Homework answers / question archive / It is a Critical Review of Literature (up to 2000 words including in-text citations, - 10% word count is acceptable)

It is a Critical Review of Literature (up to 2000 words including in-text citations, - 10% word count is acceptable)

Business

Share With

It is a Critical Review of Literature (up to 2000 words including in-text citations, - 10% word count is acceptable). You are required to do the following:

1. Based on the topic , prepare critical literature review of the 5 articles (article # 6-10).

2. Your review should identify key gaps in the literature, similarities and differences in theories and methods used, similarities and controversies in findings, compare and contrast themes of the articles, and make a final future research recommendation.

SUBMIT IN WORD FORMAT

YOU ARE REQUIRED TO CITE AND PROVIDE A REFERENCE LIST IN HARVARD FORMAT.

Technological Forecasting & Social Change 120 (2017) 163–175 Contents lists available at ScienceDirect Technological Forecasting & Social Change Origins and emergence of exploration and exploitation capabilities in new technology-based ?rms Are Jensen* , Tommy H. Clausen Nord University Business School, Bodø, Norway A R T I C L E I N F O Article history: Received 19 July 2016 Received in revised form 9 February 2017 Accepted 6 March 2017 Available online 9 May 2017 Keywords: Routines Exploration Exploitation Learning Search New technology based ?rms (NTBFs) A B S T R A C T The essence of new technology-based ?rms (NTBFs) is to generate new technologies and employ them in practice, contributing to a business sector in which many approaches to innovation have been exhausted. NTBFs are typically established around a founding team that possesses few resources but speci?c knowledge and a promising idea; NTBFs lack most of the advantages of incumbent organizations. To have a chance to succeed, these ?rms need to develop their organizational capabilities, particularly those for exploration and exploitation. Capabilities are fundamental building blocks of ?rms and key to organizational functioning. However, we have little understanding of how capabilities emerge within new organizations, including NTBFs. Therefore, we examine the origins of exploration and exploitation capabilities in NTBFs. A model of capability emergence is sourced from the literature, highlighting the role of routines for deliberate learning. A set of hypotheses concerning the antecedents and effects of routines for deliberate learning is tested, using a sample of 84 NTBFs and partial least squares structural equation modeling (PLS-SEM). This analysis offers empirical support for our model and hypotheses. Hence, the paper provides knowledge on the origins of NTBFs’ exploration and exploitation capabilities and, in particular, the role of routines for deliberate learning in this regard. © 2017 Elsevier Inc. All rights reserved. 1. Introduction New technology-based ?rms (NTBFs) are often highlighted as an important source of innovation and economic growth. One reason for this is that the essence of NTBFs is the generation, development and introduction of new technologies, ideas and innovations to the business sector in particular and society more broadly. A key advantage of NTBFs, as seen from a macro perspective and a societal standpoint, is that such companies ensure that a wider variety of approaches to innovation is employed, complementing the innovation searches and efforts of NTBFs’ larger, established counterparts in the business sector (Cohen and Malerba, 2001; Cohen and Klepper, 1992). NTBFs are typically established around a founding team that possesses few resources but speci?c knowledge and a promising idea; NTBFs lack the established organizational capabilities that support the innovation search efforts of incumbent organizations. To have a chance to succeed, NTBFs need to develop their organizational capabilities, particularly those for exploration and exploitation. Capabilities are fundamental building blocks of ?rms and key to organizational functioning, as they comprise the repertoires that the * Corresponding author. E-mail address: are.jensen@nord.no (A. Jensen). http://dx.doi.org/10.1016/j.techfore.2017.03.004 0040-1625/© 2017 Elsevier Inc. All rights reserved. organization’s members have and allow the organization’s resources to be employed productively (Helfat and Lieberman, 2002; Nelson and Winter, 1982). Within the context of exploratory and exploitative innovation, Jansen et al. (2006) [1661] note the following: “Research on exploration and exploitation is burgeoning, yet our understanding of the antecedents and consequences of both activities remains rather unclear.” Recently, scholars have demonstrated renewed interest in the topic, including in the context of new ?rms (Stettner et al., 2014; Bryant, 2014; Frigotto et al., 2014). The present paper helps to clarify the antecedents and consequences of exploration and exploitation within the context of NTBFs along three avenues. First, we study the origins of capabilities for exploration and exploitation. Second, we examine their temporal relationship, as a set of pre-organizational behaviors and as an organizational proclivity for subsequent exploration and exploitation. Third, we examine the mechanisms that convert their antecedents into actual capabilities. Most theorizing on capabilities focuses on the nature of these capabilities (e.g., Lavie et al., 2010; Gupta et al., 2006) and their in?uence on ?rm performance (e.g., Baum et al., 2014; Lin et al., 2013; Sirén et al., 2012; Nielsen and Gudergan, 2012; Yamakawa et al., 2011; Tu, 2010; Hoang and Rothaermel, 2010; Uotila et al., 2009; Yalcinkaya et al., 2007) within the context of large, established organizations. However, speci?c theorizing on capability emergence is 164 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 largely lacking, which contributes to the di?culty in predicting ?rm performance and development, particularly for new organizations such as NTBFs. Thus, we have a limited understanding of how capabilities originate and emerge in new organizations, including NTBFs. More than a decade has passed since Helfat and Peteraf (2003) concluded in their discussion of capability emergence and heterogeneity that “we lack a clear conceptual model that includes an explanation of how this heterogeneity arises” (:997). We remain in the dark in this regard. The purpose of this paper is to help theorize and build empirical research on how capabilities emerge in NTBFs. This is achieved by examining in greater detail the origins and emergence of the organizational capabilities of exploitation and exploration in NTBFs. We examine this issue by sourcing a model of capability emergence from a relatively scattered literature on this topic. In particular, the sourced model highlights how founder-mangers of NTBFs—a type of ?rm for which knowledge is the primary resource—share and combine their pre-organizational and newfound knowledge through routines for deliberate learning and how these recurring actions affect their ?rm’s proclivity for exploration and exploitation. To empirically test the sourced model, we ask the following research question: What is the role of routines for deliberate learning in NTBFs’ exploration and exploitation capabilities? A set of hypotheses concerning the antecedents and effects of routines for deliberate learning in the emergence of NTBFs’ capabilities for exploitation and exploration is derived from the model and subsequently tested, using a sample of 84 NTBFs from Norway and partial least squares structural equation modeling (PLS-SEM). Our research contributes to the search literature in the following ways: First, we craft a theory-derived model of how exploitation and exploration capabilities emerge in NTBFs during the early stages of their life-cycle and highlight the role of routines for deliberate learning in this regard. However, while our model highlights these routines, we also theorize about their antecedents in the context of NTBFs. Thus, we make the argument that routines for deliberate learning do not emerge as “manna from heaven” but result from the founder-managers’ prior knowledge and their behavior, two vitally important assets and resources to which otherwise resourcescarce NTBFs have access. Accordingly, our paper helps to develop a more complete and elaborate account of the emergence of NTBFs’ exploration and exploitation capabilities. Additionally, we derive hypotheses from the model and test them in the NTBF context. Second, and related, we address an important loci of analysis issue regarding the origins of exploration and exploitation capabilities, particularly in NTBFs. Our theorizing, model, and empirical results show that organizational exploration and exploitation capabilities can emerge through interactions among individuals within the ?rm through routines for deliberate learning. While the role of inter-?rm knowledge transfer routines is explored by, e.g., Dingler and Enkel (2016), we show that similar types of routines are equally important for intra-?rm knowledge search. We also show that routines do not necessarily reduce ?exibility. In the context of NTBFs, routines can translate into both a proclivity for associated with both e?ciency (exploitation) and ?exibility (exploration) (Eisenhardt et al., 2010). Finally, our paper contributes to the literature on ?rms’ innovation search activities. This literature often assumes the existence of a large, mature organization with both completed and ongoing organizational search processes (Santos, 2003). Less is known about how search capabilities initially emerge during ?rms’ youth. We address this issue by elucidating how NTBFs, the very essence of which is innovation search, acquire their initial capacity to explore and exploit, including (some of) the processes and mechanisms involved. 2. Theory and hypothesis development Exploration and exploitation play key and distinct roles in ?rms’ development. The following explanation is offered by March (1991): “Exploration includes things captured by terms such as search, variation, risk taking, experimentation, play, ?exibility, discovery, innovation. Exploitation includes such things as re?nement, choice, production, e?ciency, selection, implementation, execution” (:71). Whereas exploitation concerns searching for e?ciency (Eisenhardt et al., 2010) or for deep knowledge in areas related to what one already knows(Katila and Ahuja, 2002), exploration describes a broad search, or searching for knowledge in areas that are new to the ?rm(March, 1991). Capitalizing on exploitation leads to modest, fairly reliable, short-term market rents, whereas capitalizing on exploration is fraught with uncertainty and the potential for greater rewards if successful(March, 1991; Andries et al., 2013; Schulz, 2001). For mature ?rms, this translates into e?ciency gains (exploitation) and avoiding competency traps and organizational inertia (exploration)(Lavie et al., 2010; Greve, 2007; Levinthal and March, 1993), often through innovation (Aloini and Martini, 2012). While most studies on exploration and exploitation focus on large and mature ?rms (e.g., Baum et al., 2014; Lin et al., 2013; Nielsen and Gudergan, 2012; Yamakawa et al., 2011; Tu, 2010; Hoang and Rothaermel, 2010; Uotila et al., 2009; Yalcinkaya et al., 2007), they are just as relevant for new ?rms (Sirén et al., 2012). For example, exploitation is essential for new ?rms, as it leads to e?ciency, standardization and reliability (Eisenhardt et al., 2010) — key survival-enhancing attributes of organizations that are attempting to develop and commercialize new technologies (Aldrich and Yang, 2012). In contrast, exploration translates into ?exibility—a key attribute for NTBFs since technology development can necessitate search in uncharted territories, thereby allowing for unexpected and promising opportunities (Eisenhardt et al., 2010). Thus, whereas exploitation reduces the liability of newness, exploration enhances and makes up opportunity recognition (Politis, 2005). We consider ?rms’ capacity to explore and exploit as two related—yet distinct—capabilities. Organizational capabilities and their de?nition represent a hotly discussed topic. In this paper, we