194Avadikyan&Lhuillery&Negassi(2016).pages M@n@gement 2016, vol. 19(4): 277-304 Technological innovation, organizational change, and product-related services 
 Arman Avadikyan � Stéphane Lhuillery � Syoum Negassi Abstract. The literature regarding the determinants of servitization emphasizes the role of organizational change and usually overlooks the role of technological change. Using an original sample of 1,129 German manufacturing firms, we reverse the hierarchy: product novelty is a main driver of product-related service (PRS) activities. It especially boosts consulting and training services. The structure of the PRS portfolio is dependent on product novelty. Organizational changes toward a more flexible company or the adoption of new advanced manufacturing processes are found, with few exceptions, hardly to influence the decision to offer a product-related service. Our results suggest however, that process innovation is positively linked to the breadth of service surrounding products, whereas organizational innovation is more prone to lead to a larger breadth of services surrounding customer offerings. Product, process, and organizational innovation are not found to be complementary drivers of product-related service offerings. Keywords: technology, innovation, organization, servitization, product- related services. INTRODUCTION When exploring the evolution of the modern manufacturing firm, scholars insist on the increasing bundling of products and services, which is variously referred to as “servitization” (Vandermerwe & Rada, 1988), “ s e r v i c i z a t i o n ” ( Q u i n n , D o o r l e y & P a q u e t t e , 1 9 9 0 ) , “ g o i n g downstream” (Wise & Baumgartner, 1999), “transition to services” (Oliva & Kallenberg, 2003), “integrated solutions” (Davies, 2004), “hybrid offerings” (Ulaga & Reinartz, 2011), or “product-service systems” (Mont, 2002). Despite some steps back (Finne, Brax & Holmström, 2013) and some reluctance (Gebauer & Fleisch, 2007), there is evidence that a large and ever-increasing proportion of manufacturing firms already adopt such servitization strategies (Cohen, Agrawal & Agrawal, 2006; Fang, Palmatier & Steenkamp, 2008; Neely, 2009; Rosen, 1998; Tuli, Kohli & Bharadwaj, 2007). Based on a large international database, Neely (2009) found that the proportion of manufacturing firms reporting to provide services is now at least 30%. The share is approximately 58% in the United States, 29% in Germany, 25% in the UK, and 11% in Japan. �277 Arman Avadikyan BETA (UMR-CNRS 7522)—Université de Strasbourg avady@unistra.fr Stéphane Lhuillery ICN Business School and BETA (UMR- CNRS 7522)—Université de Lorraine stephane.lhuillery@icn-groupe.fr Syoum Negassi Management Science Department— University of Paris 1 Panthéon- Sorbonne syoum.Negassi@univ-paris1.fr mailto:avady@unistra.fr mailto:stephane.lhuillery@icn-groupe.fr mailto:syoum.Negassi@univ-paris1.fr mailto:avady@unistra.fr mailto:stephane.lhuillery@icn-groupe.fr mailto:syoum.Negassi@univ-paris1.fr M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi The servitization process is expected to provide competitive advantage to firms: Servitized firms are more likely to compensate for the loss of revenues arising from stagnant demand or fluctuating business cycles and to initiate new growth opportunities and more stable revenues (Cohen, et al., 2006; Cusumano, Kahl & Suarez, 2015; Sawhney, 2004). Services are less capital-intensive to produce than products, and servitized firms tend to have higher margins (Wise & Baumgartner, 1999). Services are a critical differentiation factor in a firm’s strategy mix and a strategic tool to develop clients’ loyalty (Mathe & Shapiro, 1993) and to establish barriers to entry. However, empirical studies have provided scattered and contradictory findings regarding the impact of servitization: Whereas some authors have found direct positive effects (e.g., Fang, et al., 2008; Tether & Bascavusoglu-Moreau, 2012; Visnjic, Neely & Wiengarten, 2012; Visnjic & Van Looy, 2011), others have not (Robinson & Chiang, 2002; Visnjic, et al., 2012). Neely (2009) even found mixed results: Servitization has a positive link on profitability, whereas the extent of servitization has a negative effect on profitability. Findings also highlight the non-linear effects of servitization on performance and/or the differences among industries (Fang, et al., 2008; Suarez, Cusumano & Kahl, 2015; Visnjic & Van Looy, 2013). Several explanations can be put forward to explain the “servitization paradox.” A first possible explanation underlined by Cusumano, Kahl, & Suarez, (2015) is that the results regarding servitization, with very few exceptions, rely on poor data: mono-sectorial data (Eggert, Hogreve, Ulaga & Muenkhoff, 2011, 2014; Suarez, et al., 2013), small samples (Antioco, Moenaert, Lingreen & Wetzels, 2008; Fang, et al., 2008), or biased samples (on listed companies, for example, Fang, et al., 2008; Neely, 2008; Visnjic, et al., 2016) combined with the usual problems related to time lags and mismeasurement (Brynjolfsson, 1993). A second related point is that studies do not properly take into account the heterogeneity of services involved and that different services do not impact performances in the same way. For example, some services can be offered to diversify activities when usual markets and performances decline, whereas other services can be offered to support products or processes introduced in growing markets. Along with Mathieu (2001), a distinction can also be made between services supporting products (SSPs, such as installing, documenting, and repairing) and services supporting client actions (SSCs, such as financing, training, and consulting) that can provide different competitive advantages to manufacturing firms (Eggert, et al., 2011). A third strand of the literature emphasizes that human resources (HR) or organizational capabilities are not properly taken into consideration: There is a lack of managerial motivation and competency (Gebauer, Fleisch & Friedli, 2005), top-manager commitment (Antioco, et al., 2008; Gebauer, Pütz, Fischer & Fleisch, 2009), training activities (Santamaria, Nieto & Miles, 2012), and service culture and mind-set (Neely, 2009). Other authors emphasize the absence of adequate organizational arrangements to enable firms to derive value from services (Bowen, Siehl & Schneider, 1989; Davies, Brady & Hobday, 2006; Gebauer, et al., 2009; Neu & Brown, 2005; Oliva & Kallenberg, 2003) or a lack of slack resources (Fang, et al., 2008). We contend that the main caveat likely to explain conflicting results is instead that previous studies have overlooked technology. Often, the literature focuses on product quality, in which the technological or other dimensions of novelty are not explicit (Ulaga & Eggert, 2006). Thirty years ago in a seminal paper, Teece (1986) emphasized that PRSs can be a critical complementary asset that allows firms to exploit innovation. � 278 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 Whereas many contributions have explored the role of services on performances, very few have considered technology as a critical driver of servitization that is likely to influence performance. In the following, we contend that servitization cannot be analyzed without considering the sparking role of technological innovation. To reinstall the role of technology into the analysis of servitization, the present article addresses three main research questions: Is technological change a driver of PRS strategies? What is the link between technological change and organizational change as a PRS driver? To what extent do these determinants depend on the type of PRSs considered? To answer these questions, we propose an econometric model that explains PRS offerings in terms of the strategic orientations—including product innovation, process innovation, and organizational change— chosen by 1,129 German manufacturing firms. Radical product innovation is found to be the main driver of PRSs, whereas the positive influence of organizational change and process innovation are reduced. New products and new organizations are not found to be complementary determinants of PRSs. A further contribution is that technology is demonstrated to influence diverse PRSs in different manners, shaping PRS portfolios. Our analysis also challenges the standard classifications of PRSs proposed by scholars. Despite performing a cross-sectional analysis, our article offers a much more precise view than previous work in the field and addresses several gaps: It uses a large-scale sample, including multiple industries and firm sizes, and it focuses on PRSs, whereas many articles on servitization have mixed product-related services with product-unrelated services. The reminder of the paper is organized as follows. We first propose an integrative perspective regarding the determinants of PRSs in manufacturing. Exploring the literature, we particularly rejuvenate the importance of technology in addition to organizational aspects for PRS strategies. We first present the role of process/product innovations in servitization and then discuss the role of organizational practices. We then consider the hierarchy and heterogeneity of influences driving PRS offerings. Then, we present the method that we implement to test our empirical hypotheses, including the data, the different variables, and the econometric modeling. The next section presents our statistical results regarding the determinants of PRS portfolios. A final section concludes and discusses the main contributions and implications of our results. We then address limitations and future research. THEORY AND HYPOTHESES TECHNOLOGY AND PRSS Technology expands the opportunities for firms to offer not only new products but also new processes. New products may create opportunities to expand the importance and scope of existing PRSs; new products may also create opportunities for novel PRSs. In a similar manner, new process technologies may develop or redefine the manner in which products and PRSs are produced and delivered. We review the evidence regarding product and process innovation in turn. An interesting feature of the literature about servitization is that servitization is considered to be irruption of customer orientation in manufacturing organizations: Servitization is considered market- and customer-driven. Customers’ willingness to pay for a product will be particularly increased by PRSs. A particular case is when the product is innovative, either in a radical or incremental manner. In this case, PRSs � 279 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi can be considered a role model of complementary assets likely to enable firms to exploit technological innovation (Teece, 1986). Still, the link between the degree of novelty and PRSs is not clear in the literature. On one hand, PRSs are a substitute for product innovation. Oliva & Kallenberg (2003) consider that services are proposed when related to an installed base of manufacturing products to expand product life cycles and to compensate for a lack of improvement. In this setting, PRSs can be considered a substitute to address a lack of innovation capability, and radical innovation has a crowding-out effect on PRS offerings. The point can be made compatible with incremental innovation: At the end of the life cycle, incumbents introduce incremental innovation to maintain the value of their products. In this case, PRSs