During implementation, there may be compromises on aspects of the design, but the better the design, the better will be the end-result, after implementation. When designing, practitioners should not bother too much about potential implementation problems, because that would thwart the design process and could lead to sub-optimal designs (Williamson, 1975). The final stage – evaluation – concerns the extent to which the solution meets the pre-established objectives. Evaluation may take place at the end of the design process, which is called summative evaluation, but also during the process, to test and improve the design, which is called formative evaluation (Scriven, 1967; Verschuren and Hartog, 2005).

Literature on the dialogical design approach is analytical rather than prescriptive in nature. The approach corresponds with the rational approach in its view of the object of designing, namely the structural characteristics of the organization. It differs in its

Organizational design

approaches

715

view of the organization and the design process. Organizations are seen as political systems, and structures are supposed to serve the interests of the people who have the power over the resources on which the organization is most dependent (Hickson et al., 1971; Pfeffer, 1978, 1981; Mintzberg, 1983). Designing is primarily a process of building consensus or compromises about an organizational form within a certain dominant coalition. In such a process, it may happen that not the entire problem and solution space is explored, as the spaces that are incompatible with the interests of the dominant coalition are shut off. The choice of the best design is also guided by interests (Pettigrew, 1973; Buchanan and Boddy, 1992; Butcher and Clarke, 2003). This may lead to designs that are not optimal or even inconsistent from a rational perspective, but as Pfeffer observes, “these inconsistencies may be a reasonable response to the conflicting demands and interests of organizational participants” (Pfeffer, 1978, p. 240). Consent and commitment of key figures are valued higher than the norms of rationality (Freeland, 1996).

While the rational approach regards complexity as something that should be reduced from the start, and while the dialogical approach perceives complexity mainly as the product of political differences (Greenwood and Hinings, 1996), the pragmatic approach tries to absorb complexity. Assumption of this approach is that design situations have a complex or “wicked” nature (Rittel, 1972; Rittel and Webber, 1973), which means that they are unique, ambiguous, and difficult to define unequivocally. They also embrace more than only the structural characteristics of the organization. According to Schön (1983, 1987), designers cope with these situations in a process of “reflection-in-action”. Based on a hypothesis about the nature of the problem, they put a frame – i.e. an organizing model – on a design situation. On the basis of this frame they proceed and listen to the “back talk” of the situation, in order to explore it, to find out the intended and unintended consequences of their moves, and to confirm or refute the adequacy of their frame. On-the-spot experiments, in which practitioners try to make their hypotheses come true, play an important role. They violate the canons of controlled experiment, which call for objectivity and distance, but according to Schön, this does not make the experiments in practice self-fulfilling. The relation of a practitioner to the situation is transactional (Dewey and Bentley, 1949): A designer shapes a situation in a conversational way, in which the situation may “say” “I will not be shaped like that”. Designing and analyzing are intertwined processes. Designing is a process of experimentation, action-learning and improvisation, a reflective game with the problem-situation. “[I]t is a game with the situation in which he seeks to make the situation conform to his hypothesis but remains open to the possibility that it will not,” (Schön, 1987, p. 73). Bricolage, i.e. the situational tinkering with the resources at hand (Lévi-Strauss, 1966; Weick, 1993; Baker and Nelson, 2005), plays an important role in this approach. While improvising with models and ideas, designers closely watch the emerging designs and their functionalities, shaping them step-by-step on paper and in practice. Design and implementation are closely related in local experiments. Change is the aim of the design process, not a blueprint, although one may be constructed in hindsight, reflecting the design that has been created (Weick, 1993).

The rational, dialogical and pragmatic approaches of the design process have different views of the object and the process of designing, and emphasize different aspects of the design process as crucial. In our study of design practice, we will elaborate these approaches and their focal aspects. We will investigate how

MD 48,5

716

organizational designers implement ADIE-activities in practice, how they cope with the need for commitment and consent of the dominant people in the organization, and how they handle wicked problems and experiment in organizations.

Method To reconstruct how management consultants design, a research instrument has been created in which a series of in-depth interviews with experienced consultants has a central place. To explore what management consultants do, how they do it, and for which reasons, in-depth interviews are appropriate means (Kvale, 1996), but a potential drawback is that they might not reconstruct what designers really do and think. They rely on retrospective accounts of the involved designers, which may be colored and biased. Particularly when the interviewees give rationalized reconstructions of their experiences – which is an actual risk with management consultants, who are used to selling their ways of working to clients with rational methods (Werr et al., 1997) – this would be a fatal flaw of the design. The created interview instrument tries to minimize this risk.

Interview instrument The interviews started from the interviewees’ concrete actions and proceeded to the underlying reasons by inviting them to motivate, explain, judge, or justify these actions. Were the interviews to focus directly on their general approach, then there would be a risk that interviewees report “textbook-methods” or methods from their flyers, which may be different from what they actually do in practice (Carspecken and Cordeiro, 1995; Kvale, 1996). The interviews were centered round organizational design projects, which were selected on the basis of three criteria. First, the project should be recent, to ensure that the interviewee could remember the details. Second, the project should be typical for the work of the consultant, to guarantee that the interviewee had experience with this kind of projects. And third, the interviewee should consider his course of action in this project as good and productive. This does not necessarily mean that the project was a success, which is hard to measure and not unambiguously attributable to the work of the consultant, but it does mean that the interviewee, given his professional standards, regarded his way of working to be good practice. The interviewees were asked to tell what they did in these projects, starting with fairly open questions, and they were gradually led along the focal issues of this study. To complement the interview data and to check the factual data about the cases, project related documents, such as plans of approach, power-point presentations, and consulting reports were studied.

