What Takes the Wind out of Their Sails? A Micro-Foundational Perspective of Challenges for Building Dynamic Capabilities Towards Digital Business Model Innovation

Digital transformation has been identified as a driver and an enabler for new forms of BMI (Sorescu 2017; Vial 2019). This can be attributed to a large amount of data generation by companies, products, and customers (Vial 2019). Through their interconnection via the Internet of Things (IoT), those can be leveraged by, e.g., Artificial Intelligence (AI) solutions (Sjödin et al. 2021) and big data analytics capabilities (Ferraris et al. 2019; Mariani and Wamba 2020). Besides technological foundations of digital transformation and enablers of BMI, structural changes, organizational barriers and adequate strategic responses of companies undergoing digital transformation are required (Vial 2019).

One of the major trends of digital BMI can be attributed to service-driven business models (Müller and Buliga 2019). Servitization describes a form of BMI that complements or even replaces physical products as value offers (Linde et al. 2023, Paiola and Gebauer 2020). In this context, generated data can be used to create new value-adding solutions and data-driven services that are tailored to customer demands (Günther et al. 2017; Kaiser et al. 2021). Further, service-based business models monetize data by selling customer-tailored offers like implementation, consulting, or optimization. Additionally, data or insights gained can be monetized by selling it to third parties (Loebbecke and Picot 2015; Müller and Buliga 2019). In this context, software as a service (SaaS) is an approach that is based on Cloud Computing, offering a certain software flexibly via the Cloud. Often, this includes a revenue model where the customer only pays per usage when a certain software is required rather than selling it to customers (Susarla et al. 2009).

Similarly, the IoT plays a crucial role offering significant potential to create new business models. Through company-spanning interconnection of products, humans and customers, large amounts of data can be generated, transmitted, and analyzed to create value for several stakeholders (Langley et al. 2021; Laudien and Daxböck 2016; Leminen et al. 2020; Paiola and Gebauer 2020). In this context, digital platforms are viewed as a facilitator of IoT-based BMI by directly connecting multiple stakeholders that previously relied more on peer-to-peer data exchange. Further, so-called multi-sided platforms enable the integration of various groups of providers and customers, such as connecting both while selling analytics data to third parties (Tian et al. 2021; Veile et al. 2022).

Extant research has analyzed several forms of antecedents for enabling digital BMI, such as the role of digital and organizational capabilities (Müller et al. 2021; Matarazzo et al. 2021; Muhic and Bengtsson 2021; Soluk et al. 2021; Warner and Wäger 2019; Witschel et al. 2019, 2022), cultural issues (Tronvoll et al. 2020) such as entrepreneurial orientation (Ciampi et al. 2021), or the role of strategy (Brenk et al. 2019; Müller et al. 2021; Soto Setzke et al. 2021; Warner and Wäger 2019).

In addition, several technological, organizational, and environmental challenges, barriers and risks for implementing digital BMI have been examined (Brillinger 2018; Favoretto et al. 2021; Kiel et al. 2017; Linde et al. 2021; Muhic and Bengtsson 2021; Rachinger et al. 2019).

Likewise, extant literature investigates the mechanisms of the digital BMI process, i.e., its stages and progressions (Garzoni et al. 2020; Linde et al. 2023; Rong et al. 2018; Tesch et al. 2017), organizational processes (Garzella et al. 2021; Kamalaldin et al. 2020; Latilla et al. 2020; Warner and Wäger 2019; Witschel et al. 2019), relationships (Linde et al. 2021, 2023; Paiola and Gebauer 2020; Rachinger et al. 2019), and supporting tools or frameworks (Kaiser et al. 2021; Linde et al. 2021, 2023).

Nevertheless, the understanding of how companies approach digital BMI is still nascent and fragmented (Caputo et al. 2021; Hausberg et al. 2019). While there are several publications regarding digital skills and organizational capabilities (for an overview, see Bresciani et al. 2021; Konopik et al. 2022; Verhoef et al. 2021), the DCs perspective on digital BMI could still be extended (Bresciani et al. 2021; Vial 2019; Warner and Wäger 2019; Witschel et al. 2019, 2022), as described in more detail in the following two subsections.