surround incremental innovation but are still a substitute for radical innovation. PRSs can furthermore be considered as a means for incumbents, especially in mature industries, to create switching costs for their customers. The switching costs induced by PRSs are less critical for recent and radically new products. A final substitution possibility is also that the introduction of a radically new product shrinks the market for services: A product with radically new auto- diagnostic and auto-repair properties, for example, would dissuade firms from proposing maintenance and repair activities. A new machine with radically new user-friendly controls would also dissuade firms from offering training sessions to their customers. On the other hand, the literature about servitization often implicitly or explicitly considers technology as a threat to customers likely to face products that they do not need or do not understand. For instance, Bowen, et al. (1989) asserted that customer services should be a requisite consequence when manufacturing firms differentiate their products from others through innovative technologies and designs. Users may be hesitant to adopt innovative technologies because of uncertainties about product performance and quality. This phenomenon is in agreement with the marketing literature, in which extensive PRSs may be required to overcome barriers to the adoption of innovative products by customers (Araujo & Spring, 2006). PRSs can therefore be critical, especially for radical innovations, as their novelty is difficult for users to comprehend, and firms must help customers learn about and appropriately perform new product functionalities. The complementarity view is also supported by the fact that PRSs can be an interesting means of appropriation implemented by pioneers to prevent imitation (Levin, et al., 1987). The empirical literature does not significantly help clarify the two views. The influence of product innovation on related services, including the degree of novelty of technological innovation, has been barely taken into consideration in empirical studies (Fang, et al., 2008; Neely, 2009; Tether & Bascavusoglu-Moreau, 2012; Visnjic & Van Looy, 2011; Visnjic, et al., 2012). Some empirical insights have come from expert interviews: Matthyssens & Vandenbempt (1998) confirmed, for instance, a strong and positive relationship between leadership in product innovation and market leadership in industrial services. Econometric evidence is, however, still rare and vague: Some scholars have found empirical evidence that firms with low innovative capabilities may counterbalance their weakness through higher complementary assets (Rammer, Czarnitzki & Spielkamp, 2009), and PRSs may be a commonly used asset in this respect. However, Eggert, et al. (2011) found indirect evidence for a positive impact of product novelty on PRSs. More precise is the positive and direct link found by Santamaria, et al. (2012), where R&D intense firms are more likely to innovate in services. Visnjic, et al. (2014) found a positive and complementary role in terms of long-term performance for the adoption of � 280 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 a service-oriented business model and technological innovation approximated by R&D intensity. In the two previous references, product- related services and product-unrelated services were not differentiated, but the former can be positively linked to technological innovation and the latter negatively linked to technological innovation. Based on these different works, we propose a first general hypothesis related to product innovation: Hypothesis 1: Firms that introduce product innovations, especially pioneer products, are more likely to offer PRSs. Because servitization is a general downstream orientation in which integration of customer and service units is performed, scholars have explored the role of adopted technologies in the servitization process. ICTs and advanced manufacturing technologies (AMTs) have been the main technologies investigated. Contrary to product innovation, scholars have been able to incorporate new technological processes as a means to improve servitization: For instance, ICTs were first documented as an enabler of servitization (Barras, 1986; Kellogg & Nie, 1995; Lightfoot, Baines & Smart, 2011; Windahl, Andersson, Berggren & Nehler, 2004). Information technologies and techniques help firms identify and adapt to customers’ characteristics and needs (Froehle, 2006; Nonaka & Teece, 2001) and to transform and exploit external knowledge (Joshi, Chi, Datta & Han, 2010). These technologies and techniques thus help firms define new business opportunities and differentiate their PRS offerings. ICTs reduce the delivery costs of services: By lowering internal information, coordination, production, and transaction costs, they help transform potential services into profitable ones. ICTs are also interesting because they allow new PRSs. For example, car rental companies can propose navigational systems. ICTs can also be intertwined with complementary technologies (using sensors for example) to produce or deliver new services. Customers can also track online, step-by-step, the processing and delivery of their products. The strong linkages between firms and clients increase switching costs and develop future product and service opportunities. ICTs are, however, only one part of the technology used in manufacturing companies. They are often combined with AMTs—defined as a group of hardware- and software-based technologies designed to improve the efficiency and effectiveness of firms in the design, manufacturing, and testing of products—that can lead to more flexible product design, combining manufacturing efficiency and customer differentiation (Hofmann & Orr, 2005; Lei, Hitt & Goldhar, 1996) . AMT 1 resources and competencies may allow customization of product functionalities, quality, or reliability with the possible fast delivery of small batches. Integrating computers and networks with databases, firms with recent AMTs are more likely to encourage their customers to interact directly on specifications, production, and delivery, thereby allowing for the synchronization of upstream CAD (computer-aided design) and CAE (computer-aided engineering) with downstream CAM (computer-aided manufacturing) and PDM (product data management) activities. AMTs thus create a manufacturing system environment more favorable for PRSs, thereby enabling higher-value-added and extended scope of PRSs to be offered to customers. By integrating and standardizing the management principles of services, implementing AMTs in the back office may support and create new opportunities for the effective delivery of services in the � 281 1. The point is very similar for the case of flexible manufacturing systems (Buzaco= & Yao, 1986). M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi front office. Empirical work provides many examples of service orientation and high customization in manufacturing firms that are significantly driven by not only ICTs but also remote-monitoring technologies (e.g., Antioco, et al., 2008; Barabba, et al., 2002; Davies, et al., 2006; Grubic, 2014). Mixing product and process dimensions, some technologies such as cloud computing and/or digitalization have even been demonstrated to disrupt entire industries and create entirely new service opportunities (Fichman, Dos Santos & Ze, 2014 and references therein; Sultan, 2013). Indirect evidence is provided by labor studies that demonstrate that firms with AMTs are employing skilled labor (Doms, Dunne & Troske, 1997), especially in service-related industries (Kaiser, 2000), thereby suggesting that firms that adopt AMTs are more likely to identify and propose PRSs to their customers. Specific evidence linking AMT innovations with services is, however, still scarce: Santamaria, et al. (2012) proposed the main empirical analysis, based on a large sample of Spanish firms. Their results support the general hypothesis that firms with advanced manufacturing processes are the ones that are more service-oriented. Still, whereas ICTs can be considered a general source of servitization, AMTs are more related to manufacturing products and thus should be particularly relevant to achieve a subset of PRS strategies. Based on the reviewed literature, we pose a second hypothesis regarding technology: Hypothesis 2: Firms that introduce new advanced manufacturing technologies are more likely to offer PRSs. FLEXIBLE ORGANIZATION AND PRSS In the literature about servitization, firms establish appropriate organizational arrangements to manage the service production process, in which customers’ needs and market changes are key (Bowen, et al., 1989; Davies, 2004; Galbraith, 2002; Gebauer, et al., 2005, 2009; Mathieu, 2001; Nambisan, 2001; Neu & Brown, 2005; Oliva & Kallenberg, 2003). The literature has first explored the possibility of achieving successful services through organizational design and the creation of a separate service unit, such as a service function or a service division (Gebauer, et al., 2005, 2009; Oliva & Kallenberg, 2003; Oliva, et al., 2012). The authors usually contend that a dedicated service unit is the only means to develop service businesses inside manufacturing firms (see also O’Reilly & Tushman, 2004). The creation of a separate function allows some autonomy, some specialization effects, and a specific attention to clients. However, scholars have stressed that this type of organizational arrangement is not a sufficient condition for achieving effective servitization: It may validate the co-existence of product and services within the manufacturing firm, but it does not guarantee that services are customer-centered, and it may hamper the transversal development of specific PRSs that would satisfy the clients’ need (e.g., Gebauer, et al., 2009; Miller, Hope, Eisenstat, Foote & Galbraith, 2002). Scholars consider two dimensions of enabling adaptation and subsequent efficient service activities: decentralization and integration. To be able to propose services, including PRSs, manufacturing firms should first achieve decentralization of their units (Acemoglu, Aghio, Lelarge, Van Reenen & Zilibotti, 2007; Gebauer, et al., 2005; Neu & Brown, 2008), and delegate decisions and control to lower hierarchical levels. Decentralization allows firms to more efficiently manage the use of information dispersed among employees and thus allow specialists to design and efficiently � 282 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 manage customer-centric activities (Galbraith, 2002) . A problem is that 2 the autonomy derived from decentralization may damage interactions with other intra-firm units and competencies. Integration is thus considered a second organizational trait required by firms aiming at servitization. Firms should integrate the different units through different practices, such as standardization, formalization, information systems, or cross-functional teams (Galbraith, 2002; Mathe & Shapiro, 1993). Implementation of these costly practices is considered effective because it lowers communication and coordination costs among units through better and faster responsiveness between the service unit(s) and the other functional units or customers (Bowen, et al., 1989; Turunen & Toivonen, 2011). Among these integration practices, teams seem to be considered a main solution to rapidly develop