Sample For the selection of interviewees, a purposive sampling strategy ( Johnson, 1990) was followed. Two main considerations played a role. First, we concentrated on experienced consultants who had obtained a good reputation in the field (Glückler and Armbrüster, 2003), for we assumed that these consultants had developed mature approaches for designing organizations. The second consideration for selecting interviewees was to choose consultants from different working areas and consulting firms, in order to cover the heterogeneity in the domain. To identify consultants with a good reputation, a survey was done among the members of the Dutch association of management consultants (the Ooa) and the council of management consultancies (the ROA), the two management-consulting institutes in The Netherlands. Our assumption

Organizational design

approaches

717

was that in management consulting, although it is not a full profession (Visscher, 2006), fellow practitioners are able to judge each other’s competence to some extent. In this survey, respondents were asked to nominate consultants they considered to be very good, in their own field of expertise and in general. Interviewees were selected predominantly on the number of nominations they received. Besides, a spread in working area and consulting firm played a role. As a result, a group of 27 consultants was selected, of which 24 consultants were actually willing to cooperate, and 21 were willing to discuss concrete projects extensively. Table I gives an overview of these consultants and their projects. This group contained self-employed consultants, consultants from big and middle-sized national firms, as well as consultants from big international firms. They had different educational backgrounds and different sectors and branches in which they worked. Some of them were all-round consultants, while others had specialized in HRM, strategy consulting, quality management, or the administrative organization.

Data analysis The interviews were taped and transcribed. As a first step in the analysis, reports were created containing an extensive summary of the discussed project and approach, as

Consultant Consulting firm Project

B National Structure for two merging large national newspaper publishers C National Strategy design for a university E Self-employed Cultural and structural change for a middle-sized engineering

firm F Big national Structure and processes for a large insurance company G Big international Corporate structure for the national branch of a large

international consulting firm I National Structure for a merged middle-sized educational support

institute J Big national Processes for an R&D department of a large multinational

company K Self-employed Structure and processes for a organization for mentally

handicapped L Big national Business plan for a new service for a large trade union M National Structure and processes for the national branch of a worldwide

charity N National Partner-system for a large architectural firm O Big international Strategy for a middle-sized textile company P Big international Corporate structure for a multinational in the food sector Q Self-employed Structure for a national railroad maintenance organization R National Strategy, structure, and staffing for a local home-care

organization S Big national Structure for a middle-sized consulting firm T Self-employed Top-structure for a regional educational institute U Big international Administrative processes for a governmental department V National Management team structure for a governmental department W National Cultural and structural change for a middle-sized, family owned

wholesaler Y Big national Top-structure for an institute for higher education

Table I. Interviewed consultants and studied projects

MD 48,5

718

well as memos with analytical comments (Miles and Huberman, 1994). For further analysis, the report was coded using the ADIE model. Analysis, design, implementation and evaluation activities were identified, issues of consent, commitment and experimentation were sorted out, relevant arguments and contingencies in relation to these activities were articulated. In the next step, we used the three typical design approaches to code the data anew, classifying each consultant on the basis of what he did and his view on the design and the design process. After that, we elaborated the design activities for each type and clustered the underlying argumentation and mentioned conditions, in order to explore relevant contingencies. In the next section, the results will be presented.

Rational design approach Six of the interviewed consultants ( J, N, P, S, U and V) followed a rational approach.

Analysis In consultant U’s and V’s projects, the definition of the problem and the establishment of objectives, which is a key activity in the rational approach, was a straightforward process. In these cases, the clients had created a stratification of designs, in which higher-level designs imposed objectives on lower level designs. In particular, the design of an organizational structure imposed objectives on the design of a management team structure and a process design. In these situations, the establishment of objectives was fairly simple.

In the other three projects, the consultants put much effort in finding a core problem – the “real problem” in their terms – and in establishing objectives. They made an inventory of problems, mostly through interviews with the client and other people in the organization, to figure out what was going on. Finding the real problem was considered to be a difficult, cyclical process. According to consultant S, people in organizations mostly start by telling “lies and fabrications”. Consultant N said:

In almost every question a consultant gets, you know from previous experiences that there are a great many underlying questions that should be answered first, before you can really address the question posed. And you very often see that quite different, unstated, subjects play an important role; they somehow agreed internally to put subject A on the agenda with the consultant, when, in fact, it is all about subject C, but no one dares to talk about C at that moment.

In the analysis of the problem situation, models played an important role. Most consultants used a standardized, elaborated model to decompose and analyze the problem situation. Consultant S, for example, used a variant of the EFQM-model, and consultant P a variant of the 7S-model. Although they did not use the same model for every problem, the number of models in use was limited. Rather than assembling a unique model for each situation, they used generic, “verified” models, which would only be tailored to the situation in details. By using these models, the consultants said that they were able to safeguard the rationality of the analysis.

