Employees' agency in the formalisation of knowledge-intensive business service processes : A cross-case comparison

1 Employees’ agency in the formalisation of knowledge-intensive business service processes: A cross-case comparison Abstract Purpose – This study aims to identify and explain how different kinds of knowledge-intensive business service processes (KIBS processes) can be formalised without excessively limiting employees’ agency, and thus flexibility in value creation. Previous research acknowledges the need to balance flexibility and formalisation but does not investigate how employees achieve this balance in various types of KIBS processes. Design/methodology/approach – This paper employs a qualitative multiple-case study approach to compare employees’ agency in six successful formalisation projects targeting different types of KIBS processes in three firms. Through a systematic mapping of employees’ agency across the cases, this study reveals alternative patterns of formalisation that enable agency. Findings – The findings reveal notable differences in employees’ agency in the studied cases. When KIBS processes were scale-intensive and/or the culture favoured conformity, formalisation projects were centrally organised, and employees obeyed codified process templates, even though some agency remained. When KIBS processes were smaller scale and/or the culture favoured freedom, employees conducted formalisation projects autonomously and retained higher levels of agency in formalised KIBS processes. Practical implications – Firms and business units providing KIBS should choose their formalisation approaches locally based on the scalability of the KIBS process, their employees’ skill levels, knowledge bases and culture. Choosing the right approach enables flexibility while striving for efficient processes. Originality/value – Previous studies suggest that formalisation benefits only some KIBS, but this comparative approach shows that a variety of KIBS processes benefit from formalisation if the formalisation approach is tailored to the context. Alternative patterns of formalisation are revealed to guide further empirical research on the topic. Keywords: Formalisation, agency, value-creation processes, knowledge-intensive business services (KIBS) Authors: Tiina Tuominen & Miia Martinsuo This is the accepted manuscript of the article, which has been published in Journal of service theory and practice, 2019, 29(1), 45-70. https://doi.org/10.1108/JSTP-10-2017-0184 2 1. Introduction Service firms must be able to flexibly support customers’ value creation (Brozovic et al., 2016). Often, such flexibility depends on employees’ agency, that is, employees’ ability to exercise individual judgment in value-creation processes (Feldman and Pentland, 2003) in order to match the firm’s and the customers’ processes and resources in meaningful ways (John et al., 2006). Employees’ agency plays a major role in value-creation processes related to knowledgeintensive business services (KIBS) (hereafter KIBS processes), such as sales and delivery processes in software, business-to-business (B2B) consultancy and research and development (R&D) services (Strambach, 2008; Aarikka-Stenroos and Jaakkola, 2012). However, many firms are formalising their KIBS processes to increase efficiency and quality (Jaakkola, 2011; Rahikka et al., 2011; Bettiol et al. 2015). This paper defines formalisation as intentional efforts to make service sales and delivery processes more uniform and replicable by limiting variation and generalising knowledge (e.g., Organ and Greene 1981; Feldman and Pentland, 2003; Pentland and Feldman, 2008). The study aims to identify and explain how different kinds of KIBS processes can be formalised without excessively limiting employees’ agency, and thus flexibility in value creation. Previous studies report different effects of formalisation on employees’ agency. Formalisation can support agency by enabling knowledge sharing and creating shared meaning, transparency and trust (Kärreman et al., 2003; Brivot, 2011; Rahikka et al., 2011). It can also excessively constrain agency if it limits employees’ ability to support customers’ situated value creation (Morris, 2001; Vaast and Levina, 2006; Sandholtz, 2012). These outcomes may depend on how the formalisation project is implemented — particularly on how employees are taken into account (Vaast and Levina, 2006; Brivot, 2011; Sandholtz, 2012; Wright et al., 2012). Furthermore, some KIBS processes may be more amenable to formalisation than others 3 (Consoli and Elche-Hortelano, 2010; Pina and Tether, 2016), which indicates that employees need more freedom in some KIBS processes. However, these differences have not been systematically and empirically studied at the level of the KIBS process. Many studies assume a uniform firm-level approach to formalisation (Morris and Empson, 1998; Morris, 2001; Kärreman et al., 2003; Rahikka et al., 2011; Bettiol et al., 2012) or analyse experiences at the individual level (Haas and Hansen, 2005; Skjølsvik, 2016), without making comparisons across KIBS processes. Because comparative studies are lacking, it is not known exactly how employees’ agency should be taken into account in different types of KIBS processes. Understanding employee perspectives in greater detail can significantly contribute to the service studies addressing flexibility and formalisation (Brozovic et al., 2016; Skjølsvik, 2016). In addition to making employees’ work meaningful, choosing a suitable formalisation approach for each KIBS process benefits firms and customers by supporting flexibility (Kärreman et al., 2003; Brivot, 2011). To address this topic, this study compares employees’ agency in six formalisation projects and in their outcomes, formalised KIBS processes, in three firms. The firms represent different businesses (management consultancy, insurance and software services), and the selected formalisation projects target KIBS processes that varied in terms of type of service, knowledge base and scale intensity of the process. The firms’ employees perceived all six projects as successful. By comparing these successful formalisation projects, the study creates new knowledge about how different kinds of KIBS processes can be formalised without excessively limiting employees' agency. The paper demonstrates that formalisation can benefit many types of KIBS processes and reveals alternative approaches that firms use to enable employees’ agency and thus flexibility in formalised KIBS processes. These findings can be further developed and tested in future empirical studies. 4 This paper begins by discussing the existing literature on the nature of KIBS processes, employees’ agency in KIBS and the formalisation of KIBS processes. Next, the research methods and data are presented. The findings sections compare employees’ agency across formalisation projects and formalised KIBS processes and analyse the potential explanations for the differences across the cases. The discussion section considers the implications of the findings for research and practitioners. The paper ends with a summary of the limitations of this study and suggestions for further research. 