customer-specific services (see Galbraith, 2002). Teams are indeed particularly interesting because they are problem-solving entities, able to capitalize knowledge (Lenfle & Midler, 2009) and may even boost creativity, especially when complex products are provided (Boning, Ichniowski & Shaw, 2007). Both dimensions have received little systematic empirical support. Regarding decentralization, the literature implicitly considers that a better information system is a means to decentralize organizations. The already- cited results regarding the role of ICTs in servitization should be considered, in this perspective, as clues supporting the positive role of decentralization. Bowen, et al. (1989) highlighted that implementation of control mechanisms is indeed very costly because of the rapidly changing and uncertain idiosyncratic services managed by employees. In a case study, Neu & Brown (2008) confirmed that servitization is enabled by senior management through delegation of decision-making authority to mid-level managers. Until now, little empirical evidence regarding the role of decentralization has been available. To our knowledge, Eggert, et al. (2014) is the only econometric contribution based on a large sample. The authors demonstrate that decentralized manufacturing firms are able to derive significantly higher growth revenue and profit from SSCs. Regarding integration, the literature is also still largely based on anecdotal evidence or case studies. Using a sample of 107 international firms, Antioco, et al. (2008) made a significant effort to test the link between integration and service orientation. However, the authors failed to find any significant impact of cross-functional communication on the service orientation of manufacturing firms. Taking stock of these different organizational traits that a manufacturing firm should adopt to succeed in services leads us to define an organization as “organic” or “flexible” when it is characterized by few levels of hierarchy, strong decision-making power at the employee level, and strong communication among the different organizational units to better coordinate activities and employees at lower costs (Bahrami, 1996; Bowen, et al., 1989; Burns & Stalker, 1961; Miller, 1986; Youngdahl, 1996). Such decentralization and integration are expected to enable organizations to satisfy customer needs effectively, in quality, in quantity, in design, and in a timely manner. Flexible firms have better interactions with clients and other stakeholders that are also likely to provide innovative services (Plé, Lecocq & Angot, 2010; Saccani, Visintin & Rapaccini, 2014; Von Hippel, 1986). We can propose a third empirical hypothesis: Hypothesis 3: Firms that adopt a more flexible organization are more prone to propose PRSs. � 283 2. This posiFve effect is well documented for products with econometric results on large samples (e.g. Auh & Menguc, 2007). M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi DOMINANCE, COMPLEMENTARITY AND HETEROGENEITY The organizational and technological determinants of servitization are rarely compared. Polar views can be inferred from the literature: Three are worth mentioning. First, studies usually focus on the market orientation of organizational arrangements, whereas the role of technology is overlooked or considered. Scholars working on servitization would rather consider that organizational change is the dominant driver of servitization: PRSs can be based on already existing products, but some organizational arrangements are required to be able to deliver PRSs (Oliva & Kallenberg, 2003). In this view, new organizational practices are required, but their interactions with new technologies are not considered; technology is not neutral but is a context, a determinant of servitization where the interactions with organizational dimensions are not specified (e.g. Baines & Lightfoot, 2013; Baines, et al., 2016). In a second view, some scholars consider that PRSs are arranged around incremental innovations. Stable core competencies can in this case be a source of rigidities that hamper the adaptation of firms (Leonard- Barton, 1992) and thus prevent the delivery of PRSs. In this view, incremental innovators are less able to adapt and to propose new PRSs. The solution to overcome this difficulty is to complement the incremental innovations with organizational change. The combination of organizational change and incremental innovation should thus be positive for PRSs even if the sole impact of incremental innovation is likely to impact PRSs negatively. From this perspective, the difficulties may be alleviated when radical innovation is introduced, suggesting that complementary organizational change is not required anymore. A pattern of radical innovation combined with organizational change should thus have a relatively lower impact on PRSs than for incremental innovation. Third view considers that firms can build dynamic capabilities (Teece, et al., 1997) to be able to stay innovative over time, either incrementally or radically. Services are not related to an installed base but are solutions, integrated with new competencies and products allowed by organizational change (Davies, 2004). Product and process innovations are articulated with organizational change and even AMTs to propel PRSs. Innovation and organizational change should then be observed as complementary. The degree of novelty of product innovation should be positively correlated with the degree of novelty of organizational practices. In this setting, organizational change is to be deployed either for radical innovation or for incremental innovation to be able to offer PRSs. The combination of organizational change and technological change is always expected to be positively related to PRSs. In the face of such a lack of clear predictions regarding the relative importance of the different PRS determinants, empirical studies should help. In Santamaria, et al. (2012), the impact of new processes and the link with customers (decentralization) were found to be positively linked to service innovation, but the difference between the two determinants was not tested. In Antioco, et al. (2008), the modeling adopted hampered clear conclusions regarding the magnitude of the positive effects. Their results suggested that technology is dominant over organization. There is, however, a lack of studies that allow us to generalize this result. A further problem is that to our knowledge, there is no result available regarding the complementarity issues. Facing a lack of strong insights from the literature, we adopt the two following hypotheses: � 284 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 Hypothesis 4a: Technological innovation is a dominant driver of PRS offerings. Hypothesis 4b: Technological innovation and organizational change are complementary drivers of PRSs. Some authors offered more subtle and more complete analyses in which the role of the determinants of PRSs changes according to the type of PRS. Servicing profiles may differ depending on whether the role of services is to support the introduction/development of new and innovative products to the market or to complement existing products and support the already installed product base. Cusumano, Kahl, & Suarez (2015) recently proposed an interesting categorization of PRSs in which some services are oriented toward incremental innovations or to mature industries whereas others address radical innovations. The former are “smoothing” services (such as financing, repairing, training, installing, and documenting) that enable firms to maintain their margins, whereas the latter are “adapting” services (such as software and consulting) that accompany radically new products to create a margin. The degree of technological novelty will thus support different types of PRSs. However, scholars often use the older SSP and SSC classification (see Mathieu, 2001) to argue that market pioneers should first propose SSPs and then afterward propose SSCs in their PRS portfolio once the product novelty vanishes with its related SSP opportunities. According to Eggert, et al. (2011, 2014), the heterogeneity is due to an arbitrage among competing resources: Through learning, radical product innovators benefit from comparative advantages in specific resources that enable them, at low cost, to complement their product with SSPs. Laggards are not able to develop such competitive advantages but have relatively more resources left to invest in SSCs. Despite its reliance on anecdotal evidence, this type of co-evolution has been widely accepted by scholars. Based on a large sample of 414 German mechanical engineering firms, Eggert, et al. (2011) made a breakthrough contribution that demonstrated not only a positive and higher impact of SSPs for pioneer firms (with novel products) but also a higher impact of SSCs for laggards (products that are not new to the market but new to the firm) on profit growth. The heterogeneity in organizational change should be consistent with the heterogeneity in product innovation. The integration and responsiveness allowed by flexible organizations particularly encompass a better ability to interact with external bodies such as suppliers and customers. An interpretation derived is that flexibility should thus boost the opportunities to propose customer-oriented services (SSCs) and to develop close customer–supplier relationships that are less critical for developing product-oriented services (SSPs). Once again, the empirical results are not very helpful: Antioco, et al. (2008) do not find any differences in the roles of integration practices on PRSs, and cross- functional communication is found to be influential neither on SSCs nor on SSPs. In Eggert, et al. (2014), a positive role of decentralization is undeniably found for SSCs, whereas the role of decentralization is not even explored for SSPs. Finally, we can consider that AMTs, including ICTs, are likely to help firms improve their manufacturing capabilities and to integrate them, plan them, and control them efficiently. Dedrick, Gurbaxani & Kraemer (2003) confirmed in a review of the empirical literature that IT enables fundamental changes in organizational structures. AMTs should, for instance, impact warehouse management, delivery, installation, or repair aspects, whereas financing or consulting activities should be enabled and influenced by cheaper and deeper communication with customers � 285 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi regarding production processes and costs. However, it seems more difficult to determine whether AMTs are more prone to influence SSPs than SSCs. Antioco, et al.’s empirical results (2008) corroborated the broad influence of ICTs with an impact on both SSP and SSC differentiation. According to theoretical and empirical insights, we propose the following minimal hypothesis: Hypothesis 5: The impact of technological and organizational change depends on the type of PRS. METHOD DATA: THE EUROPEAN MANUFACTURING SURVEY (EMS) Every two years since 1993, the Fraunhofer Institute for Systems and Innovation Research (ISI) has monitored technological and organizational changes in the German manufacturing industry using a written survey (“German Manufacturing Survey”) mailed to manufacturing companies. The survey addresses diffusion patterns of new organizational, managerial, and technological changes in addition to cooperation, relocation, and performance issues (Kirner, Kinkel & Jaeger, 2009; Lay, Copani, Jager & Biege, 2010). Since 2003, the survey has been extended to other European countries in the form of the EMS. The second EMS conducted during 2006, and used here, surveyed 13,426 manufacturing firms from Germany with more than 19 employees that belonged to various manufacturing sectors (from 15 to 37 in the NACE rev.1 classification). 