Teece et al. (1997) define DCs as a “firm’s ability to integrate, build, and reconfigure internal and external competences to address rapidly changing environments. DCs thus reflect an organization’s ability to achieve new and innovative forms of competitive advantage given path dependencies and market positions” (p. 516). DCs are a result of an organizational learning process and are non-imitable, idiosyncratic, and company specific. Strong DCs enable a firm not only to adapt to changes in their business environments but also to shape them actively (Schoemaker et al. 2018). Teece (2007) suggests a threefold processual conceptualization, i.e., the ability of firms to sense and shape new digital business opportunities and threats, seize these digital opportunities, and maintain competitiveness by enhancing, combining, protecting, and reconfiguring a firm’s tangible and intangible assets. In short, DCs can be seen as essential to avoid disruption in changing environments (Schoemaker et al. 2018) on strategic and performance levels (Scheuer and Thaler 2022; Witschel et al. 2022). DCs can be traced back to the behavioral theory of the firm (Cyert and March 1963), suggesting that firms search for superior alternatives for their required performance criteria and trying to implement them in their organizations (Pandza and Thorpe 2009).

DCs have received attention in the field of BMI (e.g., Achtenhagen et al. 2013; De Silva et al. 2021; Heider et al. 2021; Inigo et al. 2017; Leih et al. 2015; Loon et al. 2020; Mezger 2014; Weimann et al. 2020). Likewise, the role of DCs in digital transformation has been investigated by several studies (e.g., Akter et al. 2020; Cannas 2021; Ellström et al. 2022; Ghosh et al. 2022; Magistretti et al. 2021; Sousa-Zomer et al. 2020; Tortora et al. 2021; Vrontis et al. 2020). Combining both research streams, studies acknowledging the role of DCs in digital BMI have started to evolve (Matarazzo et al. 2021; Muhic and Bengtsson 2021; Soluk et al. 2021; Warner and Wäger 2019; Witschel et al. 2019, 2022).

Hanelt et al. (2021) describe DCs as a mechanism to enable BMI as an outcome of digital transformation. Further, they particularly highlight the interrelations between both dimensions alongside contextual conditions resulting from digital transformation. Karimi and Walter (2015) further highlight that DCs and capabilities related to digital platforms are interrelated, as investigated in three out of six cases in this study.

Nevertheless, recent reviews state a limited understanding of how organizations build DCs (Konopik et al. 2022; Kraus et al. 2022; Steininger et al. 2021; Verhoef et al. 2021). In particular, most studies primarily describe DCs for digital BMI as a subdimension and do not offer a comprehensive view of underlying mechanisms, but rather on an aggregate level (Rachinger et al. 2019; Muhic and Bengtsson 2021). Examples include the enabling role of specific DCs such as knowledge exploitation, risk management, and marketing capabilities (Soluk et al. 2021) or big data analytics as a subtype of DCs (Ciampi et al. 2021).

Despite emerging findings that strong DCs are essential for digital BMI (Soluk et al. 2021; Witschel et al. 2022), DCs remain criticized in current literature as an overly abstract and highly aggregated analytical construct. It lacks explanatory power and insufficient consideration of the lower levels of DCs, the so-called micro-foundations (Arndt et al. 2022; Bojesson and Fundin 2021; Felin et al. 2012, 2015; Teece 2007). Following Felin et al. (2012) these micro-foundations encompass individuals, processes, interactions and structures, and determine how DCs are developed, manifested and implemented within an organization. The individual level relates to individual skills, cognition, knowledge, or experience and predict individual or collective behavior and actions. Processes and interactions determine how integration, cooperation, and coordination occur, while structures specify the conditions that hinder or facilitate information and knowledge processes, knowledge building, coordination, and individual or collective actions (Bendig et al. 2018; Felin et al. 2012, 2015; Teece 2007). In a nutshell, DCs are described as a “multilevel phenomenon” (Wilden et al. 2016, p. 1027), and thus require an analysis of the role of these subordinate micro-foundations and their interactions to increase the theoretical understanding of the development of DCs (Arndt et al. 2022; Bojesson and Fundin 2021; Foss and Pederson 2016; Magistretti et al. 2021; Santa-Maria et al. 2022; Schilke et al. 2018; Vial 2019). This enables to provide deeper insights for practitioners on how companies can successfully drive BMI (Helfat and Peteraf 2015, Loon et al. 2020; Witschel et al. 2022).