2. Theoretical Framework 2.1 Employees’ agency and flexibility in KIBS processes Services are generally viewed as processes (Lovelock, 1991; Edvardsson et al., 2005) which comprise performances that require flexibility (John et al., 2006). Employees’ agency – their ability to influence their work, respond to new situations and make informed situated judgments (Feldman and Pentland, 2003) – ensures that these flexible performances support value creation and meet customer expectations (Ojasalo, 2001; Edvardsson et al., 2012). Flexibility and agency are inherent parts of any service process, but they are highlighted in KIBS processes where employees develop unique solutions for customer firms (Alvesson, 2004). KIBS might be offered by specialised KIBS firms or by other firms as part of their B2B offerings, and they involve knowledge as an important input and output (Alvesson, 2004). In KIBS processes, employees display high skill levels and professional autonomy when solving complex problems (Alvesson, 2004; Strambach, 2008; Aarikka-Stenroos and Jaakkola, 2012). Agency enables improvisation and new knowledge creation (Aarikka-Stenroos and Jaakkola, 2012; Fosstenløkken et al., 2003). Employees can also transfer tacit knowledge directly across customer contexts (Strambach, 2008) and reinforce creativity in service innovations 5 (Giannopoulou et al., 2014). As new solutions spread among employees, these solutions gradually change KIBS processes (Fosstenløkken et al, 2003; Werr and Stjernberg, 2003). Therefore, KIBS processes are often continuously developing (Alvesson, 2004). 2.2 Formalisation as a trend influencing KIBS processes The traditional view suggests that KIBS processes are informal and employees’ agency is guided by professional expertise, tacit norms and collegial control (e.g. Robertson and Swan, 2003). In contrast, many studies report the increased formalisation of KIBS processes, along with service standardisation (Bettiol et al. 2015), modularisation (Rahikka et al., 2011) and productisation (Jaakkola, 2011; Valminen and Toivonen, 2011). Informal processes are viewed as ambiguous and inefficient while striving for efficient, scalable and profitable processes (Morris, 2001; Briscoe, 2007). Also, the effective application of knowledge is suggested to increase customer value (Martelo-Landroguez and Martin-Ruiz, 2016). This paper defines formalisation as intentional efforts to make service sales and delivery processes uniform and replicable by limiting process variation and generalising knowledge (e.g. Organ and Green 1981; Feldman and Pentland, 2003; Pentland and Feldman, 2008). Formalisation is connected to the standardisation of process templates (e.g. Sandholtz, 2012) and the codification of knowledge (e.g. Brivot, 2001; Vaast and Levina, 2006). This paper assumes that formalisation increases the importance of such codified/standardised resources in KIBS processes (Organ and Greene 1981; Morris and Empson, 1998; Feldman and Pentland, 2003) – they are typical outcomes of formalisation projects (Feldman and Pentland, 2008). Here these codified resources are referred to as ‘codified process templates’, comprised of written rules, service models, process blueprints and tools which are defined in formalisation projects. 6 Formalisation can be seen to change two dimensions of KIBS processes in that it 1) increases the importance of codified process templates, which 2) subsequently changes how employees are allowed to exercise agency in KIBS processes. This study uses the term ‘formalisation’ instead of ‘standardisation’ to stress that firms can codify the process templates, but they cannot completely standardise how employees use such templates in KIBS processes. Even though codified process templates enable and constrain employees’ agency in a desired way, some agency must remain; the employees always have some embodied knowledge and some discretion in their actions (Alvesson, 1993; Edvardsson and Olsson, 1996). Thus, employee and customer interactions cannot be replicated in a standardised manner. In particular, frontline interactions depend on employees’ situated judgment and interpersonal skills (Alvesson, 1993). 2.3 Employees’ experiences of the enabling and constraining effects of formalisation Empirical studies provide mixed findings regarding the outcomes of formalisation for employees in KIBS. Table 1 presents the key aims, findings and new research gaps described in studies addressing formalisation from an employee perspective in KIBS firms, such as professional service firms. These studies, along with studies addressing KIBS processes in other types of firms (Kelley et al., 1996; Sandholtz, 2012; Vaast and Levina, 2006) and studies with somewhat different primary research focuses (Valtakoski and Järvi, 2016; Wright et al., 2012), have shown that formalisation can indeed yield benefits for employees that go beyond cost efficiency. Rahikka et al. (2011) found that the modularisation of construction engineering services helped customers to trust the firm. Brivot (2011) showed that codification increased transparency without creating resistance when employees retained control of their work in a law firm. Thus, increased process consistency may improve the professional reputation of employees (Brivot, 2011; Wright et al., 2012). Furthermore, Wright et al.’s (2012) and Kärreman et al.’s (2003) studies reported that codified templates created shared meanings 7 which helped employees cope with ambiguity. In addition to business services, studies addressing knowledge-intensive processes in other settings provide similar findings. Kelley et al. (1996) demonstrated that formalisation reduced inappropriate deviations in consumeroriented bank and insurance services. Briscoe (2007) observed that formalisation increased the temporal flexibility of doctors by decreasing the person-dependency of their work. ---- Insert Table 1: Empirical research on employee viewpoints on the formalisation of KIBS processes in KIBS firms ---- However, employees may also perceive formalisation as harmful and limiting their agency too much. It can prevent improvisation in novel customer situations, and employees’ status might be threatened (Morris, 2001). Haas and Hansen (2005) revealed that codification helped junior consultants but was time-consuming for experienced consultants. Vaast and Levina (2006) showed how standardisation failed in an IT organisation because employees and customers were not listened to, while Sandholtz et al. (2012) demonstrated that professionals could not easily internalise standards created by other experts. Studies in other contexts suggest that if codified process templates do not support actual value-creation processes, employees may deviate from the templates either intentionally or unintentionally (Orlikowski, 2000; Pentland and Feldman, 2008), leading to the decoupling of the process templates from actual processes (Sandholtz, 2012; Gondo and Amis, 2013). 