3 1,663 firms responded (response rate 12.4%). Some cleaning combined with the effect of missing values reduced the number of observations. The questionnaire uses specific questions regarding the introduction of organizational changes or technological processes performed over time. The organizational changes and product and process innovations can thus be identified over the 2003–2005 period. PRS activities are identified in 2006. Once missing values were taken into account and firms younger than four years were removed, our final sample included 1,129 firms. VARIABLES We first define the explained variables. Explanatory variables are then defined in turn. Table 1 reports the different variables and their definition. � 286 3. NACE stands for « Nomenclature staFsFque des AcFvités économiques dans la Communauté Européenne ». Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 Table 1: Definition of variables We define four different explained variables. First, we identify whether a firm proposed at least one PRS in 2006 (PRSYes is set to 1; see Table 1). A second set of variables provides more details: It is a set of seven dichotomic variables measuring the PRS offerings proposed in 2006, of the seven different PRSs listed in the questionnaire—consulting (design, consulting, and project planning), technical documentation, software development, financial services (leasing, rental, and funding), installation, training, and repairing (maintenance and repair ) . We thus 4 want to explain what types of PRSs are related to innovation and organization. To be able to further characterize the differences among PRS portfolios, we introduced the standard difference between SSCs and SSPs: The SSC index is computed as the sum of consulting, financing, and training services, whereas the SSP index is defined as the sum of documenting, software, installing, and repairing services. In fact, these indexes measure the breadth of SSPs and SSCs. A larger breadth of PRSs measures the offering of a different type of PRS but not a novel PRS. Note also that the quality of the answers regarding PRSs hampered the possibility to directly define variables regarding new PRSs without losing the main part of our sample because of missing values. � 287 4. The quesFon used was: “Which of the following services related to your products do you offer to your customers?” The “exploitaFon” service is not maintained in the following because too few firms declared that they provided producFon at the locaFon of their client. M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi Table 2: Descriptive statistics Regarding the independent variables, we define product innovators using a standard approach (see OECD, 2005) relative to the 2003–2005 period. We define two levels of product innovators: Among product innovators, we differentiate between those who introduced radical products —that is to say a product that was new to the market (PRODUCTRADIC) — a n d t h o s e w h o p r o d u c e d o n l y i n c r e m e n t a l i n n o v a t i o n s (PRODUCTINCRE) for comparison with non-innovative-product firms. Thanks to the EMS questionnaire, we further identify different process innovations: With regard to processes, the EMS survey lists 11 different types of process innovations (see Table 1) : The PROCESS index can 5 thus be computed as the sum of the new advanced manufacturing technologies implemented by firms over the 2003-2005 period . Finally, organizational traits related to PRSs are mainly characterized by integration and decentralization practices. A firm is thus considered more flexible (FLEXIBLE) when, over the 2003–2005 period , it introduced new 6 integration or new decentralization practices (see Table 1). Five control variables are considered. The EMS questionnaire enabled us to further characterize firms according to the share of their employees with a third-level educational qualification in 2005 (SKILLED). Also, subcontractors (SUBCONTRACTOR=1) are identified as well as the � 288 5. Actually, three items were deleted because they were highly specific to only a few manufacturing industries: “Clean Rooms,” “Biotech processes,” and “CatalyFc conversion.” 6. A transformaFon into a dichotomic variable was made but did not change our results. Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 age of firms. Young firms (YOUNG=1) are firms founded after 1995 (set to 0 otherwise). The size and the sectorial effects are measured, respectively, by the logarithm of the number of employees (SIZE). At the sectorial level, a set of four industry dummies (Low-Tech, Medium-Low-Tech, Medium- High-Tech, and High-Tech) defined at the two-digit level of the NACE classification and measuring the sectorial level of R&D (OECD, 2002) were also computed. These sectorial dummies approximate the degree of maturity of an industry: Low-Tech and Medium-Low-Tech firms approximate mature industries, whereas High-Tech and Medium-High-Tech industries approximate young industries. ECONOMETRIC MODEL Our model tests the determinants of PRS offerings. Any manufacturing firm can, of course, introduce PRSs. We contend, however, that firms anticipate difficulties: Those that are able to innovate and to use flexible arrangements are the ones that are more likely to derive value from PRSs. Hence, they are more likely to launch and/or maintain PRSs, and thus to subsequently offer PRSs. Adding control variables, our core model explaining PRSs is : where PRS is a set of j variables, PRS = {PRSYes, Consult, Doc, Soft, Finance, Install, Training, Repairing, SSP, SSC}; Z is a set of control variables that includes graduated, young, subcontractor, size, and industry dummies; and ei are random errors. Positive outcomes are expected because radical product innovators are also anticipated to be more prone to develop and maintain PRS offerings than incremental product innovators � . Beyond product innovation, a positive effect for process innovation: � or organizational change. � is also expected. Finally, the results regarding technology are to be compared with the magnitude of the positive effect expected for organizational change toward a flexible firm. Either technology is dominant over organization (H4), with � , or � and � as suggested by the literature on innovation, or organizational change is the dominant driver of servitization and � or � or � as suggested by the literature on servitization. To be able to compare magnitudes, marginal effects are computed for non-linear models. Additional Wald tests are conducted. Comparing the coefficients between equations for the same explanatory variable, we are able to test H5. For example, because the literature suggests that the type of PRS depends on the novelty of products, we test the hypothesis that the impact of radical innovation is the same for different PRSs (that is, and particularly SSCs and SSPs. Comparing the coefficient within an equation, we are able to test differences in magnitude among the determinants (H4a). We introduced different variables crossing technological variables (product and process) and an organizational change variable to test for complementarity (H4b). H 1 : α j 1 > 0,α j 2 > 0a n d α j 1 > α j 2 H 2 : α j 3 > 0 α j 4 > 0) α1 > α2 > α4 α1 > α4 > α2 α3 > α4 α3 < α4 α2 < α1 < α4 α1 < α2 < α4 � 289 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi With respect to control variables, employees with a third-level educational qualification are first introduced because they are usually associated with the ability of firms to adapt more efficiently from an organizational point of view (Caroli & Van Reenen, 2001). We thus expect a positive effect of the share of such employees on the ability of manufacturing firms to offer PRSs. Following Acemoglu, et al. (2007), young firms have less experience with technologies and organizational arrangements. Following this general idea, we therefore expect that they may be less able to offer PRSs than other firms. It can, however, also be argued that young high-tech firms encounter more financial constraints and, consequently, PRSs (including consulting and R&D, for example) are a means to survive when their products have not yet reached the market. The expected direction of the impact of the YOUNG variable is thus not well defined. A similar trade-off is expected for the variable SIZE because large firms are supposed to be more diversified and consequently more likely to propose PRSs, whereas small firms may be more flexible and close to clients. Indeed, the latter’s comparative advantage is their ability to provide PRSs to their customers. Subcontractors are less likely to provide new PRSs surrounding their products because the contractual arrangements are likely to limit the extent of their offerings to their customers. The negative effect is likely to occur in particular when subcontracting is performed for capacity and cost reasons rather than for capability reasons. The heterogeneity of determinants can also rely on sectorial differences. For example, following Cusumano, Kahl & Suarez, (2015), we can expect that firms in mature industries (Low-Tech and Low- Medium-Tech industries) are more prone to offer services such as “smoothing” services (financing, repairing, training, installing, or documenting), whereas sectors in a more rapidly developing phase (Medium-High-Tech or High-Tech) are more likely to propose “adapting” services (software or consulting). To cope with the different limited dependent variables measuring PRSs, non-linear models were implemented. A probit model was introduced to explain the dichotomic variable PRSYes. When dealing with the different types of PRSs introduced by firms, the strategic choices regarding these PRSs are not independent of each other. For example, a lack of resources may induce some arbitrage among PRSs, or some PRSs may be linked to other PRSs. For example, maintenance may be a critical after-sales service. A further possibility is that firms benefit from economies of scope (Kowalkowski, Brehmer & Kindstrom, 2009). When some installing or maintenance activities are provided, consulting services can be provided at lower cost, for example. To cope with these strategic interactions and possible economies of scope in PRS offerings, we implemented a multivariate probit model (as in Santamaria, et al., 2012). We were therefore able to take into account the unobserved strategic orientations regarding some PRS bundles offered to customers. The correlation among residuals controls for complementarity, independence, or substitutability among the different PRSs proposed by firms (Gourieroux, 2000). Hence, the PRS bundling is taken into account but not assumed, as is often the case in the literature (e.g., Antioco, et al., 2008; Eggert, et al., 2011, 2014). Still, we also implement a bivariate ordered probit model to identify the differences between SSC and SSP determinants. Deploying a multivariate model, we are thus able to test H5 and the heterogeneity of the role of explanatory variables on different PRSs. � 290 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 RESULTS As reported in Table 2, approximately 66% of firms launched at least one new product over the 2003–2005 period. 