Master of Science thesis
Executive Programme in Business Studies, specialization: Strategy University of Amsterdam, Amsterdam Business School
Alliances in project-based organizations: The influence of
selection criteria and managerial hierarchy on partner
selection processes.
31 January 2014 Student: Ir. Wouter Moorlag, 10476032 First supervisor: Dr. Joris J. Ebbers, University of Amsterdam, Amsterdam Business School Second supervisor: Drs. Erik Dirksen, University of Amsterdam, Amsterdam Business School2
Table of Content
Abstract 4 1 Introduction 5 2 Project organizations, alliances and partner selection 8 2.1 Project organizations 8 2.2 Alliances 9 2.3 Partner selection criteria 11 2.4 Partner selection process 13 3 Research method 16 3.1 Case study 16 3.2 Sample Selection 18 3.3 Data collection 20 3.4 Validity and reliability 21 3.5 Research ethics 22 4 Data Analysis 23 5 Results 24 5.1 General characteristics of the firms 24 5.2 Collaborating with other firms 27 5.3 Selection criteria 30 5.3.1 Internal selection criteria: task‐related or partner‐related 30 5.3.2 External selection criteria: client‐related 32 5.3.3 External selection criteria: partner‐related 39 5.4 Managerial hierarchy: selection criteria 41 5.5 Managerial hierarchy: decision making 44 5.6 Managerial hierarchy: partner selection process 483 6 Discussion 53 6.1 Internal and external selection criteria 53 6.2 Managerial hierarchy influencing partner selection processes 54 6.3 Partner availability and time pressure influencing partner selection processes 55 6.4 Managerial implications 56 6.5 Limitations and avenues for future research 58 6.6 Conclusions 59 Literature 60 Appendix 1: Interview protocol 68
4
Abstract
Recent studies have shown an increasing attention to time and temporality in management science, like project‐based organizations. These types of organizations are ideally suited for managing increasing product complexity, fast changing markets, innovation and market and technological uncertainty. Inter‐firm collaboration in project‐based consortia may help companies be innovative to meet the highly differentiated and customized nature of client demand. The alliance formation process can be an important determinant of a projects’ success. It is generally acknowledged that partner selection is one of the critical decisions a firm makes when forming an alliance and one of the main drivers of alliance success. A case study in the Dutch architecture industry has been performed to explore how selection criteria and managerial hierarchy influence partner selection in firms in the process of setting up consortia. Whereas previous studies focus almost exclusively on internal partner selection criteria, an important finding of our study has been that selection criteria are often externally determined, often due to specific tender requirements. These external partner selection criteria are either client‐related or partner‐related. Furthermore, the study argues that individuals, who work directly with partner firms, prioritize the selection criteria differently than those who don’t. In projects where the former have a large influence on the partner selection process, it is more likely that the partners will be selected primarily based on internal selection criteria, while projects where the later have a large influence, external selection criteria are likely to be more important. In addition, the study shows that partner availability and time pressure strongly influence the partner selection process. The process is often informal, fast and intuitive.5
1. Introduction
We live in a seemingly ever faster changing high pace world. To keep up with competition, firms operating in this increasingly challenging and dynamic environment must innovate in order to bring new or improved product offerings to the market. Inter‐firm collaboration may help companies to be innovative to meet the highly differentiated and customized nature of client demand. Clients want high‐speed, flexibility and reliability, which often requires a project‐based multidisciplinary approach that cannot be offered by one single firm. Especially globally operating small and medium sized firms are forced to form alliances. Whom to ally with? How to make sure the best partner is selected for the project and what influence does the temporal aspect have on this process? Temporary organizational forms like project‐based organizations emerge, in which diverse and specialized intellectual resources and expertise are integrated (DeFillippi and Arthur 1998; Hobday 2000; Gann and Salter 2000; Keegan and Turner 2002; Lindkvist 2004). While a project can be defined as any activity with a defined set of resources, goals, and time limit, a project‐based organization is a mechanism for coordinating and integrating the main business functions of the firm in a project (Hobday 2000). Project‐based organizations can be found in a wide range of industries, including consulting and professional services, cultural industries, high technology industries (Sydow et al. 2004) and can refer to either the entire firm, or multi‐firm consortia (Hobday, 1998). These temporary project organizations are a fast and flexible mode of an organization that can reduce traditional barriers to innovation, like large investment needs and a long time frame (Sydow et al. 2004). However, our current understanding of project‐based organizations is limited. More research is needed to form a strong theoretical base to understand the structures and processes that are relevant to temporary organizing principles (Bechky & Okhuysen, 2011; Jones & Lichtenstein, 2008; Mohammed & Nadkarni, 2011). Aspects of temporary organizing, like the transient nature, have6 received little research attention in organization studies (Bakker, 2010). There is a tendency to assume organizations like alliances are formed to last. This may be part of a broader neglect of explicit attention to time and temporality in organizational research (Ekstedt et al., 1999; Bakker et al, 2013). The alliance formation process can be an important determinant in a projects’ success. Firms have to make several decisions when forming an alliance: with whom to ally (partner selection), how to allocate responsibilities and authority between partners (governance structure) and what activities to perform (alliance scope) (Eden et al.,2008). This study will focus on the first decision, partner selection, often initiated by the project leading firm that is facing questions like: What kind of expertise do we need in our team? Should we go for a firm that we trust or do we form an alliance with a firm we do not know yet? How should we prioritize the different partner selection criteria? Who should we involve in the partner selection process? What is the best way to organize the partner selection process? These questions are relevant as it is generally acknowledged that partner selection is one of the critical decisions a firm has to make when forming an alliance (see, for example, Geringer 1991; Ireland Hitt & Vaidyanath, 2002). Partner selection is cited to be one of the main drivers of alliance success (Kale & Singh 2009; Schreiner et al, 2009). Especially small and medium sized enterprises (SME’s) working in project‐based organizations, often don’t have the time or lack resources to perform a thorough analysis of a potential partner for new projects. For example, concerns about the costs for a wider search can prevent firms from looking beyond their own existing pools of social relationships (Ellis, 2000), which may not necessarily lead to the best possible alliance. Even superior alliance management may not be sufficient to overcome poor initial partner screening and selection efforts. Partner selection is an important first step in the alliance formation process. Prior partner selection researchhas been heavily focused on motivations for forming alliances as well as the selection criteria used to evaluate potential partners (Geringer, 1991,