take them to mean “the repertoires of organization members ’that are ’associated with the possession of particular collections’ of resources including the ability to utilize those resources productively” (Helfat and Lieberman, 2002, :725, referring to Nelson and Winter (1982)). Therefore, the capability of exploitation would be the ?rm’s ability to exploit its resources to increase e?ciency (Eisenhardt et al., 2010), and in the case of knowledge, further deepen it (Katila and Ahuja, 2002). The ?rm’s exploration capability would be its ability to search broadly (Katila and Ahuja, 2002) to remain ?exible (Eisenhardt et al., 2010). Since the context of this study is NTBFs, the primary resource they would exploit and explore is knowledge that is either directly relevant for innovation search or for supporting the organizational ability to conduct search, although we recognize that other resources will also come into play, especially as the ?rm develops. The above de?nition of capabilities highlights the central role of individuals in the new ?rm setting. Ultimately, it is the organizations’ members who effectuate the ?rm’s capability to exploit and explore. The founder-managers are in this case those most likely to have both an understanding of the day-to-day operations of the ?rm and the organization as a whole, making them particularly useful informants. Moreover, we consider routines as both constituents of and inputs to capabilities(Helfat and Peteraf, 2003; Zollo and Winter, 2002). Whereas routines represent ways to perform actions associated with task execution or coordination (Helfat and Peteraf, 2003), a capability indicates the ability to successfully combine and direct several activities, together with resources, toward a speci?c goal(Helfat and Peteraf, 2003; Zollo and Winter, 2002). From this perspective, a routine is something that resides within the organization and is not considered a constituent of individuals. Therefore, in NTBFs, these routines must somehow be created by their founders. While entirely new capabilities may be born and created in mature ?rms, their progenitors are often already established A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 capabilities(Helfat and Peteraf, 2003; Argyres et al., 2012). This quickly leads to “in?nite regression” types of problems, in which the emergence of an entirely new capability is explained by the existence of other capabilities residing within the ?rm (Argyres et al., 2012). As a consequence, tracing the origins of capabilities becomes an intractable problem when attempting to generalize the capability emergence phenomenon1 . Additionally, organizational inertia limits mature organizations’ ‘wiggle-room for action’, that is, their inability to perform actions outside their core expertise, making them vulnerable to technological shifts (Ho and Lee, 2015). Hence, understanding mature ?rms’ capability emergence is a very di?cult undertaking. In NTBFs, however, capabilities have not yet formed (Basu et al., 2015), and the organization’s room for action is yet to become pathdependent. NTBFs are therefore ‘clean slates’ for studying capability emergence (Helfat and Lieberman, 2002; Felin and Foss, 2009). It is important to note that capabilities never predate the ?rms’ birth (Basu et al., 2015). Hence, the elements and mechanisms of capability emergence must, at the earliest, come together when the organization is born. We draw on two major insights from the entrepreneurship and strategic management literatures to explain how this emergence and manifestation take place. In particular, (a) the entrepreneurship literature provides a lens for examining capability emergence in light of the nexus of founder-managers’ behaviors and prior relevant knowledge during the creation and early organization of a ?rm, and (b) complementing this perspective, we include theorizing from strategic management on the role of routines for deliberate learning as a catalyzing mechanism that “converts” the behaviors and prior relevant knowledge of foundermanagers into organizational capabilities. These routines for deliberate learning represent a formalization of combining behavioral and cognitive approaches to organizational learning. Such routines are intended to aid in articulating and codifying experiences into knowledge from the individual to the organizational level (Zollo and Winter, 2002). Grounded in the assumption that entrepreneurship scholars typically focus on experiential learning, Evangelista and Mac (2016) recently discovered that new ?rms greatly bene?t from deliberate learning processes in export markets. Berghman et al. (2013) report similar effects of deliberate learning for ?rms that are building innovation capacity. They ?nd that deliberate learning increases the ?rm’s absorptive capacity and, as a consequence, innovation capacity. While deliberate learning and its associated routines have yet to be examined in the new ?rm context, organizations seem to bene?t greatly from them in other highly dynamic contexts (Romme et al., 2010; Nembhard and Tucker, 2010). How deliberate learning affects capability development remains unclear, but in a recent contribution, Romme et al. (2010) ?nd, in a system dynamics simulation study, that there is considerable complexity involved in the transition from the stage of deliberate learning to capability development. This complexity arose as the interactions between the deliberate learning process and routines created tipping points, whereby organizational and environmental characteristics would in some cases limit the outcomes of deliberate learning but in other cases presented little hindrance. Romme et al. (2010)’s ?ndings highlight the importance of carefully examining the effects of routines for deliberate learning in different contexts, as their outcomes can vary greatly depending on the type of organization being studied and the level of dynamism in the ?rm’s environment. Exactly which organizational characteristics affect capability development in NTBFs remains an open question. Recently, Camps and Marques (2014) provide evidence indicating that individuals’ social interactions are necessary for developing 1 There are examples from individual case studies in which the authors attempt to untangle these microfoundations, (e.g,. Schneckenberg et al., 2015), illustrating the complexities involved in such an undertaking. 165 innovation capabilities. Moreover, Wang and Hsu (2014) illustrate the importance of inter-personal power for exploitation innovations in particular. While Zollo and Winter (2002)’s initial conceptualization of routines for deliberate learning focused primarily on ‘hard’ knowledge exchange and placed less emphasis on ‘soft’ social interactions, in practice, both types of interactions undoubtedly take place when enacting the ‘performative’ aspect of routines for deliberate learning (Feldman and Pentland, 2003). Our reasoning is illustrated in Fig. 1: prior relevant knowledge and the behavior of founder-managers are presented as antecedents of capability emergence in new ?rms; routines for deliberate learning represent an organizational mechanism converting and manifesting experiences (i.e., prior relevant knowledge and search behaviors) into organizational capabilities for exploitation and exploration. The model and its relationships are elaborated on in the sections below. Model-testing hypotheses are also developed. 2.1. Prior knowledge, behavior and capability emergence in the new ?rm context At the moment of ?rm creation, the ?rm is a collection of individuals, each of whom has unique knowledge, experiences and ways of doing things. To effectively pursue business opportunities, these individuals must act together through collective behavior while capitalizing on their combined knowledge. In combination, this can kick-start organizational knowledge creation (Smith et al., 2005). However, as a ?rm starts its life-cycle, the ?rm’s formal structure is weak, or even absent, and organizing is simple (Gartner, 1988; Katz and Gartner, 1988). NTBFs are often completely reliant on their knowledge for gaining an advantage, in both the market and technological sense (Katila et al., 2012; Bollinger et al., 1983). Those NTBFs that keep their research and development to a minimum are still reliant on their members’ technological knowledge and how the members use it within the ?rm. In fact, founding teams with considerable technological knowledge may have attained this knowledge at the expense of knowledge on how to operate a business or interact with the market. Therefore, we can be certain that knowledge is a central resource for NTBFs. Such knowledge must be discussed, made explicit, and articulated among the organization’s members to become useful for the organization as a whole. The higher the level of prior relevant knowledge among team members is, the greater the need for such externalization. Since much of the founding team’s knowledge is likely to be tacit, routines for deliberate learning enable the preservation of knowledge within the organization as a whole. Nonaka (1994), Nonaka (1991) and Nonaka and Von Krogh (2009) argue that knowledge exists on an explicit–tacit continuum. Whereas ‘true’ explicit knowledge is close to information, tacit knowledge require intuition or non-articulated mental models (Nonaka and Von Krogh, 2009). While ‘true’ tacit knowledge is by de?nition impossible to transfer to others, decreasing amounts of ‘tacitness’ increase Fig. 1. Model of exploitation and exploration capability emergence in new technology based ?rms. 166 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 the ease of codi?cation (Nonaka and Von Krogh, 2009). Highly specialized knowledge is often tacit in nature (Zollo and Winter, 2002) and is closely tied to the context in which it was created (Nonaka and Von Krogh, 2009). Such knowledge has great value but requires great effort to codify and articulate (Nonaka and Von Krogh, 2009). However, when this knowledge has been externalized, it becomes less costly to share with others (Grant, 1996a; Grant, 1996b). As a consequence, teams of founder-managers possessing high degrees of relevant prior knowledge experience a greater need to capture explicit knowledge through routines for deliberate learning. The highly tacit nature of prior knowledge should make it more important to quickly establish routines for deliberate learning at an early stage. As a consequence, the presence of large repositories of prior relevant knowledge necessitates establishing routines for deliberate learning, which acts as a catalyst. It is important to note that even in cases in which the team does not consider the capability-related consequences of sharing such knowledge, a logic of appropriateness is likely to be adopted, leading to similar outcomes (March and Olsen, 2013). Even at the very early stages of a ?rm’s life-cycle, there is a possibility that high levels of tacit knowledge can lead to di?culties in communication among team members and can consequentially be a source of competency traps. While that is the case, there is no reason to believe that early organizational members are unaware of this problem. Instead, as it becomes increasingly di?cult to convey tacit knowledge, members are likely to intensify their efforts to understand one another. That is, the early organizational members’ actual deliberate learning may be hindered by their tacit knowledge, leading to an increased focus on establishing routines for deliberate learning. Finally, Bresman (2013) ?nds that groups within an organization tend to change their routines based on the prior related experience of other groups they know. Groups that have spent considerable time on a routine in the past and were more successful at doing so will have a greater in?uence on groups just beginning the routine or in the process of searching for similar routines. Furthermore, Bresman (2013) reports that when groups ?rst come together, they tend to almost immediately search for routines that will bene?t them, supporting the notion that founding teams will quickly establish routines that they consider bene?cial for their ?rm. Murmann (2012) similarly ?nds that when recruiting high-knowledge individuals, the ?rm tends to retain subsequently