36% of firms declared that their innovations were new to the market. Only 41% of firms introduced advanced manufacturing processes during the same period (0.7 on average). New practices for a flexible organization were adopted by 12% of firms. Only 12% of manufacturing firms did not offer any PRS to their clients. Consulting activities, documentation, software, and training were the most common PRSs supplied. Table 3 indicates that firms proposed different types of PRSs with heterogeneous frequencies. It also demonstrates that firms that introduced product innovation, process innovation, and organizational innovation were more prone to be active in PRSs. The increment is, however, less obvious for process innovation or organizational innovation than for product innovation. Table 2: Descriptive statistics � 291 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi Table 3: The explained variables by our three main explanatory variables The econometric results regarding the determinants of PRSs are reported in Table 4. The next hypotheses address the impact of product innovation on PRS strategies: H1 stated that radical product innovators are more likely to introduce PRSs or that PRS portfolio structures depend on the degree of product innovation introduced by firms. Controlling for a large number of variables, product innovation is found to be a significant determinant of servitization. An incremental product innovation boosts the likelihood of offering PRSs by 4.5% (Col. 1 in Table 4), whereas the introduction of a radical product innovation during the same period makes firms twice as likely (+9%) to offer PRSs. A Wald test confirms that radical innovation has a significant and higher impact than incremental product innovation (the Wald critical value is 4.64, p<0.05). The positive effect of radical innovation holds for all PRS types except financial services (Col. 5, in Table 4) and repairing services (Col. 8), whereas incremental product innovation plays a positive and significant role only for consulting services (Col. 2) and training services (Col. 7). Computation of marginal effects of the multivariate model allows us to interpret the magnitude of the impact of radical innovation on PRSs. Table 6 indicates that radical innovation especially boosts the probability of developing consulting activities (+7%, Col. 1 in Table 1) and training activities (+7%) and also, in a minor way, documentation services (+1.4%) and repairing services (+0.8%). From Table 6 (Col. 1 and Col. 6), we can see that when significant, the impacts of radical product innovation are about twice the impacts of incremental product innovation, in agreement with H1. Radical innovation positively influences SSCs and SSPs (Col. 9 and Col. 10, respectively), whereas incremental innovation is found with a lower magnitude than radical innovation. On the whole, our results support H1 and the importance of product innovators in PRS offerings. More precisely, they support the idea that radical product innovations are critical to offer PRSs � 292 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 Table 4: The determinants of PRS � 293 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi Note: .01 - ***; .05 - **; .1 - *, N=1129. No product innovation and Medium- High Tech are taken as the references. The independence of equations (2) to (8) is rejected (the critical value is 1129.3, p<0.01). The same is true for the independence of equations (9) and (10) (the value is 343,3,p<0.01). In other words, the decisions regarding the different PRSs are not made independently. The multivariate models were implemented using the Geweke-Hajivassiliou-Keane (GHK) simulator to evaluate the integrals in the likelihood function. The bivariate ordered probit model explains the SSC and SSP indexes. SSCs and SSPs: the SSC index is computed as the sum of consulting, financing, and training services, whereas the SSP index is defined as the sum of documenting, software, installing, and repairing services. The coefficients in column (1) are marginal effects, but those in columns (2) to (10) are not. The marginal effects for columns (2) to (8) are reported in Table 5. The Wald tests are two-tailed tests. The results regarding PROCESS confirm that firms that introduced or adopted many such innovations during the past three years were more likely to launch services around their products (Col. 1 in Table 4). The adoption of one additional AMT increased the likelihood of offering PRSs by 2% (Col. 1, in Table 4). However, process innovation was found to be linked, with p<0.1, only to consulting services (Col. 2 in Table 4). In contrast with the literature, our results do not provide clear support for H2 and the role of process innovation in PRS strategies. They suggest, however, that new processes can help firms extend their SSP offerings (Col. 10 in Table 4). A similar result was found for H3. A more flexible organization was not found to influence the propensity to enter into PRS activities (Col. 1). The conclusion is similar when the different types of PRSs are scrutinized: Flexibility is found to be influential only on the likelihood of doing repairing activities (Col. 8). Our results provide weak support for H3. Organizational innovation seems to favor the number of SSCs (Col. 9). This result is in agreement with Eggert, et al. (2014). The different coefficients can be compared to look for a hierarchy among the three determinants of PRSs (H4a in Table 4). A set of Wald tests underline that radical innovation is dominant over process innovation, whereas the dominance is not found over organizational innovation. The adoption of one additional AMT increases the likelihood of offering PRSs by 2%, whereas the introduction of radical product innovation boosts the same probability by a significantly higher 9% (Wald is 3.90, p<0.05 in Col. 1, Table 4). The impact of radical product innovation is found to be dominant over process for the different services, except for financial and installation services. The dominance of radical innovation is found particularly for consulting services (Wald is 5.58, p<0.05), the only PRS for which AMTs play a role. It also confirms that radical innovation is dominant for consulting and training. Organizational innovation is found to be dominant over radical innovation only for repairing services. The dominance of product innovation is found to be weaker, as expected, when incremental innovation is considered: The results presented in Table 4 suggest that incremental product innovation is less important than organizational innovation for installing services. When SSCs and SSPs are considered, the hierarchy is, however, more refined. Regarding the SSP index, radical innovation is still the dominant driver, dominating process and organizational innovation. Incremental product innovation impact is found to be superior to process innovation. Regarding SSCs, pioneer products are the only driver, the positive influence of a more flexible organization becoming significant (at � 294 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 10%). The different results provide support for H4a, that product innovation is the main driver of PRS offerings and that—despite some influences, especially on SSPs—organizational change and process innovation are found to be less influential. The lack of significance of the parameters of cross-effect variables is unexpected (tables available upon request). It suggests that technology and organization or process innovation and product innovation cannot be considered complementary in terms of their impact on PRS offerings. H4b is thus not supported. A consequence of this result is that the parameters reported in Table 4 and Table 6 provide a reliable view of the different impacts of the different determinants of PRSs and their hierarchies. We can finally consider to what extent the role of product novelty differs according to the type of PRS (H5). To answer this question precisely, we first focus on the multivariate probit model (Col. 2 to Col. 8 in Table 4) and perform Wald tests about the heterogeneity of the positive impact of radical product innovation on PRSs. The Wald critical value is 26.39 (with p<0.01); thus, we reject the hypothesis that the coefficients of radical product innovation are equal among the different PRS types. The same result is obtained for the positive impact of incremental product innovation (Wald is 12.12 at p<0.10). More precisely, it was found that pioneers in an industry are more prone to offer consulting activities (+6.6% in Col. 1 in Table 6) and training services (+6.7% in Col. 6 in Table 6), whereas their propensity to offer documentation (+1.4% in Col. 2) or repairing services (+0.8% in Col. 7) is enhanced only weakly. In other words, the structure of PRS portfolios depends on innovation and the degree of innovation recently introduced by firms. To characterize heterogeneity, we further tested the heterogeneity of coefficients on the bivariate ordered probit model: The coefficients of radical product innovation, incremental product innovation, and even process innovation were found to be different between the two SSP and SSC classes (the Wald critical values are, respectively 13.9 with p<0.01, and 5.33 with p<0.05 for Col.9 and Col.10 in Table 4). Only organizational innovation seems not to influence SSCs and SSPs in a different manner. Our results support H5 and the idea that firms tune the structures of their PRS portfolios according to the type and strength of technological innovation. Table 5: Correlations among the residuals of the multivariate probit model � 295 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi Table 6: Partial effects of the multivariate probit model With regard to control variables, we systematically find a positive influence for the share of employees with a third-level educational qualification within a firm. The PRS strategy does not usually depend on company age: Neither age nor age squared is significant in the estimations reported in Table 5 even if young firms exhibit a significantly lower likelihood of proposing software (Col. 4). Size negatively influences the probability of proposing installation services (Col. 6). Conversely, large firms are more prone to offer financial services and training sessions (Col. 5 and Col. 7, respectively). As expected, subcontractors are less likely to be involved in PRS strategies. This result is, however, not found for consulting and documentation services (Col. 2 and Col. 3, respectively). When the coefficients of sectorial dummies are considered, our results indicate that mature (low-tech) industries are less involved in PRSs, regardless of the explained variable. The computation of marginal effects (Table 6) further documents sectorial differences: Firms that belong to mature industries are less prone to offer repairing services, installation services, or training services than are firms from medium-high-tech industries. In ferment phase (high-tech) industries, software and financial services are much more likely to be proposed by manufacturing firms (Col. 3 and Col. 4 in Table 6). Consulting services (Col. 2 in Table 4) are the only PRSs that are not less likely to be proposed in mature industries. Our results suggest that young firms are not