2.4. Employees’ agency in formalisation projects and KIBS processes The above review indicates that formalisation can enable employees’ agency if it creates shared meaning, transparency and trust, thereby clarifying employees’ work and helping them to make 8 appropriate situated judgments (Kärreman et al., 2003; Brivot, 2011; Rahikka et al., 2011; Wright et al., 2012). Formalisation can constrain employees’ agency too much if their freedom is excessively curtailed, their necessary person-dependent knowledge is downplayed, or the codified process templates are not suitable for their work (Morris, 2001; Haas and Hansen, 2005; Vaast and Levina, 2006; Sandholtz et al., 2012). These findings resonate with the idea that formalised organisation can be either ‘enabling’ or ‘constraining’ (Adler and Borys, 1996; Adler, 2012). One pressing question concerns how to enable, instead of excessively limit, employees’ agency in KIBS processes under various circumstances. The above studies on KIBS specifically and formalisation more generally point out two issues. First, previous researchers call for employee involvement in formalisation projects so that their practical expertise can be captured within the codified process templates (Vaast and Levina, 2006; D’Adderio, 2008; Pentland and Feldman, 2008; Sandholtz, 2012). When employees and other important actors are involved, trust one another and strive toward similar goals, formalisation projects can clarify complexities and create a collective understanding of the best practices (Adler and Borys, 1996; Valtakoski and Järvi, 2016). In this way, formalisation projects can leverage employees’ skills, thereby helping employees understand the principles of value-creation processes, which will in turn make them more capable of flexibly using codified process templates in customer situations (Adler, 2012). Second, attention must be paid to exactly how formalisation projects change KIBS processes. The studies mentioned above indicate that codified process templates should leave sufficient room for employee agency and allow employees to identify instances which require creativity (Kärreman et al., 2003; Brivot, 2011; Wright et al., 2012). This is possible if organisational goals are congruent with employees’ goals, power and other resources are distributed equally and codified templates allow for adjustments based on practical experiences (Brivot, 2011; 9 Adler and Borys, 1996; Adler et al., 1999). Under the opposite conditions, formalised KIBS processes can become coercive, lowering employees’ motivation and value creation capabilities. 2.5 Different approaches to formalisation for different types of KIBS processes Most of the studies discussed above analyse formalisation in either singleor two-case firmlevel settings (Morris and Empson, 1998; Morris, 2001; Kärreman et al., 2003; Rahikka et al., 2011; Bettiol et al., 2012) or in individual-level cross-sectoral surveys (Haas and Hansen, 2005; Skjølsvik, 2016). Many of the studies assume a firm-level knowledge management system or strategy (Morris and Empson, 1998; Morris, 2001; Haas and Hansen, 2005; Brivot, 2011; Bettiol et al., 2012) and do not identify differences in employees’ agency across KIBS processes between and within firms. These process-level differences are important to consider because there is considerable heterogeneity across KIBS sectors (Asheim et al. 2007; Consoli and Elche, 2010, 2013; Pina and Tether, 2016). Some KIBS processes may benefit from a rigid formalisation approach, whereas others might not. Previous research suggests that three characteristics of KIBS processes may influence formalisation outcomes: the type of KIBS process and employees’ related skill levels, the type of knowledge base, and scale intensity of the process. Consoli and Elche (2013) suggest that technology-intensive KIBS processes, excluding R&D, are amenable to formalisation because the spectrum of problem-solving options is known and front-line employees’ skill levels are relatively low. In contrast, professional service processes are difficult to formalise because they require situated problem solving and complex cognitive and social skills (Consoli and Elche, 2010). In professional services room for improvisation is needed and formalisation should focus on work methods instead of content and outcomes (Alvesson, 1993; Kärreman et al., 2003; Bettiol et al., 2012). 16 details) and evaluated against their experiences of challenges in other formalisation projects. Secondary data strengthened the analysis: follow-up interviews revealed how KIBS processes evolved after formalisation, documentation ensured adequate understanding of the studied services and workshop memos supported the analysis of the effects of formalisation. Two types of formalisation projects were identified in each of the three firms, for a total of six cases. Table 4 presents the cases and the reasons for formalisation, changes and interviewees’ estimates of the successful features in each case. ------------------------------ Insert Table 4: Cases: six formalisation projects ------------------------------ The second step was a within-case analysis addressing RQ1 and RQ2. This step iterated with the third step, a cross-case analysis, which helped to categorise the observations. Case-specific observations were first summarised, paying attention to employees’ agency in formalisation projects and KIBS processes. Then, these observations were compared across cases to identify general categories. In response to RQ1, three ways in which employees were allowed to exercise their agency in formalisation projects were identified. These categories were labelled participative centralisation, inclusive centralisation and local autonomy based on concepts used in the previous literature (e.g. Adler and Borys, 1996; Quick and Feldman, 2011) (see Chapter 4.2). With regard to RQ2, how employees were allowed to exercise their agency in formalised KIBS processes was analysed in each case, and the findings were categorised into five types, ranging from the ability to decide how to combine codified process modules in customer cases to the ability to deviate completely from the codified process template. Also, the degree of the codification of process templates was analysed because the initial analysis suggested that this influenced employees’ agency but also that the type of templates and agency could vary independently. Process templates were categorised into two types: partially or 17 completely codified process templates. These categories are explained in Chapter 4.3. Table 5 defines the empirical categories identified for RQ1 and RQ2. ------------------------------ Insert Table 5. Empirical categories and their features ------------------------------ The third step was a cross-case analysis in which the findings from different cases were compared (resulting in the categories presented above) and explanations for the differences between the cases were sought (RQ3, see Chapter 4.4). A replication tactic was used (Miles and Huberman, 1994), that is, recurring patterns were identified in the categories shown above. These patterns were then tested and modified via systematic case comparisons. Initially, all variations between cases were taken into account. Attention was paid to the situation before formalisation, particularly to the types of services, KIBS processes and knowledge bases. This analysis revealed a relationship between agency and KIBS process type. The feature most distinctively related to agency was the scale intensity of KIBS processes. Scale intensity was also linked to skill levels and knowledge bases. Additionally, cultural norms were found to be important, particularly in explaining within-firm differences (such as between Cases 1 and 2 at ManCo). These findings were compared with those of previous studies to identify and assess the theoretical explanations of the study, as well as the study’s implications for further research. 