attained related knowledge by using methodologies that the recruited members brought with them into the ?rm. While Murmann (2012)is unable to ascertain the underlying mechanisms for why this was the case, the insights provided by Bresman (2013) indicate that individuals possessing high levels of knowledge tend to bring with them, or re-establish, the routines they applied to acquire their knowledge in the ?rst place. Since the founders of the NTBFs in question are reliant on maximizing the utility of one of the few resources they have at their disposal—the knowledge that they bring with them into the ?rm— they are likely to establish routines for utilizing this knowledge to the fullest extent possible. It is likely that they will rapidly establish routines for managing their most important resource, namely knowledge, particularly if they bring with them established methodologies and a large amount of prior relevant knowledge (Bresman, 2013; Murmann, 2012). This leads to our ?rst hypothesis: Hypothesis 1. There is a positive association between NTBFs’ earlystage organization members’ prior relevant knowledge and their routines for deliberate learning. New ?rms are generally resource constrained (Katila et al., 2012; Bollinger et al., 1983). One of the few resources they have available is the knowledge encapsulated within the founding team. The success of the new ?rm is therefore dependent on effectively utilizing knowledge through the founders’ search behaviors. Such resource constraints affect the ?rm’s optimal con?guration of depth search (exploitation) and breadth search (exploration) (Cohendet and Llerena, 2003). Moreover, to be effective, individual organizational members’ search behaviors require coordination, as does the knowledge that they ?nd. Hence, the experiences gained through behavior need to be shared to constitute a basis for organizational action and collective behavior. Recognizing the cyclical nature of broad (i.e., exploration), and narrow/deep (i.e., exploitation) knowledge search, Zollo and Winter (2002)note that in circumstances in which such searches for knowledge are common, routines for deliberate learning are particularly helpful. Other scholars in addition to Zollo and Winter (2002) emphasize the importance of iterative learning processes through which members re?ect upon their new knowledge (Ciccotello et al., 2000; Edmondson, 1999, 2003; Winter, 2000), sometimes in a routine-like fashion. In the context of new ventures, routines for deliberate learning have received little attention. Zahra et al. (2006), however, argue for examining whether “learning modes” and their consequences differ with respect to ?rms’ age and competitive position. They further argue that learning-related assumptions made by entrepreneurship and strategic management scholars must be carefully examined and in greater detail, as their outcomes are re?ected at different levels of analysis. That is, “learning modes” that are bene?cial for ?rms that survive into maturity might also facilitate their demise during the early stages of their life-cycle (see Levinthal and Posen, 2007, for an example of such unintended consequences). Therefore, routines for deliberate learning act as a gate-keeper that determines what knowledge should be kept within the ?rm, thereby affecting organization-level outcomes. Re?ecting the above, during the early stages of the ?rm’s organizing, variation will be generated in founders’ behaviors; subsequent feedback will provide evidence of whether these behaviors will lead to the expected results (Nelson and Winter, 1982; Zollo and Winter, 2002). If the behaviors are deemed to be bene?cial and worth repeating, the founders may want to capture their blueprint through routines for deliberate learning (Aldrich and Yang, 2012). Additionally, when founders spend large amounts of resources and energy on certain patterns of actions (i.e., behaviors), they likely intend the organization to learn from the outcomes of those actions, thereby reducing their liability of newness (Aldrich and Yang, 2012; Politis, 2005) and uncertainty (Becker and Knudsen, 2005) while improving opportunity recognition (Politis, 2005). If such learning is bene?cial, e.g., perceived to be related to the ?rm’s ability to form capabilities, routines for deliberate learning are likely to form (Zollo and Winter, 2002). While the behaviors of deep knowledge search (exploitation behavior) and broad knowledge search (exploration behavior) will not by themselves generate routines for capturing their outcomes, research on the role of transitive memory in the formation of organizational routines might provide insights into why increased levels of deep and broad knowledge search will encourage the establishment of knowledge-capturing routines. For example, Miller et al. (2014) highlight that group members not only search for what to do through established routines but equally for who can do it. That is, group members who acquire new knowledge leave their ?rms unable to utilize this knowledge outside the group member’s speci?c tasks. One obvious way of avoiding this issue is to establish routines that both convey what should be done, and who can do it. Only then can newfound knowledge be captured for use in more complex routines and capabilities (Miller et al., 2014). Tippmann et al. (2013) arrive at similar conclusions by ?nding that socialized search and, speci?cally, organizing for knowledge circulation within the ?rm can lead to the establishment and development of routines. Since knowledge is a central resource for NTBFs and founders’ search behaviors depend on using and circulating this knowledge A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 among themselves, it is likely that such search will stimulate the emergence of routines for improving their knowledge search. Routines for deliberate learning represent such a routine. This leads us to our second hypothesis: Hypothesis 2a. There is a positive association between NTBFs’ earlystage organization members’ exploitation behavior and their routines for deliberate learning. Hypothesis 2b. There is a positive association between NTBFs’ earlystage organization members’ exploration behavior and their routines for deliberate learning. 2.2. Routines for deliberate learning and capability emergence Routines for deliberate learning may be established with the intention of capturing knowledge for the future, e.g., as capabilities; this is especially likely in settings where tacit knowledge is dominant and the likelihood of personnel turnover is high, as is typical in NTBFs. However, it is also possible that these routines emerge as a consequence of individuals wanting to exchange and combine information (Smith et al., 2005), possibly to overcome di?culties in exchanging tacit knowledge. While it might not be clear “who is driving the bus,” (March and Olsen, 2013) it is likely that these routines will emerge within NTBFs. As argued by Martin and Eisenhardt (2010), “deliberate learning activities are effective because they provide new information that elaborates [an] idea and improves its quality” (:282). Moreover, new ?rms are likely to generate knowledge that will take on a tacit nature within founder-managers, and it is also likely that the same founder-managers will have to take on other domains of responsibility, or even leave the ?rm, as time passes. As a consequence, founders establishing routines for deliberate learning positions the ?rm to convert knowledge into capabilities for the future. Little or no investment in routines for deliberate learning would suggest that the organization is relying on knowledge accumulation in a semi-automatic fashion. Since this knowledge is tied to individuals, it leaves the organization as a whole vulnerable to changes in its work-stock (Zollo and Winter, 2002). This is a particular problem in the context of new ventures. For example, in a longitudinal study, Guenther et al. (2016) ?nd that the likelihood of ?rm closure increases as early-stage members leave the organization—a possible indication that valuable knowledge is lost to the ?rm. Following our de?nition of capabilities, it becomes evident that ?rms’ capabilities are dependent on the repertoires of knowledge associated with their members. Hence, routines for deliberate learning become a way to ensure that this knowledge becomes part of the organization as a whole—in the form of capabilities—rather than solely being part of its creators. While founders’ prior relevant knowledge and their search behaviors during the early stages of an NTBF may be crucial, they are not guaranteed to lead to organizational capabilities: Some founders may leave or may take on other tasks, and new members are likely to join the organization. As a consequence, an NTBF’s overall capability to explore and exploit may deteriorate if the knowledge and behaviors of the founders are not somehow captured within the new organization. This might be one of the reasons for why changes in the founder team seem to affect the likelihood of ?rm survival (Guenther et al., 2016). A mature ?rm’s ability to exploit existing capabilities is fundamental for their e?ciency. NTBFs, however, possess few capabilities, and therefore, capability exploitation is less important. Instead, these ?rms must establish capabilities that will enable them to exploit one of their core resources: their knowledge. Such knowledge exploitation (and exploration) is extensively discussed in the strategic management literature, often in the context of dynamic capabilities (O’Reilly and Tushman, 2008). While NTBFs do not 167 yet require capability recon?guration, their knowledge exploitation capabilities directly relate to exploiting the ?rm’s primary resource— knowledge—and should therefore serve a similar role as capability exploitation does in mature ?rms. Unfortunately, the literature on how new ?rms manage knowledge has been relatively silent on the issue of how new ?rms exploit knowledge. In the search literature, however, Cohendet and Llerena (2003) ?nd that contextual factors have a profound impact on optimal strategies for managing search breadth (exploration) and search depth (exploitation). For example, resource constraints lead ?rms to broadly search, or explore, their external networks to greater extents. Conversely, depth search is di?cult to perform when resources are limited, and therefore, resource constrains lead to decreased depth search in external networks. These ?ndings are in opposition to the conventional wisdom that exploitation is cheap in terms of resources, while exploration is expensive. While this is probably true in terms of internal resource usage, when searching externally, exploration and exploitation might have opposite effects. Hence, there is reason to believe that such contextual factors will push an NTBF toward both exploitation and exploration. Individuals’ knowledge capabilities can be traced back to their knowledge-sourcing routines. These routines relate to using not only one’s own knowledge but also the knowledge of others. Those who are able to combine both sources tend to develop their personal knowledge capabilities to a greater extent (Shin and Choi, 2014). Similarly, for the NTBF as a whole, as an organizational capability is the collection of people’s knowledge and resources, facilitating learning routines for the ?rm’s members should bolster the ?rm’s capabilities. When such learning routines are established, they will likely further strengthen the ?rm’s proclivity to engage in further depth search (exploitation) and breadth search (exploration): Betsch et al. (2001) show that in the presence of strong routines, individuals tend to prefer to focus on information that strengthens the routine. That is, those who have strong routines for deliberate learning tend to focus on seeking new knowledge. In turn, this should further strengthen the ?rm’s proclivity for exploitation and exploration. In sum, these routines for deliberate learning should increase an NTBF’s proclivity for both exploitation and exploration. This leads to the next hypothesis: Hypothesis 3a. There is a positive association between NTBFs’ routines for deliberate learning and their emerging exploitation capability. Further discussion is needed