7 Brouthers et al. 1995, Child & Faulkner 1998, Doz & Hamel 1998, Ireland et al 2002, Bierly & Gallagher 2007). However, little attention has been paid to the alliance formation process. For example Cummings and Holmberg (2012) developed a tool to select the best partner. From their perspective, a potential partner would be deemed a “best fit” if its total weighed average scores on the relevant factors were to exceed those of other potential partners. As in other literature, Cummings and Holmberg don’t describe how this tool should be incorporated in the partner selection process in the firm. Important aspects of the inter‐firm partner selection process are not part of the study. How should firms allocate available resources to the partner selection process? Who will determine which selection criteria are used and what weight factor is given to each criterion? How do people in the firm at different levels in the hierarchy influence the partner selection process? How do directors, senior managers, project leaders and business development staff within the firm weigh these criteria? How does a firm come to their final decision? These are precisely the problems that are addressed in this study leading to the following main research questions: How do selection criteria that are used by firms setting up project consortia influence the partner selection, and what is the influence of hierarchy on the partner selection process? Crucial parts in the partner selection puzzle are currently missing. This undermines managerial implication of the existing literature. The rest of the study is divided into four mayor sections. The first section sets out the theoretical framework that guided our empirical study. A description of our case study performed in the Dutch architecture industry is given in the second section. We chose for the architecture industry since it is a fast‐moving, rapidly changing, competitive industry often organized through multi‐partner alliances in project‐based organizations. The third section describes the data analysis phase. The results will be presented in the fourth section. The final chapter will briefly summarize the case study results and draw conclusions.
8
2. Project organizations, alliances and partner selection
2.1 Project organizations Recently, there has been increasing scholarly interest to time and temporality in management science (Ancona et al., 2001; Mitchell and James, 2001) and rapidly growing temporary organizing principles in firms (Bakker, 2010; Bakker et al, 2013; Kenis et al., 2009). These temporary organizational forms, like project organizations, are becoming more and more common especially in our current high‐paced economy (Ekstedt et al. 1999; March 1995). Firms in all types of industries are undertaking projects, to a greater or lesser extent, as a growing part of their operations (Keegan and Turner, 2002). Project organizations are being put forward as a form that is ideally suited for managing increasing product complexity, fast changing markets, innovation and market and technological uncertainty (Hobday, 2000). In contrast to other organizational forms like matrix or functional, the project organization is one in which the project is the primary unit for production organization, innovation and competition (Lundin and Soderholm 1995, DeFillippi 2002). A project can be defined as any activity with a defined set of resources, goals and time limit (Hobday, 2000). Some industries have a long tradition of project organizations, like filmmaking (Bechly, 2006; DeFillippi and Arthur, 1998), theatre (Goodman and Goodman, 1972, 1976) and construction (Eccles, 1981; Gann and Salter 2000). Other industries are now increasingly adopting project organizational structures, including software development, advertising, biotechnology, consulting, fashion and architecture, among others (Bechky, 2006; Engwall, 2003; Ekstedt et al., 1992; Morley and Silver, 1977). Project organizations share the fact that they operate in highly innovative, fast changing industries in a complex environment. Hobday (2000) distinguishes project‐based organisations from project‐led organizations. The former are firms that organize most of their internal and external activities in projects. The latter is a milder form of a project‐based organisation and refers to firms that undertake projects while their primary ‘productive’ activity is volume‐based or operations‐oriented. In these firms the needs of projects outweigh the functional influence on decision‐making and representation to senior9 management, but some coordination across project lines occurs. This article focuses on how firms set up a project‐based organization defined as a temporary organization that dissolves as soon as the project, for which it was set up, is completed (DeFillippi & Arthur, 1998; Jones, 1996). Project‐based organizations can either be entire firms or multi‐firm consortiums (Hobday, 1998). This also includes project ecologies linking many different types of individual, organizational en institutional partners, who are co‐located within a field of cooperation and mutual influence (Grabher, 2002). Capabilities from outside are needed to complete firm capabilities. Multi‐ partner alliances between firms are therefore common in project‐based organizations. However, most literature on alliances focuses on long‐term collaboration. An explicit focus on transience and limited duration is ever more pressing since this is the core of the organization of temporary systems. 2.2 Alliances In the last 30 years inter‐organizational project ventures like alliances have become more and more common (Hagendoorn, 2002). Alliances have become a central part of firms’ competitive and growth strategies. I.e. the number of international R&D alliances has increased during the past decade because they have become a major means of developing and/or acquiring technological assets (Narula & Duysters, 2004). In this study we define an alliance as a voluntary arrangement involving durable exchange, sharing, or co‐development of new products and technologies of two or more partners (Gulati, 1995). An inter‐firm cooperation that falls between the extremes of discrete, short‐term contracts and the complete merger of two or more organizations (Contractor and Lorange, 2002) The involvement of outside project partners provide firms with flexible network solutions through limited duration partnerships (Jones and Lichtenstein, 2008; Schwab and Miner, 2008). Alliances can improve a firms’ competitive position by enhancing market power (Kogut, 1991), increasing efficiencies (Ahuja, 2000), accessing new or critical resources or capabilities (Rothaermel & Boeker, 2008), and entering new markets (Garcia‐Canal, Duarte, Criado, & Llaneza, 2002).