4. Findings 4.1 Overview of the aims and outcomes of formalisation projects in the three firms Formalisation helped the studied firms to develop larger or cross-organisational service offerings. It supported employees’ agency when selling and delivering such offerings by clarifying the function and role of specific KIBS processes, which in turn helped achieve a 18 shared understanding of the service and integrate activities that were necessary for the larger offerings. ManCo’s consultants combined their expertise to develop collective consultancy projects. SoftCo aimed to increase its project sizes and recruit new consultants. InsCo aimed to combine the expertise of its subsidiaries in a customer-centric manner. The interviewees estimated that after formalisation, services were easier to buy and sell, which increased sales across all three organisations. Formalisation made the work tangible enough to allow for profitable pricing and helped in the training of new employees. It also helped anticipate the interactions and outputs of different actors, thereby decreasing dependence on individual employees. These benefits, summarised in Table 4 above, resemble the benefits identified in previous studies (e.g. Kärreman et al., 2003; Brivot, 2011; Rahikka et al., 2011; Adler, 2012). However, the interviewees also noted that some employees and customers were dissatisfied with codified process templates that were created in formalisation projects due to individual antipathies toward or lack of skills in applying such templates. This indicates that formalised KIBS processes still relied on employees’ willingness and skills. Because these negative perceptions were exceptions and thus rarely mentioned in the studied cases, the formalisation projects were generally considered successful. Overall, the analysis suggests that two facts may explain the positive perceptions of formalisation. First, the interviewees agreed on the need to clarify KIBS processes. Second, formalisation was tailored to the context, so it allowed employees to exercise their agency in a manner that was necessary and possible. Next, differences in employees’ agency are discussed. Section 4.2 presents the differences regarding how employees were involved in formalisation projects (in response to RQ1). Section 4.3 analyses the differences in employees’ agency in 19 formalised KIBS processes (in response to RQ2). Finally, Section 4.4 suggests explanations for the differences between cases (in response to RQ3). 4.2 Employees’ agency in formalisation projects Instead of relying solely on specialised service development staff, each formalisation project featured attempts to utilise the practical experiences of employees. Nevertheless, the manner in which employees participated in the projects varied from merely providing ideas to carrying out formalisation autonomously. Here, the different ways of organising formalisation projects are labelled participative centralisation, inclusive centralisation and local autonomy (see Table 5). Each firm seemed to take a different approach: participative centralisation was identified at InsCo, inclusive centralisation at SoftCo and local autonomy at ManCo. In participative centralisation at InsCo, formalisation projects were centrally coordinated. Internal developers were in charge of formalisation projects and invited selected frontline employees to present ideas in specific development phases. This approach resembles the ‘enabling formalisation’ proposed in earlier studies of technology firms, in which employee involvement is considered important but formalisation is still a centralised effort (e.g. Adler and Borys, 1996; Adler et al., 1999). For example, in InsCo Case 1 (cases are hereafter referred to as C1, C2 and so forth), formalisation was centralised because the aim was to reduce firmand person-specific variations in KIBS processes. Cross-organisational groups, comprised of developers and support personnel, identified common customer needs, developed solutions and created marketing and sales materials, which were delivered through training to employees. Frontline employees and users presented ideas and comments in development workshops, but decisions were made within the project groups. Frontline representation was important in grasping customer needs: 20 The staff at the headquarters [the developers], who don’t have many customer contacts, they increasingly take our opinions into account; they listen to our views on customer needs and problems in our processes. We have noticed that since the collaboration became more intensive, we have been able to tackle the problems faster and more efficiently. (InsCo’s sales team leader) In inclusive centralisation at SoftCo, formalisation was also centrally controlled but employees could also exercise their agency in decision making (see Quick and Feldman, 2011), as opposed to merely presenting ideas. Both managers and employees were involved in project groups. Experienced employees had the deepest expertise of the KIBS processes, but managers wanted to ensure scalability of processes to support the firm’s growth. For example, SoftCo C2 focused on processes related to complex services, offered independently of software products. Initially, the services were customised because customers’ situations and skills varied considerably. Experienced consultants were key actors in formalisation, because they possessed much of the required knowledge. They defined the process templates gradually through experimenting with various tactics in customer projects. The identification of successful process templates required hands-on expertise: It [the service] is relatively young, and we have still been searching for what the service is really about; this is what we have been searching for, and we have developed it through trial and error. And we did not want to freeze it yet, like ‘this is it’. (SoftCo Consultant No. 1) Local autonomy was identified at ManCo, where consultants designed their own team-specific KIBS processes. The teams had complete responsibility for their services, and their decisions were not coordinated at the firm level. The teams decided whether and how to pursue formalisation. Their decisions depended on team culture and led to different outcomes. Also SoftCo had autonomous formalisation projects in the beginning, but it centralised its formalisation projects as its service areas expanded. The findings regarding local autonomy 21 differ from previous studies, which present formalisation as a centralised project (e.g. Adler and Borys, 1996), but still resonate with studies showing that expert employees’ autonomy is typical of professional service firms’ innovation activities (e.g. Sundbo, 1997). 