necessarily the ones with original PRSs or original business model proposals; rather, industry characteristics and, as formerly detailed, individual technological capabilities are more important than firm age. Finally, the positive correlations among residuals (Table 5) suggest that the different PRS activities are complementary. The positive correlation among residuals provides little evidence for bundling patterns: Some SSPs are rather highly correlated with each other, thus suggesting potential SSP bundles. The same table also indicates that SSCs are poorly complementary or even independent (consulting and financing services for example). Table 5 demonstrates that some customer-oriented services are proposed with some specific product-oriented services: Specifically, repairing is proposed with training, documentation is proposed with consulting, or training is proposed with software. These relationships are confirmed by the positive correlation among residuals in the bivariate ordered probit model (Table 4) that further shows that some SSCs are proposed with some specific SSPs. � 296 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 DISCUSSION AND CONCLUSIONS The growing importance of services led governments and scholars t o f o c u s o n k n o w l e d g e i n t e n s i v e b u s i n e s s s e r v i c e s ( K I B S ) . Deindustrialization of societies are however not systematic and manufacturing firms often propose product-related services to their customers. This paper explores the determinants of such servitization. More precisely, it examines whether technology and organizational changes facilitate servitization and whether they influence the structure of service portfolios offered by manufacturing firms. We further explore to what extent the influence depends on the type of PRS offered. In addition, we consider the hierarchy among the determinants of PRS strategy and their complementarity. The results indicate that product novelty is a main driver of PRS activities. It especially boosts consulting and training services. The structure of the PRS portfolio is dependent on product novelty. Organizational change toward a more flexible company or the adoption of new advanced manufacturing processes is found, with few exceptions, hardly to influence the choice of PRS. However, our results suggest that process innovation is positively linked to the breadth of SSPs, whereas organizational innovation drives the breadth of SSC offerings. Product, process, and organizational innovation are not found to be complementary drivers of PRSs. In the following, we first discuss our contributions to the literature and their managerial implications. We conclude by addressing the main limitations of our study and research avenues. CONTRIBUTIONS AND IMPLICATIONS Our first contribution to the literature is to provide empirical evidence that technological innovation has to be reinstalled in the PRS debate, which has historically focused on the organizational dimension. Our results underline that technology is not separable from PRSs (Baines, Lightfoot, Benedettini & Kay, 2009; Baines, Lightfoot, Evans & Neely, 2007; Gebauer, et al., 2005; Mathieu, 2001); rather, technological novelty of products allows and supports the ability of firms to offer PRSs. We further demonstrate that radical innovation is the primary determinant of PRSs and that incremental innovation is influential only on training or consulting activities, or on the number of SSPs. This study is to our knowledge the first to clearly identify such a link using a large multi-sector sample; the causality was, until now, only suggested or indirectly obtained in the literature (see Eggert, et al., 2011; Sultan, 2013; Tongur & Engwall, 2014). Our analysis offers clear consequences for the top managers who want to redefine their business model through PRSs. It first indicates that the introduction of PRSs is not a simple task that can be performed by every manufacturing firm: Managers should first evaluate the degree of novelty of their products to identify related service opportunities. Our results confirm the general insights derived from works on appropriation where services surrounding products as well as to be a leader and go down on the learning curve are important means to appropriate product innovation. Our empirical results thus also rejuvenate old results (Levin, et al., 1987) and provide details on an empirical investigation usually focused on legal appropriation tools. The importance of technology is critical also because PRSs depend on industry maturity as measured by its R&D intensity level. Our results indicate that PRSs are less prone to be delivered in (medium-tech) mature industries, at odds with the ideas that mature industries are more likely to � 297 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi propose services (see Fang, et al., 2008; Suarez, et al., 2013; Teece 1986). Our results also underline that some services are less prone to be delivered than others (Cusumano, Kahl & Suarez, 2015). If we consider that servitization is core to the transformation of business models (Visnjic, et al., 2014), our article demonstrates that business model innovations do not occur only in mature industries (as in Massa & Tucci, 2013; Sabatier, Craig-Kennard & Mangematin, 2012); rather, they are also more likely to emerge in medium-high-tech industries with more nascent traits even if high-tech industries are more prone to offer SSCs than SSPs. When firms are pioneers, top managers should thus pay more attention to complementing their product with PRSs when they are doing business in technological industries. An open question is that, whenever technological context and technological innovation are reinstalled as main drivers of servitization, how should managers in manufacturing firms manage servitization? Our results suggest that managers with novel products should particularly pay attention to opportunities in consulting and training services when they are not at the technological frontier. Prospects in documentation services and repairing services are also first to be explored. A further contribution is to investigate the characteristics of PRS portfolios proposed by manufacturing firms. Our empirical results emphasize that pioneers implement PRS portfolios that differ from those of laggards: A difference between the determinants was even found for SSC and SSP classes. Still, an empirical classification is robust when there is an internal consistency of classes and thus when the determinants of PRSs are similar within SSP and SSC classes (see Gaiardelli, Resta, Martinez, Pinto & Albores, 2014, regarding service classifications). To test the internal robustness of the two classes, we tested the equality of the same coefficients inside SSC and SSP classes. In this case, the equality of the coefficients regarding technology and organization cannot be rejected (critical values are available upon request), thereby suggesting that the heterogeneity of determinants is rather between SSP and SSC classes than within SSP and SSC classes. The different results provide unexpected support for the conceptual classification of SSPs and SSCs often adopted after Mathieu (2001) (e.g. Antioco, et al., 2008; Eggert, et al., 2011, 2014). Beyond, our results failed to identify clear empirical classes of PRS bundling strategies. It suggests that firms may achieve some economy of scope according to many PRS arrangements. An alternative interpretation is that manufacturing firms may be myopic and unable to measure the rate of returns of the different possible sets of PRSs, explaining the difficulties of choosing proper arrangements and of deriving performance from PRSs. Our results suggest that the literature considering process, product, services, and organization as an integrated system is still a normative literature: It seems that it does not explain what is going on in industry, in which the development of PRSs is mainly driven by the degree of product novelty. It may, however, be argued here that organizational traits are an overall characteristic of servitized firms and are not a determinant of individual PRSs: A more flexible organization is required when many PRSs are to be coordinated, which may explain why the coefficients are downward biased. To check this idea, we employed a count model explaining the total PRS breadth computed as the sum of the seven possible PRSs that confirms the positive and significant role of organizational change (results available upon request). The results of this robustness check are thus broader than the one we obtained for SSPs and thus broader than Eggert, et al.’s (2014) results. Managers should thus pay � 298 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 attention to improving the flexibility of their servitized organization when coordination of several PRSs is addressed. A final contribution is that our analysis of the determinants of PRSs provides a path to solve the servitization paradox. Servitization in our model depends on product innovation and not on the complementarity between technological innovation and organizational change. Our result suggests that the impact of servitization on performances may thus differ depending on the proportion of leading firms in the sample or the repartition of firms among industries. Such sample selection, or a service variable that may be correlated with error terms in a service-performance equation, may introduce inefficiency and inconsistency in the estimation of parameters (see Hamilton & Nickerson, 2003). Our article contributes to the future correction of sample selection biases or the specification of a more complex servitization-performance model where servitization is explained using an equation besides the main one regarding performance . A global implication of these clues is that the rate of return 7 of PRSs may be different and higher than suggested by the literature. Despite the apparent failure of PRS strategies among industries, managers in charge of technological pioneers and flexible firms should be less averse regarding the low rate of return of PRSs usually found. LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH Due to data constraints, the model has been kept parsimonious. Consequently, some problems have been overlooked. The three main limitations we detail now—measurement, causality, and dynamics—also delineate steps for future research. The first limitation comes from the availability and the quality of the indicators. A limitation underlined by Garcia & Calantone (2002) that we already mentioned is that we do not know whether the different PRSs are innovative: We are thus able to identify PRS innovation when the firm introduced a new category of PRS but not when some PRS innovations were done within a PRS category. Hence, we explain the level of PRSs by technological change and organizational change whereas further 8 measures on changes in PRS strategies and a distinction between non- innovative and innovative services (not provided elsewhere by competitors) would be of great value. A second shortcoming is that the same type of PRS can cover different activities. For example, training sessions can be implemented to complement the introduction of new technological products, whereas they can also be provided for mature products. A service can be sometimes classified either in an SSC or an SSP category (see Cusumano, Kahl & Suarez, 2015). A similar problem can be encountered for organizational change because a firm can adopt temporary decentralization routines to respond to particular customers (Tuli, et al., 2007). Finally, some missing variables should also be mentioned. In the present article, we propose a focus on PRSs only. Our results are original because they avoid the noise and the difficulties inherent in disentangling PRSs (differentiation) from product-unrelated services (diversification). Still, in the present contribution, we were not able to control for product-unrelated services, whereas the decisions regarding the two types of services could be interdependent and thus be investigated in a broader multivariate model. The second main limitation in our analysis is linked to the causal � 299 7. See Visnjic, Wiengarten & Neely (2016) for an a=empt to introduce both product innovaFon and serviFzaFon as complementary determinants of performances. 