For consultants with a rational approach, this first step was considered crucial, because, once the problem and the objectives had been established, the design process could be shaped as a rational problem-solving process. The process was phased and decisions were made at the end of each phase, thus controlling the design and realizing

Organizational design

approaches

719

the beforehand-specified end results. Having a controlled process was important for these consultants. Consultant U exclaimed:

It is just an industrial process, it is so orderly. So it is predictable and controllable; it’s all fantastic!

What was also considered important is to keep focus on the content of the design and to ban political processes as much as possible. When political processes become dominant, this has – according to consultant N – two main drawbacks:

First, you may just make a stupid decision, with the wrong results. Second, it may lead to a certain degeneration of the organization, that people put up with the fact that these processes are purely political, and I think that, in the long-term, that is not good for the strength of the organization.

Design Consultants with a rational approach did not ignore political processes, but they tried to bracket them. In the first phase of their projects, the interviewed consultants put much effort in obtaining commitment for the design process, in particular from the top-management and other key-figures in the organization, in order to create a “protective space”, in which they had the freedom to focus on the design without interruptions of people who try to pursue their particular interests. Within these spaces, the consultants tried to make designs as good as possible in the light of the objectives that were articulated at the beginning of the process. This is a creative process, laboratory work according to consultant J – “here a jar, there a jar, mixing, stirring, and see what you get” – in which several models and examples of earlier designs are integrated to create the best solution possible. Consultants P and J emphasized that it is important to create an ideal design, in line with the formulated objectives. They acknowledged that the design that will actually be realized is not this ideal design, but a compromise that is also based on the traditions of the organization and the capacities of the people involved. Nevertheless: the closer you can get to this ideal picture the better. And besides, an ideal picture shows the clients the consequences of their objectives, which, according to the consultants, they often do not see.

Decisions about the design were made rationally, for instance through multi-criteria analyses or other rational-choice techniques, and step by step, in particular after the analysis and the design stage. When the locus of power was situated within the group of people who made the design, they made the decisions. When it was located outside that group – and this is mostly the case in management consulting – the decisions had to be confirmed separately. Consultant U, for example, made his designs of administrative processes in a team with people from the organization, but they were always sent to top-management for approval. These never made any amendments to the design, but their approval was necessary to give it force of law. Implementation of a design was considered impossible without approval of the management.

Evaluation The evaluation of the design after the project was considered important, but very problematic. A design should be evaluated by comparing the new situation to previously formulated objectives, but the effects of a design only become clear after

MD 48,5

720

implementation, often even later, and in this period many things may happen that have effect on the success of the design. Consultant V phrased it like this:

It remains difficult to measure the effects of your actions. You should try to remove what was not working well at the beginning, but that does not imply that what has actually changed corresponds exactly with what had to be removed [. . .] Mostly you cannot measure it in terms of “this was the case, this is what we did, and this is the result.

Because of these problems, several consultants turned to measuring customer-satisfaction, a certain period after completing the project. However, customer satisfaction may also change in this period, and clients may not show their satisfaction or dissatisfaction openly to a consultant. Several consultants stressed that only when clients hire a consultant again, or recommend them to others, they were undoubtedly satisfied. In this way, consultants used their own success as a consultant as a measure for the success of their designs.

Dialogical design approach Six of the interviewed consultants (B, C, I, L, T, and Y) followed a dialogical approach. The organizations for which the consultants worked had in common that they were professional bureaucracies (Mintzberg, 1979). The consultants regarded the design situations as socio-politically complex. This means that, because the organizations lacked a powerful center, the consultants had to deal with many stakeholders, with differences in opinion, perspective, and interest, and with the presence of potential conflicts and lack of trust among them, in particular in recently merged or merging organizations. In terms of consultant T, a situation is socio-politically complex when:

[. . .] the situation is laden with conflict, the client is part of the problematique, and the suggestions you make about the content land in a complex field of opinions, interests, and perspectives. When such is the case, I will definitely not presume to know what is the best approach. I cannot do that, nor do I want to.

In socio-political complex situations, consultants cannot ignore or bracket political processes. In such a situation, a rational approach would be continuously disturbed by stakeholders invading the design space to promote their interests, and would lead to designs that are not implementable because of a lack of commitment and motivation of the key figures in the organization. Therefore, the consultants chose for a dialogical approach, in which a design is created by discussion and negotiation, in order to achieve consensus or a compromise to which people with different interests and preferences can commit themselves.

Analysis In a dialogical approach, the processes of analysis, design, implementation, and evaluation also occur, but with a different focus. The aim is not to create a design that best meets pre-established objectives, but to create an implementable design to which the key figures in the organization commit themselves. The difference shows in the analysis, in which gaining commitment and the revealing of different interests and perspectives on the problem is seen as more important than finding a core problem. Besides, shortcuts are made. Consultant I, for instance, only did a “quick and dirty” diagnosis and decided to skip further analysis. He argued that in a merger, speed is important in order to keep the parties committed and to make the process irreversible.

Organizational design

approaches

721

Gaps in the analysis could be filled later, he argued, but a loss of commitment and motivation would have ruined the process.