4.3 Employees’ agency in formalised KIBS processes The effects of formalisation projects on KIBS processes varied along these two dimensions: 1) the degree of codification of process templates (partially or completely codified templates) and 2) how employees were allowed or expected to use their agency when applying these templates in individual customer cases. These categories are shown in Figure 2 (see also Table 5). ------------------------------ Insert Figure 2: Formalised KIBS processes: codification of process templates and employees’ agency ------------------------------ Codified process templates defined the goals, scope, pricing models, contents, scripts and methods applied in KIBS processes, either completely or partially. In completely codified templates, all of these elements were defined. These templates were introduced at InsCo and in ManCo C2. None of these templates defined a single, standard way but resembled processes typical for modularised offerings in which alternative ways of responding to customer needs are defined (Bask et al., 2011). Partially codified templates were introduced at SoftCo and ManCo C1. There, only some elements were defined, while others were specified or created by employees in individual customer cases. The process templates did not seem to fully determine employees’ agency. What mattered more were the norms regarding how employees were allowed to apply the templates in customer 22 cases. In some cases, strict compliance was expected, whereas in others, employees were allowed to deviate from the templates either partially or completely. Based on these observations, employees’ agency in the formalised KIBS processes was categorised into three types. In the first type, the process templates were completely codified and deviations from the templates were not allowed, but employees could decide how to combine pre-existing process modules in customer cases. They exercised their agency when interpreting customer needs and forming customer-specific solutions. The formalised KIBS processes at InsCo (C1 and C2) and in C2 at ManCo fall under this category. At InsCo C1, service packages were accompanied by rigidly formalised B2B sales and service delivery processes to reduce firmand person-specific variations across InsCo. InsCo C2 focused on the formalisation of micro-level sales and customer care approaches so that every salesperson could easily start selling packages. Instead of presenting product details, the salespersons had to address the customer-specific business situation with compatible solutions. Formalisation aimed to ensure that everyone could do it, without leaning on individual customisation: The most important thing is that the sales models and customer-centred solution models should follow the same template, regardless of the part of the country in which the customer interaction takes place and the office in question… (InsCo Developer No. 1) One of ManCo’s teams (ManCo C2) followed a similar approach. Services and process templates were carefully codified, and their modules were changed only if the entire team agreed; individual improvisations were not allowed. The team wanted to ensure high-quality concepts and minimise entrepreneurial risks for team members. Its members even signed a team contract that ensured mutual care, an appreciation of team values and compliance with collective templates. The members preferred rigidly formalised processes, even though it contradicted the values nurtured in other teams: 23 We are systematic and disciplined; we have created models and methods that others [other teams] do not have. And we have [heard others say] we are straight-laced. We do not perceive it to be so. We have realised that once you have clear, collectively developed principles, you get a clear method that follows the principles. (ManCo Consultant No. 1) In the second type, process templates were partially codified, and employees were allowed to develop and apply new customer-specific modules. Cases at SoftCo fall under this category. In SoftCo C1, the goal was to increase and simplify service sales by standardised project scopes and predefined project modules, whereas previous projects had been developed from scratch. [We had] very little replication, and in every project, we had to wonder, from the perspective of the consultant, what to put in it and, from the perspective of the customer, what to offer. Then, we began to map systematically – okay, what are the services that we are actually already delivering, and could we make packages of some sort from them? (SoftCo Unit Manager No. 1) Clear process guidelines decreased the firm’s dependency on individual experts and helped train new consultants quickly. Nevertheless, some deviations were allowed because the range of customer problems could not be fully known prior to service delivery. In SoftCo C2, the range of customer problems was even broader, and deviation was more common. This was because services in SoftCo C2 were new and complex. Because customers’ situations and skills varied considerably, these processes could not be rigidly formalised. Even after formalisation, employees could develop new knowledge and resources to solve customer problems. In the third type, process templates were partially codified, and employees were allowed to deviate completely from the codified process template. ManCo C1 represents this category. Many teams at ManCo introduced partially codified process templates in order to explicate and leverage their collective expertise. These process templates clarified commonalities in their 24 activities and articulated collective offerings with names, value propositions, contents and project scopes. However, the application of these formalised templates in KIBS processes varied because the consultants retained the freedom to choose whether to offer the formalised services or develop their own, individual projects. This freedom was considered important not only in resolving customer problems but also in motivating consultants: I think that particularly in [our] team, many of us are not motivated to do routine projects – and one of my aims, why I do everything we have done [in our team], is that I am constantly looking for interesting variations for myself, like ‘wow, a new thing again, nice’. (ManCo Consultant No. 2) 4.4 Explanations for differences in employees’ agency in the studied cases The above findings show that the cases varied in terms of both RQ1 and RQ2: employees exercised their agency in different ways in formalisation projects and formalised KIBS processes. To analyse potential explanations for these differences, attention was paid to the types of services, the knowledge bases, and the scale intensity of the KIBS processes, as well as other differences between cases. This analysis suggests that the scale intensity of the KIBS process explains much of this variation and that scale intensity was also associated with the skill level and integration of knowledge bases underlying the KIBS processes. Table 6 compares the studied KIBS processes and responses to RQ1 and RQ2. For simplicity, employees’ agency is categorised from the lowest (+) to the highest (+++) levels of agency. ------------------------------ Insert Table 6: Cross-case comparison of KIBS process types and employees’ agency ------------------------------ Scale intensity ranged from KIBS processes which were replicated by small, intensively interacting teams (such as at ManCo) to processes replicated by functionally and 25 geographically dispersed teams (such as at InsCo). Scale and technology intensity of InsCo’s processes seemed to explain why employees’ agency in formalisation projects and formalised KIBS processes was limited. InsCo’s process templates had to ensure consistent quality and efficiency in situations where tens or hundreds of employees delivered the service. Furthermore, the frontline employees did not have all the necessary knowledge to design such effective and profitable processes. The development and delivery of services were structurally separated and involved different types of knowledge. Developers possessed the analytic knowledge required in the development of profitable and scalable processes. Frontline