on the relationship between routines for deliberate learning and exploration. From the literature on exploration and exploitation, one might assume that exploration will always be hindered by routinizing. This might not be the case in terms of the exploration and exploitation of knowledge, especially in new ?rms. In the search literature, there is evidence to suggest that for mature ?rms, only the reduction in routine variation will eventually lead to a reduction in the organization’s ability to explore new knowledge. During the early stages of the ?rm’s life-cycle, this effect is not as pronounced, if extant at all (Benner and Tushman, 2002). In fact, during the early stage of the ?rm, when the organization is small, it is more likely that routines for deliberate learning will act as a means for enabling individuals’ knowledge-sourcing routines and hence strengthen their individual knowledge capabilities (Shin and Choi, 2014). While this might not hold in larger organizations due to the di?culty of sharing information across groups, for small ?rms such as NTBFs, it is easier to transfer knowledge in this way, thereby also bolstering the organization’s ability to source knowledge, in the breadth dimension (exploration). Furthermore, Betsch et al. (2001) show that those ?rms that establish strong routines make organizational members focus on strengthening them further when faced with new situations. That is, those ?rms that emphasize routines for learning also tend to focus on 168 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 searching for knowledge when faced with new situations, including in the breadth dimension. Therefore, ?rms that establish strong routines for deliberate learning are likely to engage in breadth search, or knowledge exploration. The notion that routines will always lead to inertia and therefore a reduction in the ?rm’s exploration capability is common among scholars. While the establishment of routines may lead to inertia, the same is not necessarily the case for the ?rm’s capability to explore. In a recent contribution, Yi et al. (2016) ?nd that even inert routines can have positive effects on such knowledge search, as inert routines can force the ?rm’s members to search for new ways of overcoming the limitations imposed by the routine. Through this process, the ?rm’s members must explore new knowledge. Even when an inert routine provides no such resistance, it will still free up cognitive resources that are highly bene?cial when exploring new knowledge (Levinthal and Rerup, 2006; Weick and Roberts, 1993; Bigley and Roberts, 2001). It is important to bear in mind that we are studying the very early stages of capability emergence. It may be the case, as argued by Bryant (2014), that increased routinizing can lead to a loss of exploration capabilities as the ?rm grows larger and older. This, however, is beyond the scope of this paper. While many of the same arguments used for the previous hypotheses still hold for the relationship between routines for deliberate learning and exploration, the above shows that the conventional wisdom that routines hinder exploration is unlikely to hold for routines for deliberate learning and in the context of NTBFs. This leads to the ?nal hypothesis: Hypothesis 3b. There is a positive association between NTBFs’ routines for deliberate learning and their emerging exploration capability. 3. Method 3.1. Sample and data collection We collected data from a registry containing the contact information of founders of ?rms created with the intention of commercializing new technology through in-house research and development. This registry was established by the Norwegian government as part of a low barrier of entry program where ?rms can register their R&D projects and receive tax exemption on R&D related costs (the SkatteFUNN program). To enter the program and the registry, the founder-manager must write a description of his or her R&D projects and plans and have them approved by the program. By contacting one of the program managers at the Norwegian Research Council, we were able to secure a list of all 4-year-old ?rms in the register at the time of data collection. The list contained the e-mail address of each ?rm’s contact person and their ?rm’s unique national organization ID code. This served as a basis for sending the survey via e-mail. If the contact person was unable to reply to the survey, we encouraged him or her to instead forward it to someone in the ?rm who could. Since the ?rms are all of small size, we deemed this to be a routine procedure. Of the 385 potential respondents found in the registry, we received 97 replies, yielding a25% response rate. All the ?rms were subsequently traced—based on their unique national organization ID code—back through their history in the National Business Registry. Their corporate structures were analyzed (e.g., checking to see if the ?rm is actually a new ?rm and not a new legal entity started for tax-planning reasons). All our checks con?rmed the quality of our sample. It also served as a check of the homogeneity of the sample. t-tests comparing the pro?t and loss, capital structure measures, FTEs (full time equivalents), and employee numbers of the full sample to the population provided evidence that missing responses were in all likelihood random. Finally, we excluded 13 cases since they stated that they were sole-founders, leaving us with a ?nal sample of 84 ?rms. Before the survey was sent to the full population of ?rms we identi?ed a sub-group (excluded from the ?nal survey e-mail list) of NTBFs for testing the survey instruments’ translation and adaptation. One of the researchers involved in the research project (but not a coauthor of this paper) met with founder-managers of these ?rms and observed them while they completed the survey. During this process, the respondents were asked to describe each item and the group of items it belonged to (the measure itself) out loud to allow us to observe whether our and the respondents’ understanding of both the items and the concept in general were in congruence. Thereafter, the researcher and the respondents discussed how, e.g., translations of core terms could be improved upon such that translations and adaptations remained as faithful as possible to the original. Following these adjustments, one of the authors of this paper met with two additional founder-managers following the same procedure described above. All items and corresponding sources are shown in Table 3 in the appendix. 3.2. Dependent variables: exploration and exploitation capabilities Our measure of exploitation capabilities is based on a measure developed by Alsos et al. (2008). The measure captures the ?rms’ ability to continually extract value from their existing products and processes, particularly at the ?rm level. We use Makkonen et al. (2014) measure of exploration capability, therein as part of dynamic capabilities. Since our population is restricted to new ?rms, we omitted the last item of the scale: “Our ?rm systematically transfers resources to the development of new business activities.” The majority of our population is highly unlikely to be in a position to have resources to systematically transfer across business areas. As a consequence, the inclusion of this item would therefore act as a proxy for the ?rms’ size, favoring larger ?rms. This would decrease the measure’s accuracy in predicting exploration capability among typical, new, ?rms. This is also supported by the statistical properties of the item when included in the measurement model. 3.3. Independent variable: routines for deliberate learning This measure of deliberate learning is designed to capture routines, formal systems, and established practices in the new venture, all intended to preserve the knowledge the ?rm that has attained throughout its lifetime. We were unable to ?nd a suitable measure of routines for deliberate learning for early-stage organizing in the literature. Therefore, we developed such a measure of deliberate learning based on Zollo and Winter (2002). The initial creation of the measure was undertaken as a group effort. It was then sequentially re?ned through the testing process described above. The measure’s manifest variables are designed for the context of our population—new ?rms commercializing research and development. As a consequence, they re?ect both weak and strong degrees of formalization of deliberate learning (in terms of both articulation and codi?cation), as would be common among smaller founding teams. 3.4. Independent variables: exploration and exploitation behavior For this measure, we used items developed by Mom et al. (2009, 2007). Their measure was developed for the context of large ?rms, and as a consequence, we omitted some items that were not applicable for our population of interest. For example, an item re?ecting exploitation behavior is “Activities which clearly ?t into existing company policy.” Since our ?rms are newly formed, they are unlikely to have formalized company policies. Including this item would therefore give a less accurate measure of exploitation behavior in our population. The reader should note that these items all relate to team behaviors undertaken during innovation projects. A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 3.5. Independent variable: relevant prior knowledge Our measure of relevant prior knowledge is adapted from Zhao et al. (2013), who in turn adopted their measure from DeSarbo et al. (2005). Since we are focusing on R&D-centric ?rms, the items constituting the measure are all related to the founders’ (1)ability and experience in developing new products and services and (2) understanding of existing relevant technology markets and their development. The items re?ect their translated nature, as some nuances in the questions were lost, and others added, as part of the translation and adaptation process. 3.6. Control variables There are several factors that could in?uence behavior, learning, and capabilities in our model beyond our explanatory variables: (a) The number of members on the founding team (continuous) could affect behaviors through the presence of greater human capital; (b) larger teams would require higher degrees of formalization, e.g., routines, for learning as a consequence of the challenges inherent in organizing groups of people; (c) the number of previously undertaken research and development projects (continuous) is an indication of the team’s overall tenure; (d) and, ?nally, team climates and processes such as cohesion and affective and cognitive con?ict (psychometric, Likert-type, derived from the research literature) as informal social interactions are likely to affect exploitation and exploration behaviors (Jansen et al., 2009; Jansen et al., 2006).2 We controlled for these unrelated in?uences in various ways, e.g., including them as competing explanatory variables for routines for deliberate learning (playing the same role as prior relevant knowledge) and as direct explanatory variables of capability emergence (playing the same role as routines for deliberate learning). The core model presented in the paper remained robust across all variations. 3.7. Techniques We chose to adopt PLS-SEM as our analytical tool for two major reasons: non-normal data and a small sample size. Using co-variance based SEM, such as ML-SEM, will likely lead to erroneous conclusions in such cases (Hair et al., 2012). For our statistical procedures, we used R (R Core Team, 2015) and the semPLS package (Monecke and Leisch, 2012). Since we have an excellent overview of our full sample and population, we decided to use multiple imputation to calculate missing values in our survey (Buuren and Groothuis-Oudshoorn, 2011), insofar as the respondents did not have more than two missing response items across all the model’s variables. All calculations were inspected visually, revealing no sign of misbehavior. Since we are very con?dent that these items are missing at random (see the description of the t-tests in the method chapter), we chose to use the imputed values for our main analysis, thus giving us a larger sample size and reducing the likelihood of accepting the model due to outliers (Hair et al., 2012). For the imputation, we used predictive mean matching (Little, 1988), using the model’s manifest variables as input. As a robustness test, we also tested the models with other imputation techniques, such as Gibbs sampling (Gelfand and Smith, 1990), obtaining similar results. The analyses were also performed on the dataset without any imputed values (list-wise exclusion). In so doing, our overall results were comparable with those obtained with the imputed values. 