For the analysis of the situation, consultants made use of models. In function, these models differed from the models used in the rational approach. Their main function was to structure and improve the discussions among the key figures and to enhance their trust in the design process. Their form was situation-specific rather than generic. Dependent on the problem situation and the perspectives among the people in the organization, the consultants used several models to enlighten the analysis with the key figures. Consultant C said:

The nice thing about consulting work is that while you do have models at your disposal, you develop them interactively, together with the client [. . .]. It is an interactive complex, and often multidimensional. You can’t say “it is a strategic problem”, because it is also a negotiation problem, and also a cultural problem. So complexity makes it difficult to say “I have a model”. No, you switch between a number of models.

Design The design was not the work of a single designer or an autonomous design team, as in the rational approach, but of the key figures in the organization. Major steps in the design process were mostly made on working conferences, with several stakeholders participating. The interviewed consultants put much effort in leading those conferences, particularly in establishing a design space, during the opening rituals of the conference. Opening rituals were considered important, because if not all participants are imbued with the purpose and importance of what they are going to do, the conference may turn into a free-discussion space, where no decisions are made. Through these rituals, the participants of the conference committed themselves to the design process and positioned themselves as co-designers.

One of the important elements in this design approach is the choice of “key figures”. People were included because they occupied high positions in the hierarchy of the organization, because they possessed certain expertise, or because they were enthusiastic about the design process. The decision of who to include was regarded as complicated and delicate. Including too many people complicates the process and implies a loss of speed and efficiency. Excluding people has the risk that they would resist the outcome of the design process.

The difference between the rational and dialogical approach also shows in the creation of alternative solutions. In several projects, alternatives were constructed, but with a different emphasis than in the rational approach. There was no free exploration of the solution space, nor a quest for the best possible design. Alternatives were made in such a way that a productive discussion among key figures became possible. In one case the alternatives were made to reflect the conflicting current opinions of key figures. In another case, without conflict, they visualized the main decisions to be made, within the main direction the people had already chosen. Consultant I said:

I formulate those four [alternatives] in such a way that they offer different solutions for the direction they are heading for [. . .]. So that is the principle: take the path that they are already heading down and split it into different options, in such a way that those options reflect the dilemmas of their choices, so that a clear solution eventually emerges from it.

MD 48,5

722

In the rational design approach, techniques such as multiple criteria analysis played an important role in reducing the number of alternatives. The consultants with a dialogical approach mainly emphasized the drawbacks of these techniques. Consultant C remarked that such an analysis is rarely conclusive. At the moment of the reduction, there is often dissent about the weighing of different criteria, which may result in dissent about the best alternative. They relied mostly on facilitating the discussion and negotiation process between the key figures, aiming at consensus about one of the alternatives.

Evaluation In this approach, the quality of the design is evaluated on the basis of the satisfaction of the client and other key figures. The answer of consultant Y, asked to evaluate his project, is illustrative:

When a dean, who previously was in trouble and looked bad, says to you “I go to my work with joy again, thank you”, then that is fantastic.

For the rational approach, client satisfaction is also important, but the difference is that in the dialogical approach, this way of evaluating fits the general approach, while in the rational approach, it is chosen as second best for practical reasons.

Pragmatic design approach Nine interviewed consultants (E, F, G, K, M, O, Q, R, and W) had adopted a pragmatic design approach. They all worked on assignments they considered complex. These designs concerned the structural aspects of the organization, but also concerned less tangible aspects such as the organizational culture and local ways of working.

Analysis Consultants with a pragmatic approach also started with an analysis of the design situation. However, different than in the rational approach, they did not use generic diagnostic models, nor did they try to find “the real problem” or the actual objectives. Different than in the dialogical approach they did not focus on finding key figures and reconstructing their interests and perspectives. Their analysis was more an open exploration, in order to find out how the organization worked and where it malfunctioned. Consultants differed in their techniques of analysis. Consultant E, for instance, observed people at work, in different parts of the organization. Consultant M did interviews, to get meticulous case-descriptions of the occurring problems, which he used as an input for his diagnosis.

The consultants had in common that they explored broadly, in order to get a feeling for the entire problem situation, to make sense of it and to choose a productive frame for the further process. They also had a large and heterogeneous model repertoire, which they used to create frames that would work in their particular contexts. Consultant R, for instance, created a multilevel framework, with a strategic, structural and staffing level, which allowed him great flexibility. He started designing at the strategic level, but when he got stuck, he switched to the structural and staffing level in order to proceed, and returned to the strategic level in a later stage.

Organizational design

approaches

723

Design Consultants with a pragmatic strategy emphasized the need for flexibility in the design process. They worked one-step-at-a-time, monitored the process closely and shaped their ways of working along the way. Plans of approach were considered important, as a medium for discussing progress, coordinating efforts, and giving a sense of security, but they were drawn generally and filled in along the way. In contrast with the consultants with a rational approach, these consultant argued that at the beginning of the design process, they were still uncertain of many things and could not see the whole process clearly yet. In terms of consultant K, “Fixing one step is already a big deal, fixing two steps is tempting fate, and fixing three steps is playing God”. Making detailed plans of approach was considered a waste of time. Consultant G said:

You should never elaborate a plan in too much detail. It costs energy, and you focus your energy on thinking about what could go wrong, and that’s a waste. It often takes a lot of effort to prevent people from putting in too much energy at the beginning of a process by thinking it over meticulously. It only creates false certainty, because once you begin to make some progress, you realize that you are often in a different place than you thought you would be.