10 Child and Faulkner (1998) identified five alliance motivations categories: transaction‐cost motivations, resource‐based motivations, strategic‐positioning motivations, learning motivations and other motivations. Doz and Hamel (1998) identified three primary alliance formation motivations: co‐ opting potential rivals and complementary firms, achieving co‐specialization by combining complementary resources and learning and internalizing valuable skills. Multi‐firm consortiums refer to alliances comprising a mix of different specialist competences, which have to achieve a certain goal within the limits set as costs and time. Clients regularly interact with the consortium about the often innovative design of products and services (Hobday, 1998). Multi‐partner alliances have gradually emerged in many industries to cope with the tremendous resource requirements and/or risks involved in large‐scale projects. Multilateral alliances have the same value creation logic as bilateral alliances, yet the involvement of more than two participants complicates alliance design and governance (Doz & Hamel, 1998). Various studies have suggested that alliances accelerate the innovativeness and learning of the organizations involved (Cohen & Levinthal 1990; Goes & Park, 1997; Hagedoorn & Duysers, 2002; Lee, Lee & Pennings 2001). However alliances are complex organizational forms, involving tacit, non‐ routine and highly uncertain knowledge conversion processes fraught with ambiguity (Anand & Khanna, 2000). Therefore alliances also tend to show high failure rates (Dyer, Kale, & Singh, 2001). Studies have shown that between 30% and 70% of alliances fail, which means that they neither meet the goals of their parent companies nor deliver on the operational or strategic benefits they are supposed to provide (Bamford, Gomes‐Casseres, & Robinson, 2004). Janowicz‐Panjaitan, Bakker and Kenis (2009) conclude that inter‐organizational collaborations are significantly understudied. This is surprising, especially considering the strategic importance and increasing use of inter‐organizational collaborations in recent years (Muthusamy and White, 2005). The success of any alliance depends on some key factors that are relevant at each stage of alliance evolution (Gulati, 1998). These include (a) the formation phase, wherein a firm deciding to
11 initiate an alliance selects an appropriate partner, (b) the design phase, wherein a firm (and its partner) setup appropriate governance to oversee the alliance, and (c) the post formation phase, wherein a firm manages the alliance on an ongoing basis to realize value (Schreiner, Kale, & Corsten, 2009). This study focuses on the formation phase, and especially the partner selection. Prior research has suggested that the choice of a particular partner is an important variable affecting alliance performance (Mohr & Spekman, 1994; Park & Ungson, 1997; Parkhe, 1991), since it influences the mix of skills and resources which will be available to the alliance and thus the alliances ability to achieve its strategic objectives (Geringer, 1991). Even superior alliance management may not be sufficient to overcome poor initial partner screening and selection efforts. An important first step in the alliance formation process, which will be discussed in the next paragraph. 2.3 Partner selection criteria Firms must make three decisions when forming an alliance: with whom to ally (partner selection), how to allocate responsibilities and authority between partner (governance structure) and what activities to perform (alliance scope) (Eden et al.,2008). The three sets of decision are interrelated in that each decision is likely to affect the other two. This paper focuses on the partner selection decision. Researchers identified many explanations of why a firm selects a particular partner firm for an alliance. Brouthers et al. (1995) present a framework for analysing the likely success of alliances. They identified four criteria for evaluating alliances: complementary skills, cooperative skills, compatible goals and commensurate levels of risk focused on asymmetric transfer of information or competencies. Geringer (1988) and Geringer (1991) added a distinction between task‐related selection criteria and partner‐related selection criteria. Das and He (2006) summarize the prior alliance partner selection literature as containing the following criteria. First, task‐related criteria: Complementary products or skills; financial resources; technology capabilities or uniqueness; location; marketing or distribution systems, or established customer base;
12 Reputation and image; managerial capabilities; government relationship, including regulatory Requirements and government sales; help in faster entry into the target market; and industry attractiveness. Task‐related selection criteria are associated with the strategic resources and skills that a firm requires for its competitive success, and are more concerned with an achievement of strategic fit between partners. We also included learning aspects and performance risk in this category. Second partner‐related criteria: Strategic fit or interdependence, or compatible goals; compatible or cooperative culture and ethics; prior ties and successful prior association; trust between top managers; strong commitment; similar status, including size and structure; reciprocal relationship; commensurate risk; and ease of communication. Partner‐related selection criteria however are close to the notion of how partnering firms can work together effectively, and are thus more cooperation‐related with a focus on achieving organisational fit. We also include relational risk in this category. We will discuss a couple of selection criteria more into depth. First, scholars have emphasized resource complementarity between firms (Richardson, 1972; Nohria & Garcia‐Pont, 1991). Complementarity is the extent to which a partner contributes non‐overlapping resources to the relationship, such that one partner brings those resources or capabilities the other lacks and vice versa (Dyer & Singh, 1998; Harrigan, 1988; Mowery, Oxley, & Silverman, 1996). Resource‐based theories suggest that the greater the complementarity between partners the greater the likelihood of alliance success and many studies have found support for this. Second, scholars have emphasized the role of status and reputation. They argue that firms will consider the status of potential partners when forming alliance. Some argue that firms are likely to ally with firms of similar status (Podolny, 1994; Chung et al, 2000). Other argues that relationships between firms of different status do exist as well, i.e. in alliances between new technology‐based ventures and incumbent organizations. (Baum et al., 2005; Li & Rowley, 2002; Stuart, 2000; Castellucci & Ertug, 2010; Rothaermel & Deeds, 2004; Diestre & Rajagopalan, 2012).