While still nascent, a small number of publications has turned its focus on the micro-foundational perspective on DCs for digital BMI (Soluk et al. 2021; Warner and Wäger 2019; Witschel et al. 2019, 2022). These authors suggest a set of different micro-foundations that are particularly relevant in the digital context, while maintaining a set of established DCs and respective micro-foundations that are consequently developed and adapted along the path towards digital BMI. Witschel et al. (2019) also demonstrate that some micro-foundations relate to a specific DC dimension of sensing, seizing, or transforming. This highlights the necessity of understanding DCs and respective micro-foundations in their interrelations and orchestration (Felin et al. 2012; Feiler and Teece 2014). These capabilities are underpinned by specific skills, activities, and organizational processes and structures (Teece 2007; Warner and Wäger 2019; Witschel et al. 2019), as shown in Table 1.

Below, several challenges for building DCs’ micro-foundations for the respective dimensions of sensing, seizing, and transforming are described.

Sensing refers to the ability to identify new business opportunities and threats in order to proactively drive BMI (Teece 2014; Schoemaker et al. 2018). Digital transformation implies a high degree of environmental dynamism and short technological and innovation cycles, making market and customer scanning and early identification of trends and new technologies challenging (Achtenhagen et al. 2013; Feiler and Teece 2014; Day and Schoemaker 2016; Warner and Wäger 2019; Witschel et al. 2019). To gain a profound understanding of customer latent needs and to obtain and integrate customer feedback at an early stage, it is essential to involve customers directly in the ideation process. This, in turn, implies to adapt established ideation processes that are typically rather of closed nature in incumbent industrial firms (Mezger 2014; Warner and Wäger 2019; Witschel et al. 2019). A further challenge is the ability to understand customer requirements and to translate these into a central value proposition, while incumbent industrial firms are rather product-driven and even struggle with servitization-driven BMI (Müller and Buliga 2019; Witschel et al. 2019). Similarly, collaborations with partners and suppliers allow the successful integration and reconfiguration of external competencies (Feiler and Teece 2014). This is thus considered an important micro-foundation within sensing (Inigo et al. 2017), such as an open innovation approach (Bogers et al. 2019; Westerman et al. 2014). Likewise, while having many testbed-like approaches, both open innovation and BMI struggle in many incumbent industrial firms (Müller et al. 2021; Witschel et al. 2019).

Seizing is subsequently required to exploit new business opportunities based on the appropriate organization of key competencies and the ability to adopt an agile and iterative approach to BMI. As for sensing capabilities, this implies challenges in adopting agile methods in traditional industrial environments (Mezger 2014; Warner and Wäger 2019; Witschel et al. 2019). Customer involvement in development allows to identify problems at an early stage in order to develop a business model that fully meets customer expectations (Amit and Han 2017; Teece et al. 2016; Witschel et al. 2019). In this context, especially smaller industrial firms struggle to present non-finished offers to customers due to established logics and culture (Müller et al. 2021). Finally, when introducing digital business models, ensuring the security and robustness of data and systems is a crucial competitive factor. Similarly, many industrial firms struggle to build and continuously develop the required competencies outside their traditional domains (Kiel et al. 2017; Porter and Heppelmann 2015; Witschel et al. 2019).

Transforming enables firms to adapt, expand, and renew their resource base to create a sustainable organization for business success in the long term (Day and Schoemaker 2016; Inigo et al. 2017; Witschel et al. 2019). Hence, building and developing key competencies is seen as an essential micro-foundation (Feiler and Teece 2014; Mezger 2014; Witschel et al. 2019). In this context, effective and open internal information exchange strengthens employee commitment and motivation (Achtenhagen et al. 2013; Song et al. 2016; Teece 2007; Witschel et al. 2019). However, this is not easily implemented for many incumbent industrial firms due to traditional communication culture (Müller et al. 2021). Also, maintaining existing partners while adding new ones with complementary resources towards ecosystem development are of great importance within transforming (Bogers et al. 2019; Day and Schoemaker 2016; Mezger 2014; Warner and Wäger 2019; Witschel et al. 2019). Incumbents struggle to transform established peer-to-peer relationships into platform-like or ecosystem-driven forms of communication and exchange, especially when adding new partners (Müller et al. 2021; Veile et al. 2022).

Conclusively, while there are several insights on challenges of building micro-foundations of DCs towards digital BMI, a comprehensive overview and analysis on interrelations is still missing (Bojesson and Fundin 2021; Soluk and Kammerlander 2021; Witschel et al. 2019). Hence, this paper develops a multiple case study approach, as described in the next section.