Designing was considered an open-ended process, in which room should be created for learning and experimentation. Several consultants created design spaces for developing and testing new designs. Consultant F created a simulation game, in which employees could experiment with new structures. Consultant G did a pilot within one office of the organization he restructured. The consultants emphasized that a design space should not include just a single designer or a group of key figures, because that under-utilizes the knowledge in the organization and might lead to implementation problems. A differentiated design space should be created, in which people participate in designing the part that concerns them directly. Consultant R phrased it as follows:

I strongly believe in effecting change from the bottom up. I strongly believe in involving professionals, workers, in designing their own changes. When you have defined the strategy, the hull of the organization, and the constraints, then I would [. . .] choose to let the people who actually do the work in the organization, to let them design their own work.

A focus in this approach is on the creation of momentum. The process of collaborative learning and experimentation was spurred on by the consultants to let new order emerge and prove its workings. Models and management concepts were used to create a common language for the involved co-designers, to coordinate the designing, and to trigger further actions. Making models on-the-spot, as part of the design process, was used as a way to stabilize the order that emerges. Decisions about the design were made gradually in the course of the process. In principle, the possibility to reframe when necessary was kept open by the consultants, but it did not occur in the discussed projects.

Evaluation Formative evaluation, i.e. the evaluation and improvement of the design during the process, is internalized in the pragmatic approach. In case of a summative evaluation, after the process, the design was evaluated as an integral part of the evaluation of the accomplished change. Has the organization actually changed and does it work better, judged from a posteriori requirements? A priori criteria did not play a role, because the

MD 48,5

724

statement of clear measures at the beginning of the design process was considered impossible when dealing with wicked problems. Table II summarizes the different design approaches.

Discussion and conclusion The purpose of this study was to develop insight into the variety of design approaches, a topic that has received little attention in management literature, notwithstanding the long tradition and recent revitalization of literature on organization design theory (Romme, 2003; Van Aken, 2004; Dunbar and Starbuck, 2006). We have developed a typology, in which the different activities that make up organizational design processes are elaborated in ideal/typical process configurations or design approaches.

Our study has shown that the activities of the traditional ADIE-models can be identified in each design process, but that these activities are shaped differently and do not occur in the same order within each approach. We have also shown that socio-political activities are present in each design process, but that they are handled differently in each approach. This implies that we can move forward from the traditional “either . . . or” discussion on design processes. Designing is not just a rational process (Simon, 1969), just a political process (Pfeffer, 1981), or just a learning process (Schön, 1983). In the rational approach, problem solving has a central place, but political activities are required to establish a protected space, to get commitment from the client and to prepare for implementation, and processes of learning and

Rational approach Dialogical approach Pragmatic approach

General characteristics Rational problem-solving Focus on content and control Bracketing content-related and socio-political complexity

Discussion and negotiation Focus on commitment and consent Absorbing socio-political complexity

Reflection-in-action, experimenting and learning-by-doing Focus on momentum and flexibility Absorbing content-related complexity

Analysis Establishing clear objectives Finding the real problem Consultants’ models to rationalize the analysis

Analysis of interests, power positions and problem definitions Models, developed with the client, to enhance communication

Open and thorough exploration Multiple models, to frame the situation and to coordinate the process

Design Searching for the best solution Rational choice techniques

Searching for a consented solution Political choice techniques

Searching for a working solution

Implementation After designing After designing Parallel to designing

Evaluation A priori criteria Client satisfaction (as second- best)

Key figure satisfaction A posteriori criteria Table II. A typology of design

approaches

Organizational design

approaches

725

experimentation can occur within the steps of the design process. In the dialogical approach, political processes are central, but the ADIE-activities structure the discussions and focus the attention of the stakeholders, and learning and experimentation activities may help at certain moments to open up entrenched positions in the political game. In the pragmatic approach, rational and political activities are employed to structure the process, to evaluate progress and to gain commitment for the experiments. In actual design processes, certain activities are put on the foreground while others remain on the background (Visscher and Fisscher, 2009). The alignment of these heterogeneous activities is necessary to make an approach work in practice (Law, 1987).

However, different strategies may also disrupt each other. A pragmatic strategy, for instance, depends on momentum in the experimentation process, and will slow down when a search for consensus or rational analyses requires too much time. And rational and dialogical strategies, in their turn, depend on analysis-based or consensus-based decisions, and become disrupted when these are postponed time and again to create room for further experimentation. Consultants can mix activities from other approaches in their main strategy, to support or to repair it, but should be careful not to disturb it.

In this study, we also made a contribution to the contextualization of organizational designing, which is relevant for academics and practitioners alike (Burnes, 1996; Van Aken et al., 2007). In particular, we have shown that the value of the phase-theorem (Witte, 1972; Lipshitz and Bar-Ilan, 1996) depends on the specific context. It seems valuable in the rational approach, under the conditions of relative simplicity and a defendable design space. However, in the dialogical and especially in the pragmatic approach, ADIE-activities occur in parallel and cycles, and attempting to keep a strict order is considered to be counterproductive.