personnel held the synthetic knowledge that was required in customer cases (Strambach, 2008; Pina and Tether, 2016). The findings indicate that participative centralisation at InsCo enabled the creation of scalable processes in a situation in which engaging every employee in formalisation projects was likely impossible, considering the number of employees. Furthermore, to ensure consistent quality, employees’ agency in the formalised KIBS processes was carefully limited and supported. In organisations with less scale-intensive processes, such as at ManCo, the consultants had higher skill levels and all the necessary knowledge regarding the KIBS processes in question. They knew customers’ practical problems and the substance and structure of their services (i.e. synthetic and symbolic knowledge). Thus, they could decide whether and how to formalise their KIBS processes (local autonomy in formalisation projects). More room for agency also existed in the formalised KIBS processes, likely because the high skill level of the employees, the variable nature of customer problems and the possibility of discussing upcoming issues within teams enabled the development of consultants’ well-grounded individual judgments. SoftCo was becoming a more scale intensive service provider, and it was positioned in between ManCo’s and InsCo’s formalisation approaches at the time of the study. 32 Carlborg, P., Kindström, D. and Kowalkowski, D. (2013), "A lean approach for service productivity improvements: synergy or oxymoron?" Managing Service Quality, Vol. 23, pp. 291–304. Consoli, D. and Elche, D. (2013), "The evolving knowledge base of professional service sectors", Journal of Evolutionary Economics, Vol. 23 No. 2, pp. 477–501. Consoli, D. and Elche-Hortelano, D. (2010), "Variety in the knowledge base of knowledge intensive business services", Research Policy, Vol. 39 No. 10, pp. 1303–1310. D’Adderio, L. (2008), "The performativity of routines: theorising the influence of artefacts and distributed agencies on routines dynamics", Research Policy, Vol. 37 No. 5, pp. 769– 789. Edvardsson, B., Gustafsson, A. and Roos, I. (2005), "Service portraits in service research: a critical review", International Journal of Service Industry Management, Vol. 16, pp. 107–121. Edvardsson, B. and Olsson, J. (1996), "Key concepts for new service development", The Service Industries Journal, Vol. 16 No. 2, pp. 140–164. Edvardsson, B., Skålén, P. and Tronvoll, B. (2012), "Service systems as a foundation for resource integration and value co-creation", Review of Marketing Research, Vol. 9, pp. 79–126. Eisenhardt, K. (1989), "Building theories from case study research", The Academy of Management Review, Vol. 14 No. 4, pp. 532–550. Feldman, M.S. and Pentland, B.T. (2003), "Reconceptualizing organizational routines as a source of flexibility and change", Administrative Science Quarterly, Vol. 48 No. 1, pp. 94–118. Fosstenløkken, S.M., Løwendahl, B.R. and Revang, Ø. (2003), "Knowledge development through client interaction: a comparative study", Organization Studies, Vol. 24 No. 6, pp. 859–879. Giannopoulou, E., Gryszkiewicz, L. and Barlatier, P-J. (2014), "Creativity for service innovation: a practice-based perspective", Managing Service Quality: An International Journal, Vol. 24 No. 1, pp. 23–44. Gondo, M.B. and Amis, J.M. (2013), "Variations in practice adoption: the roles of conscious reflection and discourse", Academy of Management Review, Vol. 38 No. 2, pp. 229– 247. Haas, M.R. and Hansen, M.T. (2005), "When using knowledge can hurt performance: the value of organizational capabilities in a management consulting company", Strategic Management Journal, Vol. 26 No. 1, pp. 1–24. Hansen, M. T., Nohria, N. and Tierney, T. (1999), "What’s your strategy for managing knowledge?" Harvard Business Review, Vol. 77 No. 2, pp. 106–116. Hydle, K.M., Aas, T.H. and Breunig, K.J. (2014), "Characteristics of intrapreneurs in scaleintensive service firms", Journal of Entrepreneurship, Management and Innovation, Vol. 10 No. 2, pp. 89–118. Jaakkola, E. (2011), "Unravelling the practices of 'productization' in professional service firms", Scandinavian Journal of Management, Vol. 27 No. 2, pp. 221–230. John, J., Grove, S.J. and Fisk, R.P. (2006), "Improvisation in service performances: lessons from jazz", Managing Service Quality, Vol. 16, pp. 247–268. Kärreman, D., Sveningsson, S. and Alvesson, M. (2003), "The return of the machine bureaucracy? Management control in the work settings of professionals", International Studies of Management and Organization, Vol. 32 No. 2, pp. 70–92. Kelley, S.W., Longfellow, T. and Malehorn, J. (1996), "Organizational determinants of service employees’ exercise of routine, creative, and deviant discretion", Journal of Retailing, Vol. 72 No. 2, pp. 135–157. 33 Lehtonen, M.H., Järvi, K. and Tuominen, T. (2015), "Reflexivity in the productisation of services", International Journal of Work Innovation, Vol. 1 No. 2, pp. 161–184. Lovelock, C. (1991), Services Marketing, Prentice-Hall, Englewood Cliffs, NJ. Maister, C. (1982), "Balancing the professional service firm", Sloan Management Review, Vol. 24 No. 1, pp. 15–29. Martelo-Landroguez, S. and Martin-Ruiz, D. (2016), "Managing knowledge to create customer service value", Journal of Service Theory and Practice, Vol. 26 No. 4, pp. 471–496. Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis: An Expanded Sourcebook (2nd ed.), Sage Publications, Thousand Oaks, CA. Morris, T. (2001), "Asserting property rights: knowledge codification in the professional service firm", Human Relations, Vol. 54 No. 7, pp. 819–838. Morris, T. and Empson, L. (1998), "Organisation and expertise: an exploration of knowledge bases and the management of accounting and consulting firms", Accounting, Organizations and Society, Vol. 23 No. 5–6, pp. 609–624. Ojasalo, J. (2001), "Managing customer expectations in professional services", Managing Service Quality: An International Journal, Vol. 11 No. 3, pp. 200–212. Organ, D.W. and Greene, C.N. (1981), "The effects of formalization on professional involvement: a compensatory process approach", Administrative Science Quarterly, Vol. 26 No. 2, pp. 237–252. Orlikowski, W.J. (2000), "Using technology and constituting structures: a practice lens for studying technology in organizations", Organization Science, Vol. 11, pp. 404–428. Pentland, B.T. and Feldman, M.S. (2008), "Designing routines: on the folly of designing artifacts, while hoping for patterns of action", Information and Organization, Vol. 18 No. 4, pp. 235–250. Pina, K. and Tether, B.S. (2016), "Towards understanding variety in knowledge intensive business services by distinguishing their knowledge bases", Research Policy, Vol. 45, No. 2, pp. 401–413. Quick, K.S. and Feldman, M.S. (2011), "Distinguishing participation and inclusion", Journal of Planning Education and Research, Vol. 31 No. 3, pp. 272–290. Ragin, C.C. (1992), "'Casing' and the process of social inquiry", in Ragin, C.C. and Becker H.S. (Eds.), What is a Case? Exploring the Foundations of Social Inquiry, Cambridge University Press, New York, NY, pp. 217–226. Rahikka, E., Ulkuniemi, P. and Pekkarinen, P. (2011), "Developing the value perception of the business customer through service modularity", Journal of Business and Industrial Marketing, Vol. 26 No. 5, pp. 357–367. Robertson, M. and Swan, J. (2003), "'Control – what control?' Culture and ambiguity within a knowledge intensive firm", Journal of Management Studies, Vol. 40, pp. 831–858. Sandholtz, K.W. (2012), "Making standards stick: a theory of coupled vs. decoupled compliance", Organization Studies, Vol. 33 No. 5–6, pp. 655–679. Skjølsvik, T. (2016), "Business-to-business professional service relationships under multiple logics", The Service Industries Journal, Vol. 36, pp. 163–182. Stake, R.E. (2005), "Qualitative case studies", in Denzin, N.K. and Lincoln, Y.S. (Eds.), Handbook of Qualitative Research (3rd ed.), Sage, London, UK, pp. 443–466. Strambach, S. (2008), "Knowledge-intensive business services (KIBS) as drivers of multilevel knowledge dynamics", International Journal of Services Technology and Management, Vol. 10. No. 2–4, pp. 152–174. Sundbo, J. (1997), "Management of innovation in services", The Service Industries Journal, Vol. 17 No. 3, pp. 432–455. Sundbo, J. (2002), "The service economy: standardisation or customisation?" The Service Industries Journal, Vol. 22, pp. 93–116. 