13 Third, scholars have emphasized the role of the social compatibility in alliance formation (Bourdieu and Wacquant, 1992; Coleman, 1990; Gulati, 1995a). To mitigate risk from imperfect information about potential partners, firms are thought to rely on exchanges embedded in two forms of social capital: past ties (i.e. prior experiences between participants, path dependence) and third‐ party ties (indirect exchanges through direct partners). A partner firm must be compatible with the focal firm (Beamish, 1987) and committed to the relationship. Partner compatibility refers to the fit between partners’ working styles and cultures, whereas commitment includes not only the willingness of a partner to make resource contributions required by the alliance, but also to make short‐term sacrifices to realize the desired longer‐term benefits (Gundlach, Achrol, & Mentzer, 1995). An important aspect in partner selection is prior collaboration experience and trust. Hypothesizing preference transitivity in partner selection may be too simplistic in that a prior partnership has two effects on partner selection: increasing trust which can lead to lower transaction costs and decreasing ‘information asymmetry’ between partners, which ease knowledge transfer. The latter also has a contradictory effect because it can increase the risk of opportunistic behaviour by the partner firm. When knowledge leakage costs are high, the firm might prefer a new partner to prior partner. Trust is necessary for the parties to make a good‐faith effort toward achieving mutual goals and to not take excessive and unilateral advantage of each other, even when the opportunity to do so is available (Sabel, 1993). Moreover, trust increases the likelihood that organizational actors will exchange information and ideas because current exchanges are the most likely to be positively reciprocated at a future time (Collins & Smith, 2006). 2.4 Partner selection process Although more and more firms are part of a project‐based organization, significant numbers of alliances failed (Dacin et al. 1997; Dyer, Kale, & Singh, 2001; Arino, 2003). One of the most important reasons is the incompatibility between partners (Buyukozkan et al., 2008), this means that partner selection becomes a strategic focus point for firms (Cowan et al., 2007). Partner selection
14 policies may affect the temporary organizations performance and even determines the organizations’ ability to survive. It is therefore an important strategic issue in the formation of temporary organizations, but few scholars pay much attention to it (Feng et al., 2009). The question of why firms seek allies is studied extensively but the question why firm choose a certain partner over others has received less attention (Chung et al, 2000). Understanding how partners are selected is important from both the theoretical and managerial perspective, since cooperation often has a social dimension as well as an economic dimension (Blau, 1964). Prior partner selection research has been heavily focused on motivations for forming alliances and selection criteria factors to evaluate potential partners against (Geringer, 1991, Brouthers et al. 1995, Child & Faulkner 1998, Doz & Hamel 1998, Ireland et al 2002, Bierly & Gallagher 2007). Little attention has been paid to anotherimportant piece of the partner selection puzzle, namely the partner selection process. This process could be described as the way firms should manage the partner selection process optimally at different hierarchical levels. Recent studies are focussing more on the process of how firms learn to collaborate (see, for example, Feller et al., 2013). Very little scholarly attention has been paid to the partner selection process in general, despite work emphasizing that partner selection is one of the critical decisions a firm makes when forming an alliance (see, for example, Geringer, 1991, Ireland, Hitt & Vaidyanath, 2002). In many studies the focal temporary organization is the unit of interest. However, the intra firm process of partner selection on an interpersonal level for temporary organization is important as well, in order to get a better understanding of why firms ally with a certain partner. Most of the studies are now performed on a macro‐level, the level of the organization, instead of individuals that populate them. But how do different hierarchical levels in the organization influence the process? What about the director, business developer or project leader that has influence on the partner selection? What criteria do they use to select a partner and does this differ per management level? Recent research on hierarchy addresses the paradox that despite all organizational change towards flatter and post‐modern organization, hierarchical order is quite persistent. Also modern
15 architecture firms are said to be still very hierarchical with top‐down power and control mechanisms more comprehensive than ever before (Brown et al, 2010). By differentiate understanding of hierarchy as either formal or informal Diefenbach en Silliance (2010) research reveals that hierarchy is much more widespread than thought. Important aspects of the inter‐firm partner selection process have not been studied. How should firms allocate available resources to the partner selection process? Who will determine which selection criteria are used? What weight is given to these selection criteria? Who is allowed to give advice on potential partner? How do people in the firm at different levels in the hierarchy influence the partner selection process? How is the final decision made? How should firms manage the partner selection process? It is precisely this problem that is addressed in this paper leading to the following main research questions: How do selection criteria that are used by firms setting up project consortia influence the partner selection, and what is the influence of hierarchy on the partner selection process? A crucial part to make the partner selection puzzle complete which complements existing literature.
16
3. Research method
3.1 Case study The aim of this research is to understand the partner selection process of firms and to understand the role of the selection criteria and managerial hierarchy. A case study approach in the Dutch architecture industry was considered most appropriate for the research objectives, to do theory building, rather than theory testing and to gain in‐depth insight into the process and the how and why’s (Eisenhardt 1989; Yin 2009). Furthermore a case‐ study approach was considered most appropriate to identify partner selection criteria, to analyse the impact of the criteria and to understand how the criteria are used by different people that are involved in the decision making process and to understand when and why a certain selection criterion is more important than another. To lower the amount of ‘environmental noise’ a single industry approach is adopted to have an in depth understanding of the firms in the industry (Eisenhardt 1989). The contemporary architecture industry is a fast‐changing competitive industry often organized through multi‐partner alliances in project based organizations. One of the strengths of this industry setting is that partner selection is a daily routine for most firms. It was expected to find a sizable group of architectural firms who often work in multi‐partner alliances. Dutch architects have an international focus and are renowned for their design capabilities. Another strength of choosing this industry is that the study deals with the same kind of project based organizations, minimizing the unobserved heterogeneity that can arise in the study in multiple industry settings. The architecture industry has not yet received significant attention in organizational research. The unit of analysis is the consortium, a project‐based organization (see figure 1). The partner selection process for a specific design consortium will be examined and compared to the firm’s general approach for selecting partners. The empirical basis of this study will consist of six case studies from six different firms. Finding out how in a project the partner selection process was17 organized is a delicate matter to discuss. The exploratory nature of this study calls for a qualitative research strategy. The choice of using a case study methodology is related to the notion that ‘the interaction between a phenomenon and its context is best understood through in‐depth case studies’ (Dubois & Gadde, 2002).