On a more general level, the different design approaches can also be contextualized. Characteristics of the problem situation, the client organization and the repertoire of the consultant can be identified as conditions for the working of the different approaches. A rational design approach requires reduced or bracketed complexity. If a situation is fairly simple, which means that the problem and goals can be set clearly beforehand by the client and the consultant, and that the consultant has a well-developed repertoire to tackle the problem, and is that situation is also socio-politically uncomplicated, which means that there are no conflicting interests or that a strong and unanimous dominant coalition in the client organization can silence the dissenters, then a rational approach can work. In the cases of rational design in this research, these conditions were met. If these conditions are not met, the rational approach is likely to collapse along the way, as complexities and contingencies will challenge its chosen route and destination.

A dialogical design approach is applied in situations with considerable differences in interests and opinions and a lack of central power. If the social complexity can be reduced or bracketed, a dialogical approach is not appropriate, as it explicitly takes the differences in interests and opinions as a starting-point for discussions and negotiations. A rational approach is then more efficient and effective. In the cases of dialogical design in this research, the client organizations employed fairly autonomous professionals ( journalists, lecturers and union workers), and some were in a merging process, which made them socio-politically complex indeed.

MD 48,5

726

Pragmatic design approaches are applied in wicked (Rittel, 1972), ambiguous design situations. If an assignment is straightforward and a consultant has the expertise to tackle it, a rational approach can create a design much more efficiently and effectively. If the complexity is predominantly socio-political in its origin, a pragmatic strategy is not adequate either, because the processes of learning and experimentation then become influenced severely by political processes and the emerging design will reflect the existing socio-political structures in the organization (Pfeffer, 1981). In the pragmatic cases in this research, problems were often multi-layered and ambiguous, resulting in mutual uncertainty for both clients and consultants (Sturdy, 1997; Werr and Styhre, 2002), which implied that the design process could not be determined ex-ante (Fincham, 1999; Nikolova et al., 2009). The socio-political complexity in these cases remained limited, or was managed well on the background.

The assessment of the complexity of the problem situation can be helpful as a heuristic for organizational designers, but complexity is not supposed to determine the choice of an approach univocally. To a certain extent, complexity is malleable and negotiable, as some interviewed consultants pointed out, to make them fit a certain approach. A grey area exists for the possibilities to bracket complexity, since every situation is new, and designers have room to think optimistically that, with some luck and determination, they may be able to keep the complexity restricted. Besides, design approaches should not be seen as mere methods, which can be chosen at will, depending on the complexity of the situation. Over the years, practitioners develop a professional identity, shaped by a combination of experiences, talents, affinities, and successes, which predisposes their approach and their assessment of the complexity in a concrete design project (Visscher and Rip, 2003).

Further systematic research is required to establish the relation between context factors and design approaches more thoroughly. Complexity is one of these factors, but others such as the size and the knowledge management strategy of the consulting firm may also play a role. It can be hypothesized that the knowledge repertoire necessary to support a rational design strategy requires a consulting firm of significant size and a strategy aimed at codification and instrumentation (Hansen et al., 1999; Anand et al., 2007), while this is not required for a pragmatic or dialogical approach. The size and structure of the client organization may also play a role. The dialogical strategy is often applied in professional bureaucracies, and it may be hypothesized that rational strategies flourish in rational machine bureaucracies (Mintzberg, 1979). Furthermore, the role of the national context of the consulting and client organizations needs further exploration. As this study was carried out in the Dutch context, which is consensus-oriented, the attainment of commitment and the searching for compromises and consensus may have had more emphasis than it would get in other cultures.

A further direction for future research is to study design processes from the client’s perspective. In the current study, the point of view of the consultant was chosen, and the interaction with the client was pictured from this position. A study from a client perspective can enrich the approaches that have been found, and complement the consultants’ view of the different activities that constitute an organizational design process.

Organizational design

approaches

727

References

Akin, Ö. (1994), “Creativity in design”, Performance Improvement Quarterly, Vol. 7 No. 3, pp. 9-21.

Anand, N., Gardner, H.K. and Morris, T. (2007), “Knowledge-based innovation: emergence and embedding of new practice areas in management consulting firms”, Academy of Management Journal, Vol. 50 No. 2, pp. 406-28.

Ansoff, H.I. (1980), “Strategic issue management”, Strategic Management Journal, Vol. 1, pp. 131-48.

Baker, T. and Nelson, R.E. (2005), “Creating something from nothing: resource construction through entrepreneurial bricolage”, Administrative Science Quarterly, Vol. 50 No. 3, pp. 329-66.

Benders, J., van den Berg, R-J. and van Bijsterveld, M. (1998), “Hitch-hiking on a hype: Dutch consultants engineering re-engineering”, Journal of Organizational Change Management, Vol. 11 No. 3, pp. 201-15.

Boland, R.J. and Collopy, F. (2004), “Design matters for management”, in Boland, R.J. and Collopy, F. (Eds), Managing as Designing, Stanford Business Books, Stanford, CA, pp. 3-18.

Bucciarelli, L.L. (1994), Designing Engineers, The MIT Press, Cambridge.

Buchanan, D. and Boddy, D. (1992), Expertise of the Change Agent: Public Performance and Backstage Activity, Prentice-Hall, New York, NY.

Burnes, B. (1996), Managing Change. A Strategic Approach to Organisational Dynamics, 2nd ed., Pitman Publishing, London.

Butcher, D. and Clarke, M. (2003), “Redefining managerial work: smart politics”, Management Decision, Vol. 41 No. 5, pp. 477-87.