8 .This type of shortcoming usually occurs in the economics of innovaFon, where scholars commonly explain the level of producFvity by innovaFon (see Mairesse & Mohnense, 2010 for a survey). A first-difference approach is useful less for the precision it would bring to the present analysis of PRS changes than for the neutralizaFon of hidden determinants that are specific to each firm and are not observed variables, assuming that these unobservable variables are Fme- invariant (Wooldridge, 2010). M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi links assumed in our model. Potential problems arise when the causality between the explanatory variables and the explained variables is taken into consideration. We contend that technology and organizational variables impact PRSs, whereas some scholars may advocate that services also drive firms to launch new (integrated) products (e.g., Levitt, 1976; Davies, 2004, Lindberg & Nordin, 2008; Eggert, et al., 2011) and therefore that product innovation, process innovation and organizational changes may also be considered a consequence of PRS adoption. Firms learn from their customer when they enter into servitization; thus, they can use this accumulated knowledge to propose new products, new services and new organizational practices. We consider the possible complementarity between innovation and organizational drivers whereby some other sequences are possible: Innovation could be a determinant of organizational change (e.g., Henderson & Clark, 1990) or organizational change a determinant of product innovation (e.g., Zahra & Nielsen, 2002) or even of AMT (Cardoso, Pinheiro de Lima & Gouvea da Costa, 2012). We acknowledge that these possibilities would dilute the causal links we have claimed. In the present article, because of a lack of panel data, we can explain only one side of the story: PRSs are explained with lagged variables suggesting causality from innovation and organization toward PRSs. The aim to finally understand the complete causal links remains (see Visnjic, et al., 2014), which leads us to avenues for future research. A further step in this research would be to understand the frontiers of servitized firms. Following Vargo & Lusch (2004), who redefined the service-centered activity of product firms, we can insist on the critical role of customers in the transformation of business models. Efforts by scholars to incorporate SSCs beside SSPs in servitization studies are thus a laudable but incomplete step. Considering what has been performed for user innovation following Von Hippel’s insights (1986), the identification of customers’ characteristics and customers' roles into servitization processes seems, from our point of view, vital. However, some competencies may rely on other types of cooperative links, crafted by firms to be able to servitize their products (Windahl & Lakemond, 2010). For example, software can be sub-contracted or elaborated with different types of suppliers (see Sabatier, et al., 2012), thus suggesting that analysis of PRSs requires a more open model of servitization, à la Chesbrough (2003). A third research avenue regards not the scope of PRS analysis but rather the analysis of servitization dynamics. The literature often considers cycles of exploration and exploitation activities. Considering radical innovation as an exploration phase, some PRS cycles should also occur within a manufacturing firm. Eggert, et al. (2011) suggest that the dominance of SSC and SSP should evolve over time. Some PRSs may survive product replacement, whereas others may not (see Tripsas, 1997). An implicit assumption in the literature is that technological innovation does not destroy services. The counting of PRSs in a portfolio that we use in this article includes this potential PRS withdrawal. However, the analysis of PRS destruction or redefinition due to technological innovation remains to be measured and explored.   � 300 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 REFERENCES Acemoglu, D., Aghion, P., Lelarge, C., Van Reenen J. & Zilibotti, F. (2007). Technology, Information, and the Decentralization of the Firm. The Quarterly Journal of Economics, MIT Press, 122(4), 1759-1799. Antioco, M., Moenaert, R.K., Lingreen, A. & Wetzels, M.G.M. (2008). Organizational Antecedents to and Consequences of Service Business Orientations in Manufacturing Companies. Journal of the Academy of Marketing Science, 36(3), 337-358. Araujo, L. & Spring, M. (2006). Services, Products, and the Institutional Structure of Production. Industrial Marketing Management, 35, 797-805. Auh, S. & Menguc, B. (2007). Performance implications of the direct and moderating effects of centralization and formalization on customer orientation. Industrial Marketing Management, 36(8), 1022-1034. Bahrami, H. (1996). The emerging flexible organization: Perspectives from Silicon Valley. The California Management Review, 34(4), 55-75. Baines, T., Lightfoot, H.W., Evans, S. & Neely, A. (2007). State-of-the-art in product service-systems. Journal of Engineering Manufacture, Proc. IMechE Part B, 221(10), 1543-51. Baines, T.S., Lightfoot, H.W., Benedettini, O. & Kay, J.M. (2009). The servitization of manufacturing: A review of literature, reflection on future challenge. Journal of Manufacturing Technology Management, 20(5), 547-567. Baines, T.S. & Lightfoot H.W. (2013), Made to Serve: How Manufacturers Can Compete through Servitization and Product Service Systems, Chichester, UK: John Wiley and Sons. Baines, T., Bigdeli, A. Z., Bustinza, O. F., Shi, V. G., Baldwin, J. & Ridgeway, K. (2016). Servitization: revisiting the state-of-the-art and research priorities. International Journal of Operations and Production Management, forthcoming. Barabba, V., Huber, C., Cooke, F., Pudar, N., Smith, J. & Paich, M. (2002). A multimethod approach for creating new business models: The General Motors OnStar Project. Interfaces, Vol. 32, N. 1, 20-34. Barras, R. (1986). Towards a theory of innovation in services. Research policy, 15(4), 161-173. Boning, B., Ichniowski, C. & Shaw, K. (2007). Opportunity Counts: Teams and the Effectiveness of Production Incentives. Journal of Labor Economics, 25, 613–650. Bowen, D.E., Siehl, C. & Schneider, B. (1989). A Framework for Analyzing Customer Service Orientations in Manufacturin. The Academy of Management Review, 14(1), 75-95. Brynjolfsson, E. (1993). The productivity paradox of information technology. Communications of the ACM, 36(12), 66-77. Burns, T. & Stalker, G.M. (1961), The Management of Innovation, Londres, Tavistock. B u z a c o t t , J . A . & Ya o , D . D . ( 1 9 8 6 ) . F l e x i b l e manufacturing systems: a review of analytical models. Management science, 32(7), 890-905. Cardoso, R.D.R., Pinheiro de Lima, E. & Gouvea da Costa, S.E. (2012). Identifying organizational requirements for the implementation of Advanced Manufacturing Technologies (AMT). Journal of Manufacturing Systems, 31(3), 367-378. Caroli, E. & Van Reenen, J. (2001). Skill-Biased Organizational Change? Evidence from A Panel of British, French Establishments. The Quarterly Journal of Economics, 116(4), 1449-1492. Chesbrough, H.W. (2003), Open innovation: The new imperative for creating and profiting from technology, Harvard Business Press. Cohen, M., Agrawal, N. & Agrawal, V. (2006). Winning in the aftermarket. Harvard Business Review, 84(May), 129–138. Cusumano, M. A., Kahl, S. J. & Suarez, F. F. (2015). Services, industry evolution, and the competitive strategies of product firms. Strategic management journal, 36(4), 559-575. Davies, A. (2004). Moving into high-value integrated solutions-A value stream approach. Industrial, Corporate Change, 13(5), 727-756. Davies, A., Brady, T. & Hobday, M. (2006). Charting a p a t h t o w a r d s i n t e g r a t e d s o l u t i o n s . S l o a n Management Review, 47(3), 38–49. Dedrick, J., Gurbaxani, V. & Kraemer, K.L. (2003). Information technology and economic performance: A critical review of the empirical evidence. ACM Computing Surveys (CSUR), 35(1), 1-28. Doms, M., Dunne, T. & Troske, K.R. (1997). Workers, wages, and technology. The Quarterly Journal of Economics, 112(1), 253-290. Eggert, A., Hogreve, J., Ulaga, W. & Muenkhoff, E. (2011). Industrial services, product innovations, and firm profitability: A multiple-group latent growth curve analysis. Industrial Marketing Management, 40(2), 661-670. Eggert, A., Hogreve, J., Ulaga, W. & Muenkhoff, E. (2014). Revenue and profit implications of industrial service strategies. Journal of Service Research, 17(1), 23-39. Fang, E., Palmatier, R. & Steenkamp, J. (2008). Effect of service transition strategies on firm value. Journal of Marketing, 72(5), 1-14. Fichman, R. G., Dos Santos, B. L. & Ze, Z. (2014). Digital innovation as a fundamental and powerful concept in the information systems curriculum. Mis Quarterly, 38(2), 329-353. Finne, M., Brax, S. & Holmström, J. (2013). Reversed servitization paths: a case analysis of two manufacturers. Service Business, 7(4), 513-537. Froehle, C. M. (2006). Service Personnel, Technology, and Their Interaction in Influencing Customer Satisfaction. Decision Sciences, 37(1), 5-38. Gaiardelli, P., Resta, B., Martinez, V., Pinto, R. & Albores, P. (2014). A classification model for product-service offerings. Journal of cleaner production, 66, 507-519. Galbraith, J.R. (2002). Organizing to deliver solutions. Organizational Dynamics, 31(2), 194–207. � 301 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi Garcia, R. & Calantone, R., (2002). A critical look at t e c h n o l o g i c a l i n n o v a t i o n t y p o l o g y a n d innovativeness terminology: a literature review. Journal of Product Innovation Management, 19 (2), 110–132. Gebauer, H., (2009). An Attention-Based View on Service Orientation in the Business Strategy of Manufacturing Companies. Journal of Managerial Psychology, 24/1, 79-98. Gebauer, H., Fleisch, E. & Friedli, T. (2005). Overcoming the service paradox in manufacturing companies. European Management Journal, 23(1), 14-26. Gebauer, H. & Fleisch, E. (2007). An investigation of the relationship between behavioral processes, motivation, investments in the service business and service revenue. Industrial Marketing Management, 36(3), 337-348. Gebauer, H., Pütz, F., Fischer, T. & Fleisch, E. (2009). Service Orientation of Organizational Structure. Journal of Relationship Marketing, 8(2), 103-126. Gourieroux, C. (2000), Econometrics of Qualitative Dependent Variables, Cambridge University Press. Grubic, T. (2014). Servitization and remote monitoring