Figure 1: Unit of analysis, the consortium Case studies are a unique method for building and utilizing understandings of an empirical phenomenon that may develop existing theory. Furthermore, using the case study method will allow for retaining a holistic and meaningful characteristic of real‐life events such as organizational and managerial processes and relationships related to company strategies for product development in networks and related sources of conflict (Yin, 2009). The case study is a good tool to bring formal and informal interpersonal aspects into the partner selection process. In this study, a multiple‐case design has been used concerning both multiple social contexts and multiple events. The multiple‐ case design has been chosen in an endeavour to expose variety in reasons for partner selections and to expose a variety in partner selection processes in different selected firms. Partner 1 I.e. other architecture firm Partner 2 I.e. installation advisor Partner 3 I.e. construction advisor Partner x I.e. interior designer Consortium for a project P1 P2 P3 Px Lead architect Architecture firm (project leader, selects partners)
18 3.2 Sample Selection A theoretical sample is developed, rather than a random one (Yin, 2009), that include only Dutch architecture firms with an established reputation for being known as internationally active at multiple projects simultaneously and often have a leading role in forming consortia. The firms in the study where chosen on the basis of a study performed by the Royal Institute for Dutch Architecture (BNA) to the 30 largest architecture firms in the Netherlands. Furthermore, the selected firms are member of BNA iTop, a high‐level network of 15 internationally active Dutch architecture firms. Over 60% of all employed architects in the Netherlands work in firms that are member of the Royal Institute for Dutch Architecture and most of the larger architecture firms are member. The unit of analysis is the consortium that is setup by the respondent’s firm for a certain design project (case). The cases will be selected in advance and will all share the same characteristics. The architecture firm must have the lead role / be the project leader in the consortium. At least 2 Individuals must be involved in the partner selection process in the lead organization. At least 3 external partners must be involved in the project, including at least 1 other spatial design firm, like architects, interior designers or landscape architects. The project has to be an architectural project and in the beginning of the design process, i.e. Concept Design or Schematic Design. The design process often consists of 4 phases in which the level of detailing increases every time the projects succeeds to the next phase (see figure 2). Only one firm we contacted refused to participate; the response rate was 86%. The criteria have been chosen in order to reduce the variables that can influence the partner selection process. If an architect is not leading, then the partner selection process has a different goal and outcome. Projects must have a certain size in order to properly examine the partner selection process. Projects must be in the same stage of the design process to be comparable, since partners can change during the process and can have a different role.
19 Low High Figure 2: Architectural design phases Architecture firms are in general organized the following way. The directors are often partners in the firms, often referred to as principle architects. Senior architects form the line below the directors. They are the project leader and manage the other architects and designers. Other staff departments like public relations, administration and business development are often in place to support the directors and senior architects. Business development departments are involved in the partner selection. They often keep track existing partners on record, identify new partners and organize the application for the tender process, where input from partners is often needed as well. They often have a non‐architectural background. See figure 3 for a complete overview. Figure 3: Typical architecture firm organization Directors / partners
Senior architects Senior architects Senior architects
architects / junior architects / designers / drawers Support: Public relations / administration / business development Concept Design Schematic Design Design Development Construction Drawings Level of detail in drawings
20 Next to the architectural design activities, most architecture firms have several areas of activity. Some are involved in large scale design projects like urban planning, some are involved in small scale projects like interior design as well. The mode of generalization is analytic generalization rather than statistical generalization. The contributions from the study will develop existing research and knowledge. The analysis of the research will be qualitative, highlighting patterns. The findings will be compared to theory in order to describe the patterns which exist. An inductive approach will be used in order to develop theory as a result of the research findings, although there will be an element of deduction in that the structure of the research will be based on the academic literature. 3.3 Data collection The research is based on a series of case studies using semi‐structured interviews with key informants like partners/directors, senior architects and business development managers directly involved in and sometime responsible for partner selection and others if found appropriate. To collect empirical data from the firms, key informants who are knowledgeable about the issues being researched and who made the crucial decisions concerning these issues were interviewed. In order to select the cases that comply with the criteria, a questionnaire was shared with the contact respondent in each case study firm prior to the semi‐structured interview. The goal of the questionnaire was to select one suitable project per firm, to give insight in the contextual conditions of the projects, to gain general information about the firm such as number of employees, key activities and office locations and to select respondents at different hierarchy levels in the organization. In one case only directors were involved in the partner selection. This case is included in the study in order to have insights in how having no hierarchy influences the partner selection. A semi‐ structured interview was held with two or three respondents in each case study. An interview protocol was made. The interview included open‐ended non‐project related questions and specific open‐ended questions about the selected project. This proved to be a useful method, since
21 respondents could relate their answer to the selected project. I.e. respondents could indicate whether or not the partner selection criteria used in the selected project are typical criteria used by the firm. An in‐depth understanding of the selected project could therefore be related to the context (Yin , 2009). Each respondent was asked to answer identical questions. Most of the interviews were held in English. Three respondents indicated that they did not feel comfortable to speak in English. These interviews were held in Dutch. The method of selection of firms of case studies was through personal networking and sampling to find a number of organisations with the required characteristics. I have professional experience in the Dutch architecture industry. Effort was made to select a range of firms with these characteristics. Six organisations were required to make a comparison of partner selection processes. The short listed potential firms are all internationally operating Dutch firms and within the top 30 of most important and relevant architectural offices in the Netherlands that are seen as leaders in their field. A journal was kept that includes field notes and reflections on the progress of the study. 3.4 Validity and reliability In order to achieve internal validity multiple sources of evidence from semi‐structured interviews and questionnaires were used. The design of questions was based on understanding of the theory from the literature, and a pilot test of the interview which was used to make sure questions were understood as intended. This pilot test is also included in the results. Furthermore, transcripts of interviews were sent to the interviewees for verification. Triangulating the data by source, rather than by method, was a useful means of improving the validity and credibility of information and ensured that viewpoints represented varying levels of responsibility within the case firms. External validity will be achieved through the use of multiple cases to examine whether findings can be generalised over a number of firms. The use of multiple cases have strengthened the