Carspecken, P.F. and Cordeiro, P.A. (1995), “Being, doing, and becoming: textual interpretations of social identity and a case study”, Qualitative Inquiry, Vol. 11, pp. 87-109.

Clark, T. and Fincham, R. (2002), Critical Consulting: New Perspectives on the Management Advice Industry, Blackwell Publishers, Oxford.

Cross, N. (2000), Engineering Design Methods: Strategies for Product Design, John Wiley and Sons, Chichester.

Dewey, J. and Bentley, A.F. (1949), Knowing and the Known, Beacon Press, Boston, MA.

Doty, D.H., Glick, W.H. and Huber, G.P. (1993), “Fit, equifinality, and organizational effectiveness: a test of two configurational theories”, Academy of Management Journal, Vol. 36 No. 6, pp. 1196-250.

Dunbar, R.L.M. and Starbuck, W.H. (2006), “Learning to design organizations and learning from designing them”, Organization Science, Vol. 17 No. 2, pp. 171-8.

Faust, M. (2002), “Consultancies as actors in knowledge arenas: evidence from Germany”, in Kipping, M. and Engwall, L. (Eds), Management Consulting, Oxford University Press, Oxford, pp. 146-63.

Fincham, R. (1999), “The consultant-client relationship: critical perspectives on the management of organizational change”, Journal of Management Studies, Vol. 36 No. 3, pp. 335-51.

Freeland, R.F. (1996), “The myth of the M-form? Governance, consent, and organizational change”, American Journal of Sociology, Vol. 102, pp. 483-526.

Glückler, J. and Armbrüster, T. (2003), “Bridging uncertainty in management consulting: the mechanisms of trust and networked reputation”, Organization Studies, Vol. 24 No. 2, pp. 269-97.

Grandori, A. and Furnari, S. (2008), “A chemistry of organization: combinatory analysis and design”, Organization Studies, Vol. 29 No. 3, pp. 459-85.

MD 48,5

728

Greenwood, R. and Hinings, R. (1996), “Understanding radical organizational change: bringing together the old and the new institutionalism”, Academy of Management Review, Vol. 21 No. 4, pp. 1022-54.

Greiner, L.E. and Metzger, R.O. (1983), Consulting to Management, Prentice-Hall, Englewood Cliffs, NJ.

Haberstroh, C.J. (1965), “Organization design and systems analysis”, in March, J.G. (Ed.), Handbook of Organizations, Rand McNally, Chicago, IL, pp. 1171-212.

Hammer, M. and Champy, J. (1993), Reengineering the Corporation: A Manifesto for Business Revolution, Harper Business Books, New York, NY.

Hansen, M.T., Nohria, N. and Tierney, T. (1999), “What’s your strategy for managing knowledge?”, Harvard Business Review, Vol. 77 No. 2, pp. 106-16.

Harris, M. and Raviv, A. (2002), “Organization design”, Management Science, Vol. 48 No. 7, pp. 852-65.

Heusinkveld, S. and Benders, J. (2005), “Contested commodification: consultancies and their struggle with new concept development”, Human Relations, Vol. 58 No. 3, pp. 283-310.

Heusinkveld, S. and Reijers, H. (2009), “Reflections on a reflective cycle: building legitimacy in design knowledge development”, Organization Studies, Vol. 30 No. 8, pp. 865-86.

Hickson, D.J., Hinings, C.R., Lee, C.A., Schneck, R.E. and Pennings, J.M. (1971), “A strategic contingencies’ theory of intra organizational power”, Administrative Science Quarterly, Vol. 16 No. 2, pp. 216-29.

Johnson, J.C. (1990), Selecting Ethnographic Informants, Sage, Newbury Park, CA.

Khandwalla, P.N. (1977), The Design of Organizations, Harcourt Brace Jovanovich, New York, NY.

Kvale, S. (1996), Interviews; An Introduction to Qualitative Research Interviewing, Sage, Thousand Oaks, CA.

Lang, J.R., Dittrich, J.E. and White, S.E. (1978), “Managerial problem-solving models: a review and a proposal”, Academy of Management Review, Vol. 3 No. 4, pp. 854-66.

Law, J. (1987), “Technology and heterogeneous engineering: the case of the Portuguese expansion”, in Bijker, W.E., Hughes, T.P. and Pinch, T.J. (Eds), The Social Construction of Technical Systems: New Directions in the Sociology and History of Technology, MIT Press, Cambridge, MA, pp. 111-34.

Lévi-Strauss, C. (1966), The Savage Mind, Weidenfeld and Nicolson, London.

Lipshitz, R. and Bar-Ilan, O. (1996), “How problems are solved: reconsidering the phase theorem”, Organizational Behavior and Human Decision Processes, Vol. 65 No. 1, pp. 48-60.

Meyer, A.D., Tsui, A.S. and Hinings, C.R. (1993), “Configurational approaches to organizational analysis”, Academy of Management Journal, Vol. 36 No. 6, pp. 1175-95.

Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis: An Expanded Sourcebook, 2nd ed., Sage, Thousand Oaks, CA.

Mintzberg, H. (1979), The Structuring of Organizations, Prentice-Hall, Englewood Cliffs, NJ.

Mintzberg, H. (1983), Power In and Around Organizations, Prentice-Hall, Englewood Cliffs, NJ.