3.8. Outer model evaluation Before we begin our analysis of the relationships among search behaviors, learning, prior knowledge, and capabilities, we must 2 Surprisingly, cohesion and con?ict had very little variance across teams and high means within teams. This might be due to either social desirability bias or self-selection. 169 Table 1 Overview of hypotheses support. Hypotheses H1: H2a: H2b: H3a: H3b: Support? Prior rel.know. Exploi.behavior Explor.behavior Rout.deli.learn. Rout.deli.learn. → → → → → Rout.deli.learn. Rout.deli.learn. Rout.deli.learn. Exploi.capability Explor.capability Yes Yes Yes Yes Yes establish the validity of our measurement model. Note that our outer model is purely re?ective. First, preliminary examination of the outer loadings indicated internal reliability (Hair et al., 2012). The outer model’s composite reliability measures (Dillon-Goldstein’s q and Cronbach’s a) provide further evidence that this is the case (Hair et al., 2012) (Table 2). Second, convergent validity was assessed by the AVE measures, yielding evidence of su?cient convergent validity (Table 2). In short, the average AVE was 0.68, above the recommended 0.5 (Bagozzi and Yi, 1988). Third, discriminant validity was assessed through the Fornell-Larcker criterion (Hair et al., 2012; Fornell and Larcker, 1981), comparing the square root of the AVE of each latent variable with its inter-construct correlations. The test gave no reason for concern. Additionally, the lack of large cross-loadings (Table 4) provides further evidence that the outer model has su?cient discriminant validity. Fourth, we inspected the R2 of the latent measures. Since three of the measures are endogenous (routines for deliberate learning and the two latent variables for capabilities), only those are reported. Since the emergence of routines and capabilities is a complex phenomenon, we deemed the R2 s magnitude to be more than su?cient (Table 2). In summary, our model showed clear loadings from every manifest variable on their respective latent variable. Regarding the crossloadings, we observed a minor cross-loading between two items of exploitation behavior and exploitation capability. Since these concepts are clearly conceptually distinct, we believe that this is not a reason for concern. 3.9. Inner model evaluation We used bias-corrected accelerated (BCa) bootstrapping to both generate bootstrapped path coe?cients and con?dence intervals for the inner model (n = 5000 reported, also run with n = 200 with the same results).3 This allows the con?dence intervals to be asymmetrically distributed around the mean estimate from the full data, thereby improving estimation accuracy and statistical power (Hair et al., 2012). The intervals are reported in Table 5. (Note that since the bootstrapping procedure does not assume a symmetric sampling distribution, CI intervals based on a predetermined pvalue must be used to determine support or rejection of hypotheses; calculating the p-values post-bootstrapping from the table would lead to de?ating our p-values, incorrectly making all model relationships signi?cant (p < .05).) Direct and total effects are shown in Table 6. 4. Analyses Table 1 presents a concise list of the hypotheses and whether they were supported. Fig. 2 is also helpful. The model indicates that prior relevant knowledge stimulates routine emergence (b3.4, H1). Moreover, both exploration and exploitation behaviors have a positive association with the emergence of routines for deliberate learning, as indicated by the model (b2.4, H2a, and b1.4, H2b). Finally, our model provides evidence that using routines for deliberate learning for storing behavior-derived knowledge within the 3 For replication, seed-value of 1, R version 3.3.2, running on 64-bit Debian GNU/Linux. 170 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 Exploitation behavior β 1.5=0.25* β 1.4=0.23* β 4.5=0.26* Prior relevant knowled. β 3.4=0.24* β 2.4=0.20* Routines deliberate learning β 2.6=0.13 Exploration behavior Exploitation capability β 4.6=0.39* Exploration capability Fig. 2. Full model, ∗ p = .05. ?rm leads to the emergence of the corresponding capabilities (b4.5, H3a and b4.6, H3b). In short, we found support for all hypotheses; we also saw that the direct effect of behavior on capability emergence became nonsigni?cant for exploration behaviors, providing evidence for the need for routines for deliberate learning for the emergence of exploration capability. While this was not the case for exploitation behaviors— indicating that exploitation behaviors would develop organizational exploitation capabilities on their own—we did observe that the exploitation behaviors and routines for deliberate learning were more effective in combination than on their own. 5. Discussion and implications Building on the insights provided by Zollo and Winter (2002), empirical testing of our adapted model (Fig. 1) elucidates the emergence of exploration and exploitation capabilities in NTBFs. The model has good explanatory power, and all the hypotheses derived from the model received empirical support (Table 1). Hence, our model and choice of context provide much-needed insight (Helfat and Lieberman, 2002; Helfat and Peteraf, 2003; Felin and Foss, 2009) into the interplay between the origins of and mechanisms in?uencing the emergence of exploration and exploitation capabilities in NTBFs in particular. Our model considers prior knowledge of founder-managers and their exploration and exploitation behaviors during ?rm formation to be important origins for capability emergence in NTBFs, providing implications for their proclivity for exploration and exploitation. However, while these origins may be important sources for the emergence of the organizational capabilities of exploration and exploitation in NTBFs, their potency for doing so is most pronounced when interacted with organizational processes in the new ?rm, of which we have highlighted routines for deliberate learning as a key mechanism (Zollo and Winter, 2002). In short, our model and empirical results indicate that there is an effect of exploitation and exploration behavior on capability emergence and that this effect is mediated through routines for deliberate learning. This elaborates on prior research showing that learning-by-doing is a powerful explanation for the evolution of exploration and exploitation capability (see Helfat and Peteraf, 2003 for a discussion), and our research is not antithetical to their ?ndings. Instead, we propose that during the early stages of an NTBF’s life-cycle, the ?rm is lacking formal structures and routines that, in turn, may lead to loss of important knowledge gained from learning-by-doing. As such, there might be a more complex interplay between routinized deliberate learning and learning-by-doing whereby the ?rm needs to reach a threshold of formalization before learning-by-doing is maintained within the organization as capabilities. Our model and empirical results show that routines for deliberate learning are formed as a consequence of exploration and exploitation behaviors, indicating that founders are aware of the need to capture the knowledge that they create. When these behaviors are captured within capabilities, the focal behaviors are likely reinforced (Hannan and Freeman, 1984; March, 1991; Levinthal and March, 1993). Thus, the model may suggest the root of organizational inertia or path-dependency as emerging in NTBFs as a consequence of the need to retain knowledge within the organization as capabilities. While routines are often considered to be sources of e?ciency and discipline and may therefore reduce the exploratory capability of the ?rm (Bryant, 2014), we add to this knowledge by showing that in certain circumstances it seems like these routines can help develop NTBFs’ exploratory capability. While this is the case, these routines may have adverse future effects on capability (Bryant, 2014; Levinthal and Posen, 2007). Similar concerns are raised by Levinthal and Posen (2007) whose simulation study indicates that population selection is ine?cient due to non-optimal search strategies being chosen early on in the ?rm’s life-cycle. Further research is needed in this regard. Finally, we show that such routines are also reliant on some level of prior relevant knowledge. This indicates that familiarity with the problem domain increases appreciation for learning. It may also indicate that deep knowledge is tacit, requiring deliberate and conscious effort to spread with the organization and its members. Our theorizing, model, and empirical results illustrate routines for deliberate learning’s importance for encouraging the emergence of exploration and exploitation capability in NTBFs (Nonaka and Von Krogh, 2009). We show that relevant knowledge, both extant and developed through founder behavior during the early stages of a ?rm’s life-cycle, is not automatically imprinted in the capabilities of new ?rms. Deliberate learning needs to take place before prior relevant knowledge and search behaviors become manifest as exploration and exploitation capabilities. Hence, our research has elucidated the dynamics surrounding how organizations acquire their initial capacity to exploit old knowledge and explore new knowledge. Mastering these capabilities is closely tied to organizations’ ability to search (Nelson and Winter, 1982; Cyert and March, 1963), e.g., search for innovations (Katila et al., 2012; He and Wong, 2004; Katila and Ahuja, 2002). Indeed, in most studies, search consists of different constellations of the ?rm’s ability to explore and exploit knowledge (Katila and Ahuja, 2002; Katila et al., 2012). Despite the importance of knowledge search, most studies on the issue focus on mature organizations with existing search capabilities, thereby neglecting the issue of how knowledge search ?rst emerges within organizations. Our model and empirical results have addressed this issue in the context of NTBFs, a type of ?rm that are, by their nature, closely aligned with innovation search. Importantly, we show that ?rm heterogeneity in exploration and exploitation capabilities forms early in the life-cycle of new ventures. While founder-managers’ prior relevant knowledge and behavior are important, distinct learning A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 processes and routines initiated within the new ?rm are necessary for the emergence of ?rm heterogeneity in exploration and exploitation capabilities. This will likely affect the development of the ?rm in the long run. Further, our theorizing and results showcase the importance of organizational processes and mechanisms’ impact on the oftenposed direct relationships between characteristics of founders and the performance of their ?rms. As an example, our model and empirical results challenge the “simple” idea in imprinting theory that founder-managers directly imprint the capabilities of the ?rms they create (Marquis and Tilcsik, 2013). Our ?ndings instead reinforce the notion that imprinting effects can be managed and be a source for adaptation and development (Bryant, 2014b), at least in the early stages of the organization’s life-cycle. To management scholars, our results and model highlight that the capabilities that they theorize about, often assuming the existence of a mature organization, can trace their history back to the time when barely any organization was in place. Thus, by taking the pre-history of the ?rm into account, theorizing on the evolution of capability can be strengthened (Winter, 2012). While the behavior and prior knowledge of founder-managers are important antecedents of the emergence of capabilities within NTBFs, their in?uence requires routines for deliberate learning, particularly in the case of NTBFs’ exploration and exploitation capabilities. The historical argument may be particularly