22 external validity and provided the basis for analytical generalization (Yin, 2009). Reliability is pursued through semi‐structured interviews with questions derived from the literature in order to examine cases in the same way. The interviews and questionnaires were all administered in the same time frame and in controlled manner, with explanation to ensure participants all understood the research in the same way. All interviews were recorded which enabled us to have direct access to the original discussion and pay specific attention to any part of it at later stages. To minimize memory failure problems of the respondents, questions were asked about the recent past and only key informants having an intimate knowledge of the subject of study were interviewed (Huber and Power 1985; Kumar et al. 1993). The willingness with which the respondents participated in the research and the interest they showed during the interviewing process provided a further reason to trust the face validity of their responses (Papadakis et al. 1998). Respondents showed interest in organizing a follow‐up discussion and the Royal Institute of Dutch Architects will make the results available to all members. 3.5 Research ethics Respondents were given a clear written description of the purpose, scope and intended outcomes of the research. The type of information required for the research was clearly stated, as was the policy for anonymity and confidentiality. The research is carried out in a way that ensured confidentiality of the participant firm and the individual participants in the interviews. The goal was to encourage more candid and accurate responses to sensitive issues. The participating respondents are not named in the study. The interview questions and questionnaire were designed to examine only the practices of partner selection. No confidential product or customer information was required.
23
4
Data Analysis
An in‐depth analysis approach was used to uncover key themes. Insights from the grounded theory approach and extended case method were used as a guide to data analysis. The extended case method approach goes through many cycles of confrontation between data and theory. Each iteration is directing the analyst to additional data and drawing on additional concepts and theories. The extended case method consists of two ‘running exchanges’. One between literature review and data analysis, and one between data analysis and data collection (Burawoy 1991). ‘The first running exchange involves the interplay of existing concepts and theories and analysis of empirical data. Intensive analysis of the data and exploration of the scholarly literature occur in conjunction. Data analysis points to relevant concepts and theories in the literature, while simultaneously the literature provides conceptual frameworks to aid in the interpretation of the data. The second running exchanges calls for continuously moving back and forth between data collection and analysis’ (Glaser, 2004). During all interviews, informants were encouraged to illustrate their statements with specific events and examples from specific projects. Six cases were studied. Data from the last two case studies resulted in minimal incremental understanding (Lee, 1999). The first step in the analysis was the thorough reading of the interview transcripts, notes and questionnaires, looking for themes and patterns (Miles and Huberman, 1994). Critical passages were highlighted and coded, and initial interpretations were recorded in marginal notes. Memos were generated in the process, little pieces of insights. The emergent theoretical interpretations contained in the memos with the evidence from each case were systematically compared to assess how well or poorly they fit with the case date (Eisenhardt, 1989). This iterative process of constantly comparing emergent theory and data led to additional and often more qualified and refined memos.24 Facts and statements of the respondents and their firms and the memos were grouped to define conceptual clusters in relation to different topics in order to condense data and combine the evidence. Clustering in topics, creating data displays and showing the frequency of different events, made it easier to note regularities and patterns in the data set. Conceptual clusters are sets of closely related analytic ideas. The following clusters were introduced in the analysis: ‘collaboration with other firms’, ‘selection criteria’, ‘managerial hierarchy and ‘selection process’. These conceptual clusters formed the basis of the organization of the presentation of findings. To verify the first preliminary conclusions emerging from the data, the original field notes were re‐read to ensure that no possible similarities and differences might be overlooked (Miles and Huberman 1994). Emerging insights were checked with some of the respondents, asking for their feedback. When we felt the initial conclusions to be sufficiently tested for their ‘sturdiness’, we translated them into final conclusions. The results of the case analysis of the architecture industry are presented below. The findings are grounded by using verbatim examples from the interviews and descriptive statistics. The verbatim examples are sometimes slightly changed in order to improve the readability of the examples. Spelling errors and grammar errors were taken out and replaced by correct words or sentences. Dutch words, sentences or expressions were translated into English. It should be noted that the statistics presented are used to illustrate statements rather than to imply conclusions based on frequencies.
25
5
Results
5.1 General characteristics of the firms Table 1 lists the firms included in the study, together with their areas of activities, office locations, number of countries actively involved in, age, number of employees (in full time equivalent, FTE). By agreement with the participants, the firms included in the study are not named in this table. When reporting results from the data analysis, we made sure that none of the firms nor respondents could be identified. FirmAreas of activity Offices # countries Age (years) # FTE 1 Architecture, Urban Planning 2 10+ 35 60 2 Architecture, Urban Planning, interior architecture 3 3+ 15 40 3 Architecture, Urban Planning, interior architecture 2 3+ 30 50 4 Architecture, Urban Planning, interior architecture 2 20+ 25 100 5 Architecture, Urban Planning 2 15+ 20 70 6 Architecture, Urban Planning, interior architecture 2 10+ 20 10 Table 1: Selected firms ‐ two or three respondents per firm As table 1 show, all firms have their headquarters in the Netherlands. The firms in the study all work internationally. In comparison to the industry’s average the firms are relatively large with an average of 55 employees. The six firms in the study account for about 5% of the workers employed in the entire Dutch architecture industry. The firms listed in table 1 are all registered architecture firms and member of the Royal Institute for Dutch Architecture.