Since research on DCs’ micro-foundations and respective challenges are in a nascent research stage, this study follows an explorative multiple-case study design (Eisenhardt 1989). It is suitable for nascent research subjects (Edmondson and McManus 2007), as it enables “how” and “why” questions, examination of new insights, and in-depth understanding of specific and complex phenomena in a real-life context (Yin 2009). A multiple case study further provides increased rigor, robustness, reliability, and generalizability of findings compared to single-case studies (Eisenhardt and Graebner 2007; Yin 2009).

The sampling approach (Creswell 2013) was chosen to offer comparability while cases should still differ in certain characteristics to enable theory-building (Yin 2009). Thus, established industrial companies from German-speaking countries were selected that operate in various industries and run digital BMIs. This approach ensured better comparability among cases due to cultural proximity while offering generalizability due to a variety of industry sectors. However, it also limits the results in terms of generalizability to other geographical contexts.

Due to the novelty of the research field, the sampling approach follows a theoretical sampling strategy (Glaser and Strauss 2017), offering insights into the largest traditional industry sectors of Germany, Austria and Switzerland. All companies required to have (a) successfully completed digital BMI, (b) represent an incumbent firm in their industry and (c) represent large firms in contrast to start-ups, having at least 10,000 employees. Further, we deliberately decided to exclude companies from digital or IT-driven sectors such as the software industry. The case selection process was supported by a consulting firm to identify leading companies and contacts for interviews.

Initially, eight relevant case studies were identified. However, after the first round of data analysis, theoretical saturation was achieved (Yin 2009). We then decided to cede data collection at this point since the potential two further case companies would not have brought further insights into already represented industry sectors.

The final case study sample consists of six companies. Those are active in different industries, such as transport, logistics, energy, rail, tire, and industrial technologies (as described in Sect. 3.3) and engage in different forms of digital BMI. Still, homogenous factors include that all companies have successfully undergone digital BMI as per the selection criteria, have a long-standing corporate history that spans several decades. Further, several case companies are considered heavyweights on the German stock market index (DAX).

Semi-structured interviews with executives and senior managers represent the primary source of data collection, which allow flexibility and openness relating to the exploratory nature of this study (Yin 2009). We thus developed a semi-structured interview protocol (as shown in the Appendix) that was informed by extant literature but allowed themes to emerge during the interviews.

The experts were selected based on the criteria of experience with BMI, particularly their involvement and high degree of decision-making power in digital BMI. For each case study (except for EnerCo), one expert provided details on strategic and organizational aspects, while a second represented technical expertise. Thus, the inclusion of heterogeneous perspectives allowed for a complementary and in-depth discussion of the interview content.

Initial interviews were collected beginning in 2018 and complemented by follow-up data collection until 2019 to update and validate upcoming themes in the analysis process. To complement the interviews, verify statements and gain a deeper understanding of the case studies, various secondary data were collected from internal documents such as internal reports or corporate presentations that were made available by participants. Additionally, external documents, such as media reports, company websites and press releases were consulted to enrich the researchers understanding during the data analysis process. To ensure validity and reliability, the interviews were audio-recorded and verbatim transcribed for data analysis.

Table 2 provides further information on the sample and data sources.

As a method of analysis, this study follows an abductive logic approach, allowing inductive-deductive reasoning and iterations between theoretical concepts and empirical data (Dubois and Gadde 2002). Thereby, a theory-guided framework (Urquhart 2013) that is organized according to the three generic process dimensions of DCs—sensing, seizing, and transforming (Teece 2007) serves as an analytical framework for data analysis. It is further divided into the underlying micro-foundations relevant to digital BMI (Witschel et al. 2019) as described in Table 1. This enables the allocation of the inductively derived challenges in building DCs for digital BMI to each micro-foundation and thus increases the internal validity of findings (Yin 2009).

Open, axial, and selective coding were used to analyze data (Gioia et al. 2013). First, the interview transcripts and secondary data were carefully reviewed. Hereby, fundamental text passages (e.g., words, paragraphs, or quotes) of specific challenges that were explicitly and implicitly associated with the related micro-foundations of DCs and their related activities and processes for digital BMI were labeled with initial codes. Based on this open coding, first-order categories were identified. Next, axial coding was performed to examine thematic links and patterns among the first-order categories (Strauss and Corbin 1998).