technology: A literature review and research agenda. Journal of Manufacturing Technology Management, 25(1), 100-124. Hamilton, B. H. & Nickerson, J. A. (2003). Correcting for endogeneity in strategic management research. Strategic organization, 1(1), 51-78. Henderson, R.M. & Clark, K.B. (1990). Architectural innovation: the reconfiguration of existing product technologies, the failure of established firms. Administrative Science Quarterly, 35(1), 9-30. Hofmann, C. & Orr, S. (2005), Advanced manufacturing technology adoption—the German experience. Technovation, 25(7), 711-724. Joshi, K. D., Chi, L., Datta, A. & Han, S. (2010). Changing the competitive landscape: Continuous i n n o v a t i o n t h r o u g h I T- e n a b l e d k n o w l e d g e capabilities. Information Systems Research, 21(3), 472-495. Kaiser, U. (2000). New technologies and the demand for heterogeneous labor: firm-level evidence for the German business-related service sector. Economics of Innovation and New Technology, 9(5), 465-486. Kellogg, D.L. & Nie, W. (1995). A Framework for Strategic Service Management. Journal of Operations Management, 13(4), 323-337. Kirner, E., Kinkel S. & Jaeger A. (2009). Innovation paths, the innovation performance of low- technology firms—An empirical analysis of German industry. Research Policy, 38(3), 447-458. Kowalkowski, C., Brehmer, P. O., & Kindstrom, D. (2009). Managing industrial service offerings: r e q u i r e m e n t s o n c o n t e n t a n d p r o c e s s e s . International Journal of Services Technology and Management, 11(1), 42-63. Lay, G., Copani, G., Jager, A. & Biege, S. (2010). The Relevance of Service in European Manufacturing Industries. Journal of Service Management, 21(5), 715-726. Lei, D., Hitt, M.A. & Goldhar, J.D. (1996). Advanced Manufacturing Technology: Organizational Design, Strategic Flexibility. Organization Science, 17(3), 501-523. Lenfle, S. & Midler, C. (2009). The launch of innovative product-related services: lessons from automotive telematics. Research Policy, 38(1), 156-169. Leonard-Barton, D. (1992). Core capabilities and core rigidities: A paradox in managing new product development. Strategic management journal, 13(S1), 111-125. Levin, R.C, Klevorick, A.K, Nelson, R.R, Winter, S.G, Gilbert, R. & Griliches, Z. (1987). Brookings Papers on Economic Activity. Microeconomics, Special Issue, No. 3, 783-831. Levitt, T. (1976) The Industrialization of Service. Harvard Business Review, vol. 54, 63-74. Lightfoot, H.W., Baines, T. & Smart, P. (2011). Examining the information and communication technologies enabling servitized manufacture, Proceedings of the Institution of Mechanical Engineers. Journal of Engineering Manufacture, Part B, 225(10), 1964-1968. Lindberg, N. & Nordin, F. (2008). From products to services and back again: Towards a new service p r o c u r e m e n t l o g i c . I n d u s t r i a l M a r k e t i n g Management, 37(3), 292-300. Mairesse, J. & Mohnen, P. (2010). Using innovation surveys for econometric analysis. Handbook of the Economics of Innovation, 2, 1129-1155. Massa, L. & Tucci, C.L. (2013). Business Model Innovation. In M. Dodgson, D.M. Gann, N. Phillips (Eds.). The Oxford Handbook of Innovation. Oxford University Press: Oxford. Mathe, H. & Shapiro, R.D. (1993), Integrating Service Strategy in the Manufacturing Company, London: Chapman & Hall. Mathieu, V. (2001). Service Strategies within the M a n u f a c t u r i n g S e c t o r : B e n e fi t s , C o s t s , partnerships. International Journal of Service Industry Management, 12(5), 451-475. Matthyssens, P. & Vandenbempt, K. (1998). Creating competitive advantage in industrial services. Journal of Business & Industrial Marketing, 13(4/5), 339-355. Miller, D. (1986). Configurations of strategy and s t r u c t u r e : To w a r d s a s y n t h e s i s . S t r a t e g i c Management Journal, 7(3), 233-249. Miller, D., Hope, Q., Eisenstat, R., Foote, N. & Galbraith, J. (2002). The problem of solutions: balancing clients and capabilities. Business Horizons, 45(2), 3-12. Mont, O.K. (2002). Clarifying the concept of product– service system. Journal of cleaner production, 10(3), 237-245. Nambisan, S. (2001). Why Service Businesses are Not Product Businesses. MIT Sloan Management Review, Summer, 72-80. N e e l y, A . D . ( 2 0 0 8 ) . E x p l o r i n g t h e F i n a n c i a l C o n s e q u e n c e s o f t h e S e r v i t i z a t i o n o f Manufacturing. Operations Management Research, 2(1), 103-118. � 302 Technological innovation, organizational change, and product-related services M@n@gement, vol. 19(4): 277-304 Neu, W.A. & Brown, S.W (2005). Forming successful business-to-business services in goods-dominant firms. Journal of Service Research, 8(1), 3-17. Neu, W.A. & Brown, S.W. (2008). Manufacturers forming successful complex business services: designing an organization to fit the market. I n t e r n a t i o n a l J o u r n a l o f S e r v i c e I n d u s t r y Management, 19(2), 232-251. Nonaka, I. & Teece, D.J. (Eds.). (2001). Managing industrial knowledge: creation, transfer and utilization. Sage OECD (2002), The Measurement of Scientific and Technological Activities, Frascati Manual 2002, Proposed Standard Practice for Surveys on Research and Experimental Development OECD, OECD Publishing. OECD (2005), Oslo Manual. Guidelines for collecting, interpreting innovation data, Third edition, OECD, Paris. Oliva, R. & Kallenberg, R. (2003). Managing the transition from products to services. International Journal of Service Industry Management, 14(2), 160–172. Oliva, R., Gebauer, H. & Brann, J.M. (2012). Separate or integrate? Assessing the impact of separation between product and service business on service performance in product manufacturing firms. Journal of Business-to-Business. 19(4), 309-334 O'Reilly, C.A., III, & Tushman M.L. (2004). The Ambidextrous Organization. Harvard Business Review, 82(4), 74–81. Plé L., Lecocq X. & Angot J. (2010). Customer- i n t e g r a t e d b u s i n e s s m o d e l s : a t h e o r e t i c a l framework. M@n@gement, 13(4), 226-265. Quinn, J.B., Doorley, T.L. & Paquette, P.C. (1990). Beyond products: Services-based strategy. Harvard Business Review, 68(2), 58–67. Rammer, C., Czarnitzki, D. & Spielkamp, A. (2009). Innovation success of non-R&D-performers: substituting technology by management in SMEs. Small Business Economics, 33(1), 35-58. Robinson, W.T. & Chiang J. (2002). Product development strategies for established market pioneers, early followers, late entrants. Strategic Management Journal, 23(9), 855–866. Rosen, L.D. (1998). Service: The Next Frontier. Hospital Material Management Quarterly, 19(3), 29-34. Sabatier V., Craig-Kennard A. & Mangematin V. (2012). When Technological Discontinuities and Disruptive Business Models Challenge Dominant Industry L o g i c s : I n s i g h t s f r o m t h e D r u g s I n d u s t r y. Technological Forecasting and Social Change, 79(5), 949-962. Saccani, N., Visintin F. & Rapaccini M. (2014). Investigating the linkages between service types a n d s u p p l i e r r e l a t i o n s h i p s i n s e r v i t i z e d environments. International Journal of Production Economics, 149(2014), 226-238. Santamaria, L., Nieto M.J. & Miles, I. (2012). Service innovation in manufacturing firms: Evidence from Spain. Technovation, 32(2), 144-155. Sawhney, M. (2004). Creating growth with services. Sloan Management Review, Winter, 45(2), 34–43. Suarez, F.F., Cusumano M.A. & Kahl, S.J. (2013). Services and the business models of product firms: An empirical analysis of the software industry. Management Science, 59(2), 420-435. Sultan, N. (2013) Knowledge management in the age of cloud computing and Web 2.0: Experiencing the power of disruptive innovations. International Journal of Information Management, 33(1), 160– 165. Teece, D.J. (1986). Profiting from Technological Innovation. Research Policy, 15(6), 285-305. Teece, D., Pisano, G. & Shuen, A. (1997). Dynamic Capabilities and Strategic Management.Strategic Management Journal, 18(7), 509-533. Tether, B. & Bascavusoglu-Moreau, E. (2012). Servitization: The extent and motivations for service provision amongst UK manufacturers, Proceedings of DRUID, 19-21. Tongur, S. & Engwall, M. (2014). The business model dilemma of technology shifts. Technovation, 34(9), 525–535. Tripsas, M. (1997). Unraveling the process of creative destruction: Complementary assets and incumbent survival in the typesetter industry. Strategic Management Journal, 18(s1), 119-142. Tuli, K.R., Kohli, A.K. & Bharadwaj, S.G. (2007). Rethinking customer solutions: from product bundles to relational processes. Journal of Marketing, 71(3), 1–17 Turunen, T. & Toivonen, M. (2011). Organizing c u s t o m e r - o r i e n t e d s e r v i c e b u s i n e s s i n manufacturing. Operations Management Research, 4(1-2), 74-84. Ulaga, W. & Eggert, A. (2006). Value-based differentiation in business relationships: Gaining and sustaining key supplier status. Journal of marketing, 70(1), 119-136. Ulaga, W. & Reinartz, W. (2011). Hybrid Offerings: How Manufacturing Firms Combine Goods, Services Successfully. Journal of Marketing, 75(6), 5-23. Vandermerwe, S. & Rada, J. (1988). Servitization of business: Adding value by adding services. European Management Journal, 6(4), 314–324. Vargo, S.L. & Lusch, R.F. (2004). Evolving to a new dominant logic for marketing. Journal of marketing, 68(1), 1-17. Visnjic I., Neely, A. & Wiengarten, F. (2012). Another performance paradox ? A refined view on the performance impact of servitization, Working paper, Cambridge Service Alliance, May. Visnjic Kastalli, I. & Van Looy, B. (2013). Servitization: Disentangling the impact of service business model innovation on manufacturing firm performance. Journal of Operations Management, 31(4), 169-180. Visnjic, I. & Van Looy, B. (2011). Manufacturing Firms Diversifying into Services: A Conceptual, Empirical Assessment, proceedings of the 20th POMS conference, Orlando, Florida. Visnjic, I., Wiengarten, F. & Neely, A. (2016). Only the Brave: Product Innovation, Service Business Model Innovation, and Their Impact on Performance. Journal of Product Innovation Management, 33(1): 36-52. � 303 M@n@gement, vol. 19(4): 277-304 Arman Avadikyan & Stéphane Lhuillery & Syoum Negassi Von Hippel, E. (1986). Lead users: a source of novel product concepts. Management science, 32(7), 791-805. Windahl, C. & Lakemond, N. (2010). Integrated solutions from a service-centered perspective: Applicability and limitations in the capital goods industry. Industrial Marketing Management, 39(8), 1278-1290. Windahl, C., Andersson, P., Berggren, C. & Nehler, C. (2004). Manufacturing firms and integrated solutions: characteristics and implications. European Journal of Innovation Management, Vol. 7, No. 3, 218-28. Wise, R. & Baumgartner, P. (1999). Go downstream-the new profit imperative in manufacturing. Harvard Business Review, 77(5), 133-141. Wooldridge, J.M. (2010), Econometric analysis of cross section and panel data, MIT press. Youngdahl, W. (1996). An Investigation of service- based manufacturing performance relationships. International Journal of Operations & Production Management, 16(8), 29-43. Zahra, S.A. & Nielsen, A.P. (2002). Sources of c a p a b i l i t i e s , i n t e g r a t i o n , t e c h n o l o g y commercialization. Strategic Management Journal, 23(5), 377–398. Acknowledgments: We gratefully acknowledge ISI FhG in Karlsruhe for data access and the useful and friendly help they provided. � 304