26 In general most of the respondents understand that partner selection and other organizational related topics are important aspects that determine the success of a firm. However, in general they are more interested in designing and the design process, especially the respondents who are architects feel that way. ‘It is a creative company and most of the architects and directors are not interested at all in the business side of this business. Organisation or anything related is seen as a nasty necessity, really.’ (Business Development Manager, firm A) Most of the respondents indicate that they would like to improve the partner selection process in their firm. Most of them mention that doing part of the partner selection beforehand would be beneficial to the process. They indicate that it would be helpful to have a list or database of potential partners and already do a quick scan in advance. Better preparation would save time and might lead to other and better partners, illustrated by the following quotes: ‘[it would be good] to have a better network of possible partners. Like a certain kind of agreement with other firms in which you state that you will work together if the right occasion occurs. If the fundaments for collaboration are there, you can act very quickly. A sort of quick scan so you already know whether or not we should collaborate together with office X or Y. In my ideal situation, there is a network of partners from different perspectives in place. (...) so you can have different selection criteria for different jobs in different parts of the world, in different areas, with different firms. But not too many, or you will end up with endless possibilities.’ (Director, firm E)
27 A similar idea is confirmed by the following director who indicates that his firm should approach the selection more strategically and should be better prepared: ‘It would be good to make a list of potential partners in advance. We don’t have that now. We just start most of the partner selection processes from scratch, but it would be useful to already identify potential partner without having a specific project in mind. This would lead to better and strategically more interesting partners, and would help to cope with the short time frame when a projects comes in’ (Director, firm F) However, some respondents indicate that having a good intuition instead of a good preparation is more important. Architecture firms work in a fast changing environment, so preparation is not always effective. ‘Choosing a partner is based on intuition (…) Since you have to make some decisions in a split second, I think that this is the best way to do it. You could make more lists and try to make a database in which you evaluate former collaborations. However, there is so much change within our own company, within the companies we work together; the client is always another one. You still have to make a lot of decisions based on intuition. It makes no sense to make a science of partner selection.’ (Business Development Manager, firm D) 5.2 Collaborating with other firms The architecture industry is more and more organized through multi‐partner alliances in project based organizations. All respondents in the study acknowledge that collaborating with other firms is therefore getting more and more important and a very important strategic aspect of running their business. All respondents indicate that in most of the projects they have to work in consortia nowadays. This has changed rapidly in recent years.
28 ‘The nature of the projects that we do, dictates a large number of specialists that we need around us (...). In most projects we work in consortia nowadays’ (Director, Firm C) Multi‐partner consortia have gradually emerged in the architecture industries to cope with the tremendous requirements and/or increasingly complexity of projects. Partners are needed to bring in those resources or capabilities the other lacks and vice versa. Architects cannot do the entire project themselves. ‘[collaborating is] extremely important because we only offer a part of the services. We could not work alone, so we are always working in consortia’ (Business Development Manager, firm A) The number of the so‐called integrated projects, like total engineering or public private partnership (PPS) projects, in which architects need to team‐up with contractors and other value chain partners, is increasing as well. This is illustrated by the following quote: ‘[collaborating with other firms is] more and more important. (…) We almost always have to tender together with advisors or a local architect. In the Netherlands there are more and more total engineering projects, PPS projects, design and build etc. There is a trend to cooperate from the beginning with other firms.’ (Business Development Manager, firm B) According to the following senior architect, knowledge from partner firms is crucial in order to be innovative. Input from partners is needed in order to keep up with competition.
29 ‘Innovation comes from the questions we raise and the models we study but can only exist by the input and knowledge from specialists, so [collaborating] is very important. (Senior architect, firm A) The following director indicates that collaboration is getting more and more important because of three reasons: New European tender laws, increasing complexity of projects and the need for a type of architecture that is focussing more and more on combining knowledge from multiple firms: ‘I believe the stronger need for collaboration is the result of a couple of trends. First, of course, the European tender system. Architects have to collaborate with other firms in order to present the right reference projects to the client that will give them a high score. Second, the changing demand of clients. Projects are getting more complex and the requirements are huge. Architecture firms cannot do all of the work anymore by themselves but need specialists to help them out. Third, I think architecture nowadays is focusing less on the final image but more on the process. There is a new generation architects that is very knowledge focused and less focused on designing. They combine knowledge from different sources and like to learn from other people’ (Director, firm F) One director summarizes the situation as follows: ‘In the whole industry getting and sharing knowledge is much more important than before. The days that you as an architect were able to provide all the services that the client wanted have mostly passed. In some architecture projects you can still do all required tasks, but for the bigger assignments and especially for international jobs, you always search for a good
30 companionship with other people or you search to make a consortium. (…) It is much more important than 5 years ago’ (Director, firm C) To summarize, collaborating with other firms in consortia is becoming more and more important. Architecture firms work in consortia in most of the projects. The work for architects is getting more complex and specialized and therefore specialised firms are needed in the consortium. 5.3 Selection criteria The findings in this section emerged from the interplay of literature, especially on partner selection criteria. The literature distinguishes task‐related criteria from partner‐related criteria. Task‐ related selection criteria are more skill‐related, concerned with an achievement of strategic fit between partners. Partner‐related selection criteria are more cooperation‐related with a focus on achieving organisational fit (Das & He, 2006). 5.3.1 Internal selection criteria: task‐related or partner‐related The respondents in the study use a wide variety of criteria when selecting a partner. E.g. Complementary expertise, reference projects, trustworthiness, similar work attitude, quality of work, and reputation. All 35 different selection criteria are listed in table 3. Due to increasing complexity of projects and requirement of specialist knowledge most of the respondents acknowledge that partners should have complementary skills. Some indicate that having partners with similar skills could even create friction in the team. ‘Before you choose or select a partner you want to work with, you have to ask yourself: What are we missing? What could the other party bring that makes you a better company?’ (Business Development Manager, firm D)
31 ‘In order to make a good project nowadays, you need this complementary, knowledge, skills, and experience in order to do the project in the best possible ways’ (Director, firm E) However, one director indicates that partners do not always need to be complementary. Sometimes it is just nice to collaborate with another firm that is not complementary, because it can lead to interesting discussions and new insights during the design process: ‘Complementarily is important for a project, however, we can also collaborate very well with a firm that is not complementary to us. It is just a lot of fun and can create interesting debates… As long as there is a mutual respect’ (Director, firm F) Furthermore, respondents indicate that having a partner with a similar work attitude on various levels is very important as well: ‘I think there needs to be a kind of connection or overlap. Could be style, or could be on a more social level. Something that tells you that it’s easy to work with them (...). They don’t have to work exactly the same of course, but somehow they give you the feeling that it’s going to be a pleasant way of working’ (Senior architect, firm E) ‘Most important selection criterion is the way that you collaborate with the other firm. How smooth the collaboration goes and that you have the feeling that they can listen to what you have to say and that together you can make the best project’ (Director, firm F) Apart from being complimentary and having a compatible work attitude, reliability and having a good reputation are often mentioned among respondents as important selection criteria.