Different sub-challenges such as “prioritization and selection of the right idea” and “find problem to developed solution” were aggregated into the overarching second-order theme “selecting the right idea”. Furthermore, the sub-challenges “lack of knowledge about customer requirements” and “difficulties to find and understand market need” were aggregated into “capturing and understanding market requirements”.

To avoid confirmation bias (Gioia et al. 2013), data analysis up to this point was performed inductively, i.e., without considering the theory-informed micro-foundations of DCs for digital BMI. Finally, selective coding was performed combining the second-order themes, e.g., “selecting the right idea” and “capturing and understanding market requirements” to the aggregate dimension “recognizing market dynamics and trends” that represents the respective micro-foundation of sensing capabilities.

This approach resulted in 98 underlying sub-challenges (first-order concepts) and 39 categories of challenges (second-order themes), referring to 13 micro-foundations of DCs.

Figure 1 presents the date structure on sensing, which is further described below.

Figure 2 illustrates the data structure of the seizing dimension of DCs.

Figure 3 shows the data structure of DCs’ transforming dimension.

In general, IoT platforms are expected to face more challenges than SaaS business models. This is attributed to the higher complexity of IoT platforms in comparison to SaaS business models. For IoT platforms, a multitude of actors on a platform must be cross-linked, exacerbating the challenges in hardware and software co-development, data management, data mining, privacy, and data security (Lee and Lee 2015, Veile et al. 2022). As illustrated in Table 3, digital BMI driven by IoT platforms overall face significantly more challenges (n = 76), particularly in seizing and transforming, than for SaaS solutions (n = 53). However, if one considers the sensing dimension, a different picture emerges. Figure 4 summarizes patterns relating to the similarities and differences between the two business model types.

In the sensing dimension, more mentions of challenges for SaaS (n = 19) compared to IoT platforms (n = 11) occur. Further, SaaS business models have specific challenges not relevant for IoT platforms, such as formation of technical silos or deadlocked mental patterns that impede the idea-generation phase.

In the dimension of seizing, IoT platforms are confronted with most challenges (26 out of 41). Aspects such as difficulties encountered in the organization of development teams or lack of experience towards agile methods are more pronounced for IoT platforms. In addition, cooperation with external service providers and organizational efforts towards customer integration are challenges that are not found for SaaS.

Regarding the absolute mentions within the dimension of transforming, by far most challenges (39 out of 58) are encountered for IoT platforms. As stated, this can be attributed to the high complexity of IoT solutions and digital platforms with multiple actors (Tian et al. 2021; Veile et al. 2022). This becomes particularly evident for the micro-foundations sustainable provision and development of key competencies, complex bundling and coordination of information, providing fast customer solutions, and complexities within the technical partner integration.

Since DCs and their micro-foundations are attributed to people, processes, and structures (Barney and Felin 2013; Felin et al. 2012), Table 4 in the Appendix illustrates different patterns that explain how and where challenges are manifested.

Notably, most challenges are found at the individual and process levels, while the structural level is mainly relevant for the dimension of transforming. Moreover, the results show that as the role of customers and partners in building DCs for digital BMI increases (Witschel et al. 2019), an increasing number of challenges occurs at the interface between internal and external levels.

Further, several sub-challenges are interrelated within and between different building blocks, as well as across second-order challenges. For instance, regarding the ability to recognize market dynamics and trends, the underlying (sub‑)challenges are primarily rooted in individuals. Those are also manifested, albeit less pronounced, at the level of processes and structures. On the one hand, lack of individual (or group) knowledge or limited cognition implies difficulties in identifying and understanding customer and market requirements; alternatively, technological developments negatively influence a company’s ability to recognize new business opportunities (Grégoire et al. 2010). On the other hand, lacking transparency and knowledge on competitors and markets constrains the ability of individuals to recognize customer requirements and market developments (Teece 2007). Comparably, structural (e.g., formation of functional silos) or individual challenges (e.g., presence of old thought patterns) impede cognitive abilities of individuals to identify new managerial approaches (Helfat and Peteraf 2015). Overall, these results underline Felin et al. (2012) who emphasize the interdependencies across micro-foundations.

Regarding RQ 1, the study illustrates that building DCs for digital BMI is a complex and multifaceted process characterized by various challenges in all three dimensions of DCs, while transforming being most affected. Further, concerning RQ 2, through adopting the micro-foundational lens, the study allows a more nuanced understanding of the prevailing challenges facing each specific micro-foundation. The results highlight that challenges can be attributed to clusters and different forms of digital BMI, whereby IoT platforms encounter more difficulties in building DCs that SaaS-driven BMI.