Nikolova, N., Reihlen, M. and Schlapfner, J.-F. (2009), “Client-consultant interaction: capturing social practices of professional service production”, Scandinavian Journal of Management, Vol. 25 No. 3, pp. 289-98.

Pahl, G. and Beitz, W. (1996), Engineering Design: A Systematic Approach, Springer, Berlin.

Pettigrew, A.M. (1973), The Politics of Organizational Decision-making, Tavistock, London.

Organizational design

approaches

729

Pfeffer, J. (1978), Organizational Design, AHM Publishing, Arlington Heights, IL.

Pfeffer, J. (1981), Power in Organizations, Pitman, Marshfield, MA.

Rittel, H.W.J. (1972), “On the planning crisis: systems analysis of the first and second generation”, Bedrifts Okonomen, Vol. 8, pp. 309-96.

Rittel, H.W.J. and Webber, M.M. (1973), “Dilemmas in a general theory of planning”, Policy Sciences, Vol. 4, pp. 155-69.

Romme, A.G.L. (2003), “Making a difference: organization as design”, Organization Science, Vol. 14 No. 5, pp. 558-73.

Roozenburg, N.F.M. and Eekels, J. (1995), Product Design: Fundamentals and Methods, John Wiley, Chichester.

Schön, D.A. (1983), The Reflective Practitioner, Basic Books, New York, NY.

Schön, D.A. (1987), Educating the Reflective Practitioner: Toward a New Design for Teaching and Learning in the Professions, Jossey-Bass, San Francisco, CA.

Scriven, M. (1967), “The methodology of evaluation”, in Tyler, R.W., Gagné, R.M. and Scriven, M. (Eds), Perspectives of Curriculum Evaluation, Rand McNally, Chicago, IL, pp. 39-83.

Short, J.C., Payne, G.T. and Ketchen, D.J. (2008), “Research on organizational configurations: past accomplishments and future challenges”, Journal of Management, Vol. 34 No. 6, pp. 1053-79.

Simon, H.A. (1945), Administrative Behavior: A Study of Decision-making Processes in Administrative Organization, Free Press, New York, NY.

Simon, H.A. (1969), The Sciences of the Artificial, MIT Press, Cambridge.

Sturdy, A. (1997), “The consultancy process; an insecure business?”, Journal of Management Studies, Vol. 34 No. 3, pp. 389-413.

Suchman, L.A. (1987), Plans and Situated Actions: The Problem of Human-machine Communications, Cambridge University Press, Cambridge.

Van Aken, J.E. (2004), “Management research based on the paradigm of the design sciences: the quest for field-tested and grounded technological rules”, Journal of Management Studies, Vol. 41 No. 2, pp. 219-46.

Van Aken, J.E., Berends, H. and Van der Bij, H. (2007), Problem Solving in Organizations: A Methodological Handbook for Business Students, Cambridge University Press, Cambridge.

Van den Ven, A.H. and Joyce, W.F. (1981), Perspectives on Organization Design and Behavior, Wiley, New York, NY.

Van Strien, P.J. (1997), “Towards a methodology of psychological practice, the regulative cycle”, Theory and Psychology, Vol. 7 No. 5, pp. 683-700.

Verschuren, P. and Hartog, R. (2005), “Evaluation in design-oriented research”, Quality & Quantity, Vol. 39, pp. 733-62.

Visscher, K. (2006), “Capturing the competence of management consulting work”, Journal of Workplace Learning, Vol. 18 No. 4, pp. 248-60.

Visscher, K. and Fisscher, O.A.M. (2009), “Cycles and diamonds: how management consultants diverge and converge in organization design processes”, Creativity and Innovation Management, Vol. 18 No. 2, pp. 121-31.

Visscher, K. and Rip, A. (2003), “Coping with chaos in change processes”, Creativity and Innovation Management, Vol. 12 No. 2, pp. 121-8.

MD 48,5

730

Weick, K.E. (1993), “Organizational redesign as improvisation”, in Huber, G.P. and Glick, W.H. (Eds), Organizational Change and Redesign: Ideas and Insights for Improving Performance, Oxford University Press, New York, NY, pp. 346-79.

Werr, A. and Styhre, A. (2002), “Management consultants: friend or foe?”, International Studies of Management and Organization, Vol. 32 No. 4, pp. 43-66.

Werr, A., Stjernberg, T. and Docherty, P. (1997), “The functions of methods of change in management consulting”, Journal of Organizational Change Management, Vol. 10 No. 4, pp. 288-307.

Williamson, O. (1975), Markets and Hierarchies. Analysis and Antitrust Implications, Free Press, New York, NY.

Witte, E. (1972), “Field research on complex decision-making processes; the phase theorem”, International Studies of Management and Organization, Vol. 2 No. 2, pp. 156-82.

Yoo, Y., Boland, R.J. and Lyytinen, K. (2006), “From organization design to organization designing”, Organization Science, Vol. 17 No. 2, pp. 215-29.

Zeleny, M. (1982), Multiple Criteria Decision-making, McGraw-Hill, New York, NY.

Corresponding author Klaasjan Visscher can be contacted at: [email protected]

Organizational design

approaches

731

To purchase reprints of this article please e-mail: [email protected] Or visit our web site for further details: www.emeraldinsight.com/reprints

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.