important for scholars of exploration and exploitation. Balancing exploration and exploitation within incumbent technology ?rms has proven to be challenging, which has implications for how (mature) organizations search for innovation (Levinthal and March, 1993; Raisch et al., 2009; Andriopoulos and Lewis, 2008). Uncovering the source of their initial divergence can prove fruitful for understanding why and how some ?rms tend to focus on exploitation, exploration, or, for a select few, both. For example, founder-managers who achieved success by searching broadly for knowledge through exploration and subsequently re?ning it through exploitation may prefer the incumbent ?rm to organize such activities sequentially (so-called “temporal ambidexterity” (Tushman and O’Reilly, 1996)). Those who succeeded by engaging in both simultaneously may prefer the incumbent ?rm to organize such activities at the same time (so-called “contextual ambidexterity” (Gibson and Birkinshaw, 2004)). Martini et al. (2015), Belderbos et al. (2010) and de Visser et al. (2010) provide evidence that the type of ambidexterity, e.g., the degree of separation between the two types of capabilities, has important performance implications. This further cements the importance of understanding the initial emergence of exploration and exploitation capabilities. Forecasting scholars have shown interest in how ?rms can develop such ambidexterity (in terms of strategy, not knowledge) (e.g., Bodwell and Chermack, 2010). By studying the phenomenon by starting from its emergence, their research would become even more powerful. Our model and research provide means for understanding this process for the emergence of exploitation and exploration capabilities in NTBFs and highlights the importance of establishing routines for deliberate learning for knowledge capture. Future research could examine how these routines can be crafted in such a way that they lead to the simultaneous emergence of exploration and exploitation capabilities, their emergence in turn, or even the failure of one or both to emerge. When exploring determinants of a ?rm’s ability to grow and generate revenue, entrepreneurship scholars often draw upon explanations of behavior and prior knowledge. We provide additional insights into how these two dimensions of new venturing affect the ?rm beyond early-stage performance measures. There is strong evidence for the ‘stickiness’ of capabilities and routines once they have become part of an organization (see, e.g., Levinthal and March, 1993; Levinthal and Posen, 2007; Sydow et al., 2009; Hannan and Freeman, 1984; Atuahene-Gima, 2005). Our model indicates 171 that entrepreneurship scholars can produce insights into how these routines and capabilities ?rst come to light. Unfortunately, entrepreneurship scholars tend to ignore routines as a concept, despite their usefulness for reducing the liability of newness (Aldrich and Yang, 2012; Stinchcombe, 1965). Recognizing that incumbent ?rms are likely to be affected by their historical development of routines and capabilities (Winter, 2012), we urge entrepreneurship scholars to reconsider routines as an avenue of research. While, e.g., routines for learning may be unable to provide short-term revenues or growth (Wilden et al., 2013), our model indicates that they have important consequences for the ?rm in the long run, at least if the NTBF’s early-stage capabilities become manifest in the organization. In sum, we have crafted a theory-derived model of how exploration and exploitation capabilities emerge and become manifest in NTBFs during the early stages of their life-cycle, highlighting the role of routines for deliberate learning. In doing so, we have combined insights and theorizing from the entrepreneurship and strategic management literatures. Capability emergence, particularly in new, but also incumbent, organizations, is a phenomenon that lies in a “gray area” domain between strategic management and entrepreneurship. Arguably, it has been largely unaddressed because neither literature has the theoretical apparatus to address it fully. Our model showcase the potency of combining and merging insights from these two bodies of literature for understanding key phenomena on the border between them. Further, in our theorizing on this issue, we have introduced concepts and vocabulary that have the potential to aid further theorizing. A distinct element of our model is the difference between origins and mechanisms in capability emergence. We argue that the nexus of prior knowledge and behavior undertaken by founder-managers during start-up constitutes the origins of capability emergence within NTBFs. However, these origins are not automatically translated into input for the emergence of capabilities in NTBFs in the absence of routines for deliberate learning. These routines acted as a catalyzing mechanism that “converted” search behaviors and prior relevant knowledge of founder-managers into search capabilities for exploration and exploitation. Our research strongly suggests that NTBFs with better-developed routines for deliberate learning will be more able to utilize and channel the two central assets that most NTBFs have—prior knowledge and behavior—as input for the development of organized innovation search, in our case exempli?ed by these new organizations’ proclivity for exploration and exploitation. Subsequent research should examine the in?uence of sources and mechanisms other than those investigated and theorized on in this paper. Future research should test the model using data from new ?rms in other types of settings. We have argued that the NTBF setting has important implications for the relationships in the model. Moreover, the research suffers from the cross-sectional nature of the data. On the one hand, a longitudinal research design would allow us to make causal claims about the model’s directional relationships. On the other hand, the lack of longitudinal data is probably less of a concern for explaining capability emergence since the time span from the organization’s birth to the emergence of its capabilities is short. Notwithstanding, detailed and “thick” accounts, for instance through observational data, of how capabilities for exploration and exploitation emerge in new ?rms would help to re?ne and extend the model, perhaps suggesting new relationships and new intervening mechanisms between founder behaviors, routines for deliberate learning, and capability emergence. Moreover, an additional avenue for future research would be to examine whether, and to what extent, the processes and mechanisms affecting the emergence of exploration and exploitation capabilities in an NTBF in?uence the subsequent evolution of these capabilities during the ?rm’s incumbency. 172 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 Appendix Table 2 Model diagnostics. Exploitation behavior Exploration behavior Prior rel. know. Rout. deliber. learn. Exploitation capability Exploration capability Cronbach’s a Dillon-Goldstein’s q No. manifests R2 R2 -corrected 0.74 0.77 0.87 0.70 0.71 0.82 0.84 0.86 0.91 0.84 0.83 0.89 3 3 4 3 3 3 . . . 0.17 0.16 0.19 . . . 0.14 0.14 0.17 Table 3 Manifest variables for the measurement model. All are Likert-type variables, 7-points. Exploration capability We systematically bring together creative and knowledgeable persons within the ?rm to identify new business opportunities. We systematically bring together creative and knowledgeable persons from outside the ?rm to help identify new business opportunities. Exploitation capability We systematically benchmark the ?rm with the best in the industry. We have developed routines to recon?gure our resources in new ways. Exploration behavior Evaluating diverse options with respect to products/services, processes, or markets. Undertaken activities requiring great adaptability. Exploitation behavior Undertaken activities which we knew well beforehand how to perform. Undertaken activities which were completed by using the ?rm’s existing knowledge. Routines for deliberate learning Discussions on the ?rm’s development are routinely systematized for review in the future. The experiences we have are stored and systematized in written form in spreadsheets, databases, or similar. Prior relevant knowledge We have experience in identifying technological changes and trends in the industry in which we now compete in. We have experience in new product development. We posses unique knowledge, enabling us to develop just as unique products. Item name Source We systematically search for new business concepts through observation of processes in the environment. Makkonen et al. (2014) We continuously work to take out e?ciency gains in the organization. Alsos et al. (2008) Searching for new possibilities with respect to products/services, processes, or markets. Mom et al. (2009, 2007) Undertaken activities we already have extensive experience with. Mom et al. (2009, 2007) The experiences we have are subject to frequent (repeating) discussion. Self-developed, based on Zollo and Winter (2002) We have experience in developing new products, services, or processes. Zhao et al. (2013) Table 4 Latent variable loadings > .5. exploit_b1 exploit_b2 exploit_b3 explore_b1 explore_b2 explore_b3 knowledge1 knowledge2 knowledge3 knowledge4 learning1 learning2 learning3 exploit_c1 exploit_c2 exploit_c3 explore_c1 explore_c2 explore_c3 Exploitation behavior Exploration behavior Prior relevant knowledge Routines delib. learn. Exploitation capability Exploration capability 0.89 0.79 0.71 . . . . . . . . . . . . . . . . . . . 0.88 0.88 0.71 . . . . . . . . . . . . . . . . . . . 0.81 0.83 0.92 0.83 . . . . . . . . . . . . . . . . . . . 0.75 0.87 0.77 . . . . . . . . . . . . . . . . . . . 0.80 0.65 0.91 . . . . . . . . . . . . . . . . . . . 0.81 0.94 0.83 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 173 Table 5 Bootstrapped coe?cients and con?dence intervals, p = .05, (p = .1). b1.4 b2.4 b3.4 b1.5 b4.5 b2.6 b4.6 Coe?cients Bias Std.error Lower Upper 0.23 0.20 0.24 0.25 0.26 0.13 0.39 0.01 0.01 0.04 0.02 0.01 0.01 0.01 0.10 0.08 0.07 0.10 0.11 0.10 0.10 0.01 (0.07) 0.04 (0.09) 0.14 (0.12) 0.02 (0.07) 0.01 (0.06) -0.10 (-0.07) 0.16 (0.20) 0.41 (0.38) 0.36 (0.33) 0.36 (0.33) 0.40 (0.38) 0.44 (0.41) 0.30 (0.28) 0.57 (0.53) Table 6 Top: Signi?cant, p < .05 (p < 0.1) direct paths (path coe?cients). Bottom: Total effects (path coe?cients and indirect effects) through all paths. Exploitation behavior Exploration behavior Prior rel. know. Rout. deliber. learn. Exploitation capability Exploration capability Exploitation behavior Exploration behavior Prior rel. know. Rout. deliber. learn. Exploitation capability Exploration capability Exploitation behavior Exploration behavior Prior rel.know. Rout. deliber. learn. Exploitation capability Exploration capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.23 0.20 0.24 . . . 0.23 0.25 0.24 . . . 0.25 . . (0.26) . . 0.30 0.05 0.06 0.26 . . . . . 0.38 . . 0.09 0.21 0.09 0.39 . . References Aldrich, H.E., Yang, T., 2012. Lost in translation: cultural codes are not blueprints. Strateg. Entrep. J. 6 (1), 1–17. ISSN 1932-443X. Aloini, D., Martini, A., 2012. Exploring the exploratory search for innovation: a structural equation modelling test for practices and performance. Int. J. Technol. Manag. 61 (1), 23–46. ISSN 0267-5730. Alsos, G., Borch, O., Ljunggren, E., Madsen, E., 2008. Dynamic Capabilities – Conceptualization and Operationalization, Anaheim, CA. Andries, P., Debackere, K., Van Looy, B., 2013. Simultaneous experimentation as a learning strategy: business model development under uncertainty. Strateg. Entrep. J. 7 (4), 288–310. ISSN 1932-4391. wOS:000328000700002. Andriopoulos, C., Lewis, M.W., 2008. Exploitation-exploration tensions and organizational ambidexterity: managing paradoxes of innovation. Organ. Sci. 20 (4), 696–717. ISSN 1047-7039. Argyres, N.S., Felin, T., Foss, N., Zenger, T., 2012. Organizational economics of capability and heterogeneity. Organ. Sci. 23 (5), 1213–1226. ISSN 1047-7039. Atuahene-Gima, K., 2005. Resolving the capability: rigidity paradox in new product innovation. J. Mark. 69 (4), 61–83. ISSN 0022-2429. Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. J. Acad. Mark. Sci. 16 (1), 74–94. ISSN 0092-0703, 1552-7824. Basu, S., Sahaym, A., Howard, M.D., Boeker, W., 2015. Parent inheritance, founder expertise, and venture strategy: determinants of new venture knowledge impact. J. Bus. Ventur. 30 (2), 322–337. ISSN 0883-9026. 81d7fc8c-6a12-47be-ac96-16d802934bf5. Baum, J.A.C., Cowan, R., Jonard, N., 2014. Does evidence of network effects on ?rm performance in pooled cross-section support prescriptions for network strategy? Strateg. Manag. J. 35 (5), 652–667. ISSN 0143-2095. wOS:000334183900002. Becker, M.C., Knudsen, T., 2005. The role of routines in reducing pervasive uncertainty. J. Bus. Res. 58 (6), 746–757. ISSN 0148-2963. Belderbos, R., Faems, D., Leten, B., Looy, B.V., 2010. Technological activities and their impact on the ?nancial performance of the ?rm: exploitation and exploration within and between ?rms*. J. Prod. Innov. Manag. 27 (6), 869–882. ISSN 1540-5885. Benner, M.J., Tushman, M., 2002. Process management and technological innovation: a longitudinal study of the photography and paint industries. Adm. Sci. Q. 47 (4), 676–707. ISSN 0001-8392. Berghman, L., Matthyssens, P., Streukens, S., Vandenbempt, K., 2013. Deliberate learning mechanisms for stimulating strategic innovation capacity. Long Range Plan. 46 (1–2), 39–71. ISSN 0024-6301. Betsch, T., Haberstroh, S., Glöckner, A., Haar, T., Fiedler, K., 2001. The effects of routine strength on adaptation and information search in recurrent decision making. Organ. Behav. Hum. Decis. Process. 84 (1), 23–53. ISSN 0749-5978. Bigley, G.A., Roberts, K.H., 2001. The incident command system: high-reliability organizing for complex and volatile task environments. Acad. Manag. J. 44 (6), 1281–1299. ISSN 0001-4273 1948-0989. Bodwell, W., Chermack, T.J., 2010. Organizational ambidexterity: integrating deliberate and emergent strategy with scenario planning. Technol. Forecast. Soc. Chang. 77 (2), 193–202. ISSN 0040-1625. Bollinger, L., Hope, K., Utterback, J.M., 1983. A review of literature and hypotheses on new technology-based ?rms. Res. Policy 12 (1), 1–14. ISSN 0048-7333. Bresman, H., 2013. Changing routines: a process model of vicarious group learning in pharmaceutical R&D. Acad. Manag. J. 56 (1), 35–61. ISSN 0001-4273 1948-0989. Bryant, P.T., 2014. Self-Regulation and Entrepreneurial Ambidexterity. Exploration and Exploitation in Early Stage Ventures and SMEs, vol. 14 of Technology, Innovation, Entrepreneurship and Competitive Strategy. vol. 14. Emerald Group Publishing Limited., pp. 15–37. DOI: 10.1108/S1479-067X20140000014001. http://dx.doi. org/10.1108/S1479-067X20140000014001. Bryant, P.T., 2014b. Imprinting by design: the microfoundations of entrepreneurial adaptation. Enterp. Theory Pract. 38 (5), 1081–1102. ISSN 1540-6520. Buuren, S. v., Groothuis-Oudshoorn, K., 2011. mice: multivariate imputation by chained equations in R. J. Stat. Softw. 45 (3), 1–67. Camps, S., Marques, P., 2014. Exploring how social capital facilitates innovation: the role of innovation enablers. Technol. Forecast. Soc. Chang. 88, 325–348. ISSN 0040-1625. Ciccotello, C.S., Conrad, E.J., Fekula, M.J., Grant, C.T., 2000. Financial decisions and top management team composition. Adv. Account. 17, 91–109. ISSN 0882-6110. Cohen, W.M., Klepper, S., 1992. The tradeoff between ?rm size and diversity in the pursuit of technological progress. Small Bus. Econ. 4 (1), 1–14. ISSN 0921-898X. 1573-0913. Cohendet, P., Llerena, P., 2003. Routines and incentives: the role of communities in the ?rm. Ind. Corp. Chang. 12 (2), 271–297. ISSN 0960-6491. wOS:000182646700006. Cohen, W.M., Malerba, F., 2001. Is the tendency to variation a chief cause of progress? Ind. Corp. Chang. 10 (3), 587–608. ISSN 0960-6491. Cyert, R.M., March, J. G., A behavioral theory of the ?rm, Englewood Cliffs, NJ 2. DeSarbo, W.S., Anthony Di Benedetto, C., Song, Michael, Sinha, I., 2005. Revisiting the miles and snow strategic framework: uncovering interrelationships between strategic types, capabilities, environmental uncertainty, and ?rm performance. Strateg. Manag. J. 26 (1), 47–74. ISSN 1097-0266. de Visser, M., de Weerd-Nederhof, P., Faems, D., Song, M., van Looy, B., Visscher, K., 2010. Structural ambidexterity in NPD processes: a ?rm-level assessment of the impact of differentiated structures on innovation performance. Technovation 30 (5–6), 291–299. ISSN 0166-4972. Dingler, A., Enkel, E., 2016. Socialization and innovation: insights from collaboration across industry boundaries. Technol. Forecast. Soc. Chang. 109, 50–60. ISSN 0040-1625. Edmondson, A., 1999. Psychological safety and learning behavior in work teams. Adm. Sci. Q. 44 (2), 350–383. Edmondson, A.C., 2003. Speaking up in the operating room: how team leaders promote learning in interdisciplinary action teams. J. Manag. Stud. 40 (6), 1419–1452. ISSN 0022-2380. wOS:000184934200004. Eisenhardt, K.M., Furr, N.R., Bingham, C.B., 2010. CROSSROADS—Microfoundations of performance: balancing e?ciency and ?exibility in dynamic environments. Organ. Sci. 21 (6), 1263–1273. ISSN 1047-7039. Evangelista, F., Mac, L., 2016. The in?uence of experience and deliberate learning on SME export performance. Int. J. Entrep. Behav. Res. 22 (6), 860–879. ISSN 1355-2554. Feldman, M.S., Pentland, B.T., 2003. Reconceptualizing organizational routines as a source of ?exibility and change. Adm. Sci. Q. 48 (1), 94–118. ISSN 0001-8392, 1930-3815. Felin, T., Foss, N.J., 2009. Organizational routines and capabilities: historical drift and a course-correction toward microfoundations. Scand. J. Manag. 25 (2), 157–167. 174 A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 18 (1), 39–50. ISSN 0022-2437. Frigotto, M.L., Coller, G., Collini, P., 2014. Exploration and Exploitation from Start-up to Sale: A Longitudinal Analysis through Strategy and MCS Practices. Exploration and Exploitation in Early Stage Ventures and SMEs, vol. 14 of Technology, Innovation, Entrepreneurship and Competitive Strategy. vol. 14. Emerald Group Publishing Limited., pp. 149–179. 10.1108/S1479-067X20140000014004 DOI: 10.1108/S1479-067X20140000014004. Gartner, W., 1988. ”Who is an Entrepreneur?” Is the Wrong Question, Entrepreneurship Theory and Practice. Entrep. Theory Pract. 12 (4), 11–32. ISSN 0022-2437. Gelfand, A.E., Smith, A.F.M., 1990. Sampling-based approaches to calculating marginal densities. J. Am. Stat. Assoc. 85 (410), 398–409. ISSN 0162-1459. Gibson, C.B., Birkinshaw, J., 2004. The antecedents, consequences, and mediating role of organizational ambidexterity. Acad. Manag. J. 47 (2), 209–226. ISSN 0001-4273. wOS:000221772600003. Grant, R.M., 1996a. Prospering in dynamically-competitive environments: organizational capability as knowledge integration. Organ. Sci. 7 (4), 375–387. ISSN 1047-7039. Grant, R.M., 1996b. Toward a knowledge-based theory of the ?rm. Strateg. Manag. J. 17 (S2), 109–122. ISSN 1097-0266. Greve, H.R., 2007. ‘Exploration and exploitation in product innovation’. Ind. Corp. Chang. 16 (5), 945–975. ISSN 0960-6491, 1464-3650. Guenther, C., Oertel, S., Walgenbach, P., 2016. It’s all about timing: age–dependent consequences of founder exits and new member additions. Enterp. Theory Pract. 40, 843–865. ISSN 10422587. http://dx.doi.org/10.1111/etap.12148. Gupta, A.K., Smith, K.G., Shalley, C.E., 2006. The interplay between exploration and exploitation. Acad. Manag. J. 49 (4), 693–706. ISSN 0001-4273. 1948-0989. Hair, J.F., Sarstedt, M., Pieper, T.M., Ringle, C.M., 2012. The use of partial least squares structural equation modeling in strategic management research: a review of past practices and recommendations for future applications. Long Range Plan. 45 (5–6), 320–340. ISSN 0024-6301. Hannan, M.T., Freeman, J., 1984. Structural inertia and organizational change. Am. Sociol. Rev. 49 (2), 149–164. ISSN 0003-1224. wOS:A1984ANQ8200002. He, Z.-L., Wong, P.-K., 2004. Exploration vs. exploitation: an empirical test of the ambidexterity hypothesis. Organ. Sci. 15 (4), 481–494. ISSN 1047-7039. Helfat, C.E., Lieberman, M.B., 2002. The birth of capabilities: market entry and the importance of pre-history. Ind. Corp. Chang. 11 (4), 725–760. ISSN 0960-6491. 1464-3650. Helfat, C.E., Peteraf, M.A., 2003. The dynamic resource-based view: capability lifecycles. Strateg. Manag. J. 24 (10), 997–1010. ISSN 1097-0266. Ho, J.C., Lee, C.-S., 2015. A typology of technological change: technological paradigm theory with validation and generalization from case studies. Tech. Forcasting Soc. Chang. 97, 128–139. ISSN 0040-1625. Hoang, H., Rothaermel, F.T., 2010. Leveraging internal and external experience: exploration, exploitation, and R&D project performance. Strateg. Manag. J. 31 (7), 734–758. ISSN 1097-0266. Jansen, J.J.P., Van Den Bosch, F.A.J., Volberda, H.W., 2006. Exploratory innovation, exploitative innovation, and performance: effects of organizational antecedents and environmental moderators. Manag. Sci. 52 (11), 1661–1674. ISSN 0025-1909. Jansen, J.J.P., Tempelaar, M.P., van den Bosch, F.A.J., Volberda, H.W., 2009. Structural differentiation and ambidexterity: the mediating role of integration mechanisms. Organ. Sci. 20 (4), 797–811. ISSN 1047-7039. wOS:000267982000007. Katila, R., Ahuja, G., 2002. Something old, something new: a longitudinal study of search behavior and new product introduction. Acad. Manag. J. 45 (6), 1183–1194. ISSN 0001-4273. 1948-0989. Katila, R., Chen, E.L., Piezunka, H., 2012. All the right moves: how entrepreneurial ?rms compete effectively. Strateg. Entrep. J. 6 (2), 116–132. ISSN 1932-4391. wOS:000304905900003. Katz, J., Gartner, W.B., 1988. Properties of emerging organizations. Acad. Manag. Rev. 13 (3), 429–441. ISSN 0363-7425, 1930-3807. Lavie, D., Stettner, U., Tushman, M.L., 2010. Exploration and exploitation within and across Organizations. Acad. Manag. Ann. 4 (1), 109–155. ISSN 1941-6520. Levinthal, D.A., March, J.G., 1993. The myopia of learning. Strateg. Manag. J. 14 (S2), 95–112. ISSN 1097-0266. Levinthal, D., Posen, H.E., 2007. Myopia of selection: does organizational adaptation limit the e?cacy of population selection? Adm. Sci. Q. 52 (4), 586–620. ISSN 0001-8392, 1930-3815. Levinthal, D., Rerup, C., 2006. Crossing an apparent chasm: bridging mindful and less– mindful perspectives on organizational learning. Organ. Sci. 17 (4), 502–513. ISSN 1047-7039. Lin, H.-E., McDonough, E.F., Lin, S.-J., Lin, C.Y.-Y., 2013. Managing the exploitation/exploration paradox: the role of a learning capability and innovation ambidexterity. J. Prod. Innov. Manag. 30 (2), 262–278. ISSN 1540-5885. Little, R.J.A., 1988. Missing-data adjustments in large surveys. J. Bus. Econ. Stat. 6 (3), 287–296. ISSN 0735-0015. Makkonen, H., Pohjola, M., Olkkonen, R., Koponen, A., 2014. Dynamic capabilities and ?rm performance in a ?nancial crisis. J. Bus. Res. 67 (1), 2707–2719. ISSN 0148-2963. March, J.G., 1991. Exploration and exploitation in organizational learning. Organ. Sci. 2 (1), 71–87. ISSN 1047-7039. March, J., Olsen, J., 2013-09-05. The Logic of Appropriateness. Oxford Handbooks Online, Retrieved 30 Apr. 2017, from: http://www.oxfordhandbooks.com/view/ 10.1093/oxfordhb/9780199604456.001.0001/oxfordhb-9780199604456-e-024. Marquis, C., Tilcsik, A., 2013. Imprinting: toward a multilevel theory. Acad. Manag. Ann. 7 (1), 195–245. ISSN 1941-6520. Martin, J.A., Eisenhardt, K.M., 2010. Rewiring: cross-business-unit collaborations in multibusiness organizations. Acad. Manag. J. 53 (2), 265–301. ISSN 0001-4273 1948-0989. Martini, A., Neirotti, P., Aloini, D., 2015. Finding the way to ambidexterity: exploring the relationships among organisational design, knowledge creation and innovation. Int. J. Innov. Manag. 19 (04), 1550045. ISSN 1363-9196. Miller, K.D., Choi, S., Pentland, B.T., 2014. The role of transactive memory in the formation of organizational routines. Strateg. Organ. 12 (2), 109–133. ISSN 1476-1270. Mom, T.J.M., van den Bosch, F.A.J., Volberda, H.W., 2009. Understanding variation in managers’ ambidexterity: investigating direct and interaction effects of formal structural and personal coordination mechanisms. Organ. Sci. 20 (4), 812–828. ISSN 1047-7039. wOS:000267982000008. Mom, T.J.M., Van Den Bosch, F.A.J., Volberda, H.W., 2007. Investigating managers’ exploration and exploitation activities: the in?uence of top-down, bottom-up, and horizontal knowledge in?ows*. J. Manag. Stud. 44 (6), 910–931. ISSN 1467-6486. Monecke, A., Leisch, F., 2012. semPLS: Structural equation modeling using partial least squares. J. Stat. Softw. 