34 Vaast, E. and Levina, N. (2006), "Multiple faces of codification: organizational redesign in an IT organization", Organization Science, Vol. 17 No. 2, pp. 190–201. Valminen, K. and Toivonen, M. (2011), "Seeking efficiency through productisation: a case study of small KIBS participating in a productisation project", The Service Industries Journal, Vol. 32 No. 2, pp. 273–289. Valtakoski, A. and Järvi, K. (2016) "Productization of knowledge-intensive services: enabling knowledge sharing and cross-unit collaboration”, Journal of Service Management, Vol. 27 No. 3, pp. 360–390. Werr, A. and Stjernberg, T. (2003), "Exploring management consulting firms as knowledge systems", Organization Studies, Vol. 24 No. 6, pp. 881–908. Winter, S.G. and Szulanski, G. (2001), "Replication as strategy", Organization Science, Vol. 12 No. 6, pp. 730–743. Wright, C., Sturdy, A. and Wylie, N. (2012), "Management innovation through standardization: consultants as standardizers of organizational practice", Research Policy, Vol. 41 No. 3, pp. 652–662. Yin, R.K. (1994), Case Study Research: Design and Methods (2nd ed.), Sage, Thousand Oaks, CA. -------------------------------- Insert Appendix 1 -------------------------------- 35 Table 1. Empirical research on employee viewpoints on the formalisation of KIBS processes in KIBS firms Source Method and context Aims and findings Scope choices and identified gaps justifying further research Bettiol et al., 2012 Qualitative two-case study: design and communicatio n service firms. Aim: Analyses the relationship between standardisation and creativity in KIBS innovation. Shows that knowledge can be codified effectively in complex, creative service work (i.e. design). These KIBS firms can balance creative outputs with standardised working methods. Studies employees’ viewpoints of firm-level standardisation and creativity in KIBS. Explicates a further need to 1) explore how the changes in the entrepreneurial roles and strategic approaches may affect codification of the working method and 2) to compare firms in different industries to strengthen and generalise the results. Brivot, 2011 Single-case study with multiple methods in a large law firm. Aim: Examines whether centralised knowledge management systems contribute to a shift in power from professionals to managers in a professional service firm. Focuses on how a knowledge management system is used for knowledge codification and reuse, and how knowledge control may facilitate the re-distribution of power in the firm. Reveals that codification is useful when employees remain in control of their work: social control and selfcontrol regulate knowledge use and retain individuals’ independence. Studies employees’ viewpoints of the firm implementing a knowledge management system that changed the work practices. Single-case setting does not enable comparison. Focuses on the formalisation of knowledge management systems. Analyses control, not employee agency. Calls for more research on the bureaucratisation of service work. Haas and Hansen, 2005 Interviews and survey with 191 respondents in a management, audit and tax consulting firm. Explores the possibility that utilising the firm’s knowledge resources to complete important tasks can have negative performance consequences. Tests the link between personal and codified knowledge and winning the bid, along with some moderating variables. Reveals that the usefulness of codification depends on the employees’ level of experience: juniors benefit from it more than seniors. Studies knowledge use statistically from the individuals’ perspective in the front end of customer projects. Focuses on bids (i.e. certain customer assignments) rather than the formalisation approach or task. Points out the importance of situational factors in affecting the value of knowledge resources for the performance of task units within firms. Kärreman et al., 2003 Two-case study with a consulting firm and a researchAim: Intends to reveal that also knowledge intensive firms may become bureaucratised and that such firms require context-sensitive management. Studies standardisation of practices at the level of the firm, through personnel’s experiences. Focuses on big firms. 36 Source Method and context Aims and findings Scope choices and identified gaps justifying further research intensive pharmaceutical firm. Demonstrates that traditional managerial control and standardisation of practices prevails also in knowledge-intensive firms. Bureaucratic mechanisms may help to cope with the inherent ambiguity in knowledge work, but core work processes should not be standardised (“selective bureaucratisation”). Firm size may be related to bureaucratisation, but cultural differences may be more important in specifying the type of standardisation. Identifies a need to understand other types of knowledgeintensive firms and “selective bureaucratisation” better. Morris, 2001 Longitudinal qualitative single-case study in a consulting firm. Aim: Examines whether and how the paradox of knowledge sharing vs. knowledge as a source of professional status is resolved in a professional service firm. Suggests that codification is suitable for certain tasks and employees, but not all. Individuals approach codification and their professional roles in different ways depending on their background and capabilities. Studies consultants’ experiences of the firm’s approaches to codification. Single-case setting does not enable comparison. Emphasises knowledge as property; particularly relevant to professional (complex) services. Difference between simple and complex services pointed out as a factor relevant to the suitability of codification/standardisation. Morris and Empson, 1998 Comparative exploratory two-case study: accounting and consulting firms. Aim: Examines forms of knowledge and knowledge management strategies within professional service firms; develops a framework linking the firm’s knowledge base and organisational structure. Argues that the nature of the knowledge base is central in defining the firms’ knowledge management strategies (codified vs uncodified). A variety of knowledge bases can be leveraged in the same firm. Organisational culture, context and human judgment may affect whether and how codification can be used. Compares knowledge management strategies between two firms, based on the personnel’s experiences. Points out the multiplicity of knowledge bases within the same firm. Expresses a need to 1) compare systematically firms with different knowledge bases and strategies and 2) control for other background factors such as size, market environment, technology and age. Rahikka et al., 2011 Single-case study in a professional engineering and construction service firm. Aim: Examines the influence of modular services on customers’ value perception in professional services. Suggests that service modularity is linked to trust and customers’ value perceptions. Shows that different aspects of modularity (processes, organisation, service modules) are Analyses a firm’s approaches to modularisation, both through employees’ and customers’ perceptions. Single-case setting does not enable comparison. Focus on customers’ view and service offering – employees’ experiences of service work or process not covered. 