32 ‘An ideal partner should add value to the project, (…) is reliable and also has a good name to the outside world. Then you have the strongest possible team’ (Director, firm C) Status and reputation differences between firms seem to be less relevant when selecting a partner. Sometime a high‐status or high reputation firm would collaborate with a low status firm because this firm has a unique skill or experience, which would give the team an advantage when the client rates the team. 5.3.2 External selection criteria: client‐related Apart from the above stated task‐related and partner‐related criteria that strive to a strategic or organizational fit between firms, most of the respondents in the study indicate another type of selection criteria that is more externally driven, the external selection criteria. These criteria can be divided into two categories: client‐related and partner‐related. What client‐related selection criteria are will be illustrated by the following quotes. ‘I came up with the idea to involve firm [name] based in the UK. It had to do with how the tender document was set up. We could gain extra points for previous projects where we had been cooperating on. It is strange that we had to collaborate with a foreign office. You would prefer to do this Dutch project with a Dutch office (…) However, it was the only way to get the highest amounts of points. The weird thing is that the tender decides which partner is the best’ (business development manager, firm B) ‘How it works in most of the competitions is that you are trying to fill in the scoring document on the basis of the criteria of the client’ (director, firm C)
33 ‘The people in the office who have suggested taking part in a certain competition often have already formed an idea of which partners would be the best to take into the team following the requirements of the tender. Sometimes it is very specific what kind of expertise should be involved in the consortium. (…) It wouldn't make sense to go for the first idea if the requirements of the tender are not fulfilled’ (director, firm E) In other words, sometimes partners are selected not based on task‐related or partner‐ related criteria. The decision driver is an external source, often the tender procedure or indirectly the client. In comparison to the more internally orientated task‐related and partner‐related criteria, the described criteria are actually often externally determined. We therefore call them external selection criteria and they are often the result of the tender procedure which is very common in the current architecture industry. A tender procedure is the process of making a proposal in response to an invitation or request for tender. Tendering organisations seek other firms to respond to a particular need, such as the supply of services, and will select a firm, or group of firms, that meets their needs. In general, tendering processes are used for procurements or contracts involving substantial amounts of money. I.e. in 2014 national governments must use a tender procedure according to EU tender rules for all procurements exceeding 204.000 euro. However, tendering can also be used by private sector firms. Tender request documents outline what the requesting organisation’s needs are. These documents also outline the particular requirements, selection criteria, and instructions that are to be followed. Interested suppliers will then prepare a tender. This can include documents that outline the proposal, provide information on qualifications, competencies and experience. The submitted tenders are then evaluated with regard to defined criteria. The five offers that best meet all of the requirements outlined in the request (i.e. highest amount of points), will be selected for the next phase. In architecture it is common that for the second phase multiple firms are invited to take part
34 into a design competition. The selected firms need to make a design that will be evaluated. The winner of the second phase wins the contract. In figure 4 the tender process is described. Figure 4: Tender process contracting authority‐ phase 1. Clients predefine in the first phase of the tender what kind of expertise they want in a consortium and what other criteria they use. In the architecture industry it is very common that clients ask for reference projects, turnover numbers and use of quality systems. The tender request documents often describe a rating mechanism, so architectural firms can ‘calculate’ what kind of partner firms they need in order to get a certain amount of points. Table 2 shows a rating mechanism that was used of a tender for a new building of the European Investment Bank. Table 2: Tender rating structure for new building European Investment Bank Luxembourg in 2012 Criterion No. Criterion name Points by criterion Allocation of points by sub‐ criterion 1 Quality of the team 50 1.1 Integrated team approach 25 1.2 Multidisciplinary 25 2 methodology 50 2.1 Design Methodology 20 2.2 Execution methodology 30 3 References 55 3.1 Architecture 15 3.2 Structural engineering 10 3.3 Technical installations engineering 10 3.4 Integrated Design and Project Team 10 3.5 Building Information Modelling 10 4 Quality of services 15 TOTAL 170 Step 1 Prepare Step 2 Request Step 3 Receive Step 4 Evaluate Step 5 Award