48 (3), 1–32. Murmann, J.P., 2012. The coevolution of industries and important features of their environments. Organ. Sci. 24 (1), 58–78. ISSN 1047-7039. Nelson, R.R., Winter, S.G., 1982. An Evolutionary Theory of Economic Change. Harvard University Press. Nembhard, I.M., Tucker, A.L., 2010. Deliberate learning to improve performance in dynamic service settings: evidence from hospital intensive care units. Organ. Sci. 22 (4), 907–922. ISSN 1047-7039. Nielsen, B.B., Gudergan, S., 2012. Exploration and exploitation ?t and performance in international strategic alliances. Int. Bus. Rev. 21 (4), 558–574. ISSN 0969-5931. Nonaka, I., 1991. The knowledge-creating company. Harv. Bus. Rev. 69 (6), 96–104. Nonaka, I., 1994. A dynamic theory of organizational knowledge creation. Organ. Sci. 5 (1), 14–37. ISSN 1047-7039. Nonaka, I., Von Krogh, G., 2009. Perspective-tacit knowledge and knowledge conversion: controversy and advancement in organizational knowledge creation theory. Organ. Sci. 20 (3), 635–652. O’Reilly, C.A., III, Tushman, M.L., 2008. Ambidexterity as a dynamic capability: resolving the innovator’s dilemma. Res. Organ. Behav. 28, 185–206. ISSN 0191-3085. Politis, D., 2005. The process of entrepreneurial learning: a conceptual framework. Enterp. Theory Pract. 29 (4), 399–424. ISSN 1540-6520. c0a3a0ed-c468-4e40a0a8-b41385acd853. Raisch, S., Birkinshaw, J., Probst, G., Tushman, M.L., 2009. Organizational ambidexterity: balancing exploitation and exploration for sustained performance. Organ. Sci. 20 (4), 685–695. ISSN 1047-7039. wOS:000267982000001. Romme, A.G.L., Zollo, M., Berends, P., 2010. Dynamic capabilities, deliberate learning and environmental dynamism: a simulation model. Ind. Corp. Chang. 19 (4), 1271–1299. ISSN 0960-6491. R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Santos, F.M., 2003. The coevolution of ?rms and their knowledge environment: insights from the pharmaceutical industry. Tech. Forcasting Soc. Chang. 70 (7), 687–715. ISSN 0040-1625. Schneckenberg, D., Truong, Y., Mazloomi, H., 2015. Microfoundations of innovative capabilities: the leverage of collaborative technologies on organizational learning and knowledge management in a multinational corporation. Tech. Forcasting Soc. Chang. 100, 356–368. ISSN 0040-1625. Schulz, M., 2001. The uncertain relevance of newness: organizational learning and knowledge ?ows. Acad. Manag. J. 44 (4), 661–681. ISSN 0001-4273. 1948-0989. Shin, D.-H., Choi, M.J., 2014. Ambidextrous information search: linking personal and impersonal search routines with individual performance. Inf. Technol. Manag. 15 (4), 291–304. ISSN 1385-951X, 1573-7667. Sirén, C.A., Kohtamäki, M., Kuckertz, A., 2012. Exploration and exploitation strategies, pro?t performance, and the mediating role of strategic learning: escaping the exploitation trap. Strateg. Entrep. J. 6 (1), 18–41. ISSN 1932-443X. Smith, K.G., Collins, C.J., Clark, K.D., 2005. Existing knowledge, knowledge creation capability, and the rate of new product introduction in high-t ?rms. Acad. Manag. J. 48 (2), 346–357. ISSN 0001-4273 1948-0989. Stettner, U., Aharonson, B.S., Amburgey, T.L., 2014. The Interplay Between Exploration and Exploitation in SMEs. Exploration and Exploitation in Early Stage Ventures and SMEs vol. 14 of Technology, Innovation, Entrepreneurship and Competitive Strategy. vol. 14. Emerald Group Publishing Limited., pp. 3–13. DOI: 10.1108/S1479-067X20140000014020. http://dx.doi.org/10.1108/S1479067X20140000014020. Stinchcombe, A.L., 1965. Social Structure and Organizations. Handbook of Organizations. pp. 142–193. Sydow, J., Schreyögg, G., Koch, J., 2009. Organizational path dependence: opening the black box. Acad. Manag. Rev. 34 (4), 689–709. ISSN 0363-7425, 1930-3807. Tippmann, E., Mangematin, V., Scott, P.S., 2013. The two faces of knowledge search: new solutions and capability development. Organ. Stud. 34 (12), 1869–1901. ISSN 0170-8406. Tu, C., 2010. Balancing exploration and exploitation capabilities in high technology ?rms: a multi-source multi-context examination. Ind. Mark. Manag. 39 (4), 672–680. ISSN 0019-8501. Tushman, M.L., O’Reilly, C.A., 1996. The ambidextrous organizations: managing evolutionary and revolutionary change. Calif. Manag. Rev. 38 (4), 8–30. ISSN 0008-1256, 2162-8564. Uotila, J., Maula, M., Keil, T., Zahra, S.A., 2009. Exploration, exploitation, and ?nancial performance: analysis of S&P 500 corporations. Strateg. Manag. J. 30 (2), 221–231. ISSN 1097-0266. A. Jensen, T. Clausen / Technological Forecasting & Social Change 120 (2017) 163–175 Wang, C.-H., Hsu, L.-C., 2014. Building exploration and exploitation in the high-tech industry: The role of relationship learning. Technol. Forecast. Soc. Chang. 81, 331–340. ISSN 0040-1625. Weick, K.E., Roberts, K.H., 1993. Collective mind in organizations: heedful interrelating on ?ight decks. Adm. Sci. Q. 38 (3), 357–381. ISSN 0001-8392. Wilden, R., Gudergan, S.P., Nielsen, B.B., Lings, I., 2013. Dynamic capabilities and performance: strategy, structure and environment. Long Range Plan. 46 (1–2), 72–96. ISSN 0024-6301. Winter, S.G., 2000. The satis?cing principle in capability learning. Strateg. Manag. J. 21 (10-11), 981–996. ISSN 0143-2095. wOS:000165188900003. Winter, S., 2012. Capabilities: their origins and ancestry. J. Manag. Stud. 49 (8), 1402–1406. ISSN 1467-6486. Yalcinkaya, G., Calantone, R.J., Gri?th, D.A., 2007. An examination of exploration and exploitation capabilities: implications for product innovation and market performance. J. Int. Mark. 15 (4), 63–93. ISSN 1069-031X. Yamakawa, Y., Yang, H., Lin, Z.J., 2011. Exploration versus exploitation in alliance portfolio: performance implications of organizational, strategic, and environmental ?t. Res. Policy 40 (2), 287–296. ISSN 0048-7333. Yi, S., Knudsen, T., Becker, M.C., 2016. Inertia in routines: a hidden source of organizational variation. Organ. Sci. 27 (3), 782–800. ISSN 1047-7039. Zahra, S.A., Sapienza, H.J., Davidsson, P., 2006. Entrepreneurship and dynamic capabilities: a review, model and research agenda*. J. Manag. Stud. 43 (4), 917–955. ISSN 1467-6486. 175 Zhao, Y.L., Song, M., Storm, G.L., 2013. Founding team capabilities and new venture performance: the mediating role of strategic positional advantages. Enterp. Theory Pract. 37 (4), 789–814. ISSN 1540-6520. Zollo, M., Winter, S.G., 2002. Deliberate learning and the evolution of dynamic capabilities. Organ. Sci. 13 (3), 339–351. ISSN 1047-7039. wOS:000175510900009. Are Jensen is a Ph.D. candidate at Nord University Business School (Norway). He holds a MSc. in Management with a specialization in management accounting and control. Research interests include new technology based ?rms, technology entrepreneurship, incubators, group dynamics, and decision making. Besides researching entrepreneurs, Are used to be one. This helps him understand the di?culties founders are faced with on a daily basis. Tommy Høyvarde Clausen is a Professor of Entrepreneurship at Nord University Business School (Norway). He is also an Adjunct Professor at the Centre for Entrepreneurship, University of Oslo (Norway). He has a PhD in Innovation Studies from the University of Oslo. Research interests include new technology based ?rms, incubators, innovation strategy, entrepreneurial opportunities, imprinting, and origins & evolution of ?rm heterogeneity. He has published in journals such as Research Policy, Industrial & Corporate Change, Technovation and Journal of Technology Transfer. The current issue and full text archive of this journal is available on Emerald Insight at: www.emeraldinsight.com/0025-1747.htm MD 53,2 Drivers of technology commercialization and performance in SMEs 338 The moderating effect of environmental dynamism Received 17 March 2014 Revised 9 July 2014 Accepted 14 November 2014 Taekyung Park Department of Business, Yeungnam University, Gyeongsan, Republic of Korea, and Dongwoo Ryu Department of Management, Yeungnam University, Gyongsan, Republic of Korea Abstract Purpose – The purpose of this paper is to explore the effects of small- and medium-sized enterprises (SMEs’) R&D capability and learning capability on their technology commercialization by focussing on the moderating effect of environmental dynamism. Design/methodology/approach – Based on a review of the literature on organizational capability, technology commercialization, and environmental dynamism, various hypotheses were developed and tested using a sample of 179 SMEs in Korea. Non-response bias using t-test and common method bias was assessed. Findings – The results indicate that their R&D capability and learning capability were significant drivers of their technology commercialization, which in turn influenced their business performance. Environmental dynamism was found to moderate the relationship between technology commercialization and business outcomes. These results suggest that SME managers should place greater emphasis on strengthening their organizational capability and dealing with turbulent business environments. Originality/value – Few studies have explored the drivers of technology commercialization and their effects on business performance. To fill this gap in the literature, the present study examines the effects of firms’ R&D capability and learning capability on technology commercialization in the context of SMEs, focussing specifically on the moderating effect of environmental dynamism. The study contributes to the literature by extending the research horizon to firms’ technology commercialization capability, providing a better understanding of the pivotal role of technology commercialization and its key drivers and environmental factors in boosting performance. Keywords Management strategy, Business environment Paper type Research paper Management Decision Vol. 53 No. 2, 2015 pp. 338-353 © Emerald Group Publishing Limited 0025-1747 DOI 10.1108/MD-03-2014-0143 1. Introduction In a fast-changing business environment, an organization’s ability to produce differentiated products or services becomes one of the most important sources of its competitive advantage (Diericks and Cool, 1989; Barney, 1991). This intangible asset allows a vast majority of small- and medium-sized enterprises (SMEs) with resources to compete in an effective manner. Because of this potential to allow SMEs to be competitive, more SMEs are emphasizing their organizational capability. In essence, this capability accumulates within the firm (Barney and Hesterly, 2006; Day and Wensley, 1988). According to the resource-based view, a firm’s organizational resources are focussed on its internal capability, which determines the extent to which a firm sustains its competitive advantage (Diericks and Cool, 1989; Grant, 1991). For firms to be competitive, accumulated resources must be valuable, rare, and difficult to imitate (Barney, 1991). Growing attention has been paid to a SMEs’ technology commercialization capability, which facilitates the marketing of technologies (Zahra and Nielsen, 2002; Ho et al., 2011; Park and Rhee, 2013). Technology commercialization involves a series of processes in which ideas are acquired and extended to knowledge for the development, manufacturing, and marketing of products (Mitchell and Singh, 1996) and is taken into account as a follow-up process that transforms various types of technology assets such as patents, designs, and know-how into profits. From this perspective, technology commercialization is likely an overarching concept covering the whole process of the product concept (product definition, design, sample, and pre-validation processes), including effective production and marketing. In this regard, successful technology commercialization can serve as an important capability that can satisfy consumers based on various cost-, speed-, quality-, and technology-related factors (Zahra and Nielsen, 2002). Previous studies have explored the antecedents of technology commercialization, including the entrepreneurial culture (Conceicao et al., 2002); manufacturing capability (Zahra and Nielsen, 2002); innovation characteristics (Rogers, 2003); resources and innovation capability (Chen, 2009); and networks (Park and Rhee, 2013). However, ...