37 Source Method and context Aims and findings Scope choices and identified gaps justifying further research linked to different customer value outcomes. Emphasises the customers’ role in specifying whether modularisation is possible and whether employees can consider alternatives in the service event. Skjølsvik, 2016 Interview study with 51 buyers of consulting services in 16 large organisations and 27 consultants. Aim: Explores the contradictory logics of formalisation (goods logic) vs. service logic in professional service relationships. Finds that formalisation appears differently at the firm level and at the individual level, and differently for different customers. The need to balance the contradicting logics at both the individual and firm levels defines the nature of the relationship. Focuses on individuals’ viewpoints to goods vs. service logics, both from the buyers’ and service providers’ perspective in the service relationship. Focuses on a certain type of service (consulting) – identifies a need to 1) understand other types of services, 2) study different personal service relationships (instant vs long) and 3) study multiple logics and co-produced service value. 38 39 Table 2. The studied firms ManCo SoftCo InsCo Industry Management consulting Software and consulting Financial and insurance services Nature of business Professional KIBS services Professional and technology-based KIBS services and software products B2B service products and supplementary professional KIBS services (focus of the study) Type of knowledge base Synthetic with symbolic knowledge elements Synthetic but software development involves analytic elements Synthetic but service development involves analytic features KIBS processes before the studied formalisation projects Small-scale, customised processes Mediumand smallscale processes including both formalised and customised elements Scale-intensive, formalised processes with some customised elements Nr of personnel (category) <100 <100 <5000 Turnover MEUR (category) <10 <10 <2000 Table 3. Summary of the data ManCo SoftCo InsCo Interviewees’ positions 10 consultants, including managing partners 2 top managers 1 top manager 3 unit managers 2 unit manager 2 account managers 3 development managers 1 marketing manager 1 chief underwriter 2 team leaders 4 business/sales developers 1 developer 2 sales team leaders Primary interviews 9 9 11 Follow-up interviews 6 6 5 2 group interviews, 5 persons interviewed 2–3 times 4 persons interviewed twice 2 persons interviewed 2–3 times Secondary data Documentation, observation in 2 workshops Documentation, observation in 3 workshops Documentation, observation in 2 workshops 40 Table 4. Cases: Six formalisation projects Case Type of KIBS process Reason for formalisation Formalisation actions Perceived effects on KIBS processes InsCo C1 (Case 1): Formalisation of cross-organisational processes related to B2B packages. Consultancy as part of B2B insurance packages. Scale and technology intensive. Most processes were formalised but subsidiary firms operated independently. Products difficult for customers to comprehend. Synergies between firms and processes were missing. Cross-organisational, masscustomised packages were defined with modular structures and process templates. Individual deviations were not allowed. Sales were increased, personnel were proud of new offerings. Cross-organisational knowledge increased. Customer satisfaction depended on salespersons’ skills. InsCo C2: Formalisation of micro-level sales interactions related to B2B packages. Scaleand technology-intensive processes. From the customers’ perspective, coordination between services was lacking. Selling modularised offerings required ensuring and unifying employees’ sales competences. Sales process templates were defined. No deviations were allowed. Customer centricity increased: customer knowledge was prioritised over technical knowledge. Variations remained as some salespersons used their old tactics. SoftCo C1: Formalisation of consultancy services and processes related to software products. Technology-intensive but customised processes with medium-scale intensity: contents depended on individuals’ skills. Dependence on ‘heroic’ consultants problematic due to growth. Project sizes and modules were standardised. Templates for sales and delivery processes were created. Variations in customer cases were allowed to some extent. Sales, profitability and project sizes increased. Consultants were socialised more rapidly. Informal knowledge sharing was easier. But some customers disliked standardised contents. SoftCo C2: Development and formalisation of new consultancy services and processes. Technology-intensive, smallscale processes in new and complex service area, where a small team developed the service through experimentation. Service was too ambiguous for the customers and salespersons. Sales processes were too resource intensive. Sales and delivery process templates were created, but details were not specified. Some modules were codified, others evolved through experimentation. Delivery was easier, even though KIBS processes were still evolving and varied across customer cases. Sales material was improved but could be made clearer. 41 ManCo C1: Formalisation of key consultancy services and processes within teams. Small-scale, customised processes: consultants sold their personal expertise. Some standardised methods were shared among consultants. Need to improve knowledge transfer, quality and predictability in sales. Desire to sell collective projects without compromising individual autonomy. A number of services and KIBS process templates were defined, including names, value propositions, pricing models and tools/contents. Individual variations and deviations were allowed. Collective projects increased team spirit. Sales were increased and delivery was easier. Consultants used process templates depending on personal desires and customers’ needs. ManCo C2: Rigorous formalisation of team’s consultancy offerings and processes. Small-scale, customised processes: consultants sold their personal expertise. Some standardised methods were shared among consultants. Desire to create reliable team-level offerings and to minimise members’ entrepreneurial risk. Team contracts, KIBS process templates and competence requirements rigidly defined. No individual variation allowed. Sales and profitability increased alongside larger projects with ensured quality. Safer income for members. Common language with customers. However, some members changed teams due to inability to follow rules.