In doing so, this research contributes to the emerging research stream that examines DCs from a micro-foundational perspective with respect to digital BMI (Soluk and Kammerlander 2021; Warner and Wäger 2019; Witschel et al. 2019) by investigating the so-far lacking understanding of challenges towards building DCs. While proactively building and managing DCs is a certain asset towards digital BMI, we argue that comprehensive understanding of the underlying challenges opens a valuable perspective for research and adapting management strategies (Helfat and Peteraf 2015; Radic et al. 2022). Therefore, this study provides empirical insights from six case studies that illustrate the challenges of building DCs relevant for digital BMI from a micro-foundational lens. In doing so, the study responds to numerous calls to extend the understanding of micro-foundations in the context of digital transformation in general (Konopik et al. 2022; Steininger et al. 2021; Verhoef et al. 2021; Vial 2019), particularly factors preventing firms from building DCs (Bojesson and Fundin 2021; Favoretto et al. 2021; Helfat et al. 2007; Kraus et al. 2022; Soluk and Kammerlander 2021; Witschel et al. 2019). More specifically, this research extends literature by making the following contributions:

First, the study comprehensively investigates challenges towards building DCs for digital BMI. Therefore, it confirms and extends extant knowledge that digital BMI is challenging for incumbents in several regards. To begin with, digital BMI requires established routines and path-dependencies to consider new forms of value offer, such as turning from product-driven approaches to service-driven or data-driven value offers (Paiola and Gebauer 2020). More precisely, value creation exceedingly requires new partners with complementary competencies or even ecosystem-based approaches while incumbents often prefer internal value creation to dominate (Müller et al. 2021). Finally, value capture mechanisms that are not built upon sales of products might challenge incumbents, requiring to rethink their logic of generating profit towards fulfilling customer requirements (Müller and Buliga 2019; Witschel et al. 2019). Further, the paper provides insights into different types of digital BMI, resembling the complexity of IoT-based platform BMI in challenges of building DCs and its micro-foundations (Tian et al. 2021; Veile et al. 2022) in comparison to SaaS BMI.

A second contribution to literature is that challenges associated with building DCs in the context of digital BMI have different origins and manifest themselves at the individual, process-related, and structural levels. This extends research on micro-foundations of DCs (Barney and Felin 2013; Felin et al. 2012), which has rather focused on facilitating effects than challenges regarding micro-foundations. While individual- and process-related issues represent most challenges, structure-related issues manifest themselves less often. This might be related to the fact that digital BMI is often structurally easier to separate as it includes own value capture mechanisms rather than digital products alone that must be integrated into existing value capture structures (Rachinger et al. 2019). Conclusively, building DCs for digital BMI is rooted on different levels that are intertwined and should thus not be regarded as standalone traits (Barney and Felin 2013; Felin et al. 2012; Magistretti et al. 2021).

Third, this study also considers the nature of underlying challenges. Through examining the different types of challenges, we provide insights on requirements to change corporate culture as well as knowledge and information exchange towards digital BMI (Eriksson 2014; Foss and Saebi 2015), suggesting that organizational transformation requires the ability to change corporate culture. Partially in contrast to extant findings (Favoretto et al. 2021), purely technological or data-related challenges show a less important role for digital BMI when considering building DCs. Further, as shown by Müller et al. (2021) in their study on absorptive capacity and its role to provide ambidexterity and thus enable digital BMI, this study confirms the role of recognizing market dynamics and trends within sensing and the appropriate organization of development competencies within seizing. Moreover, the transforming dimension highlights effective know-how exchange and internal communication, and to scale the business model through partnerships as the most challenging areas. This extends Witschel et al. (2019) regarding the role of transformation as an important enabler of DCs, but an even more important challenge to address in order to develop micro-foundations of digital BMI.

Conclusively, this paper is among the first to provide a comprehensive understanding of building DCs towards digital BMI while taking a micro-foundational perspective. Further, the study illustrates that digital transformation is not just about technology, but rather it is an issue of human factors and their capabilities, processes, and structures that must be regarded in interplay with technological potentials (Björkdahl 2020; Kane et al. 2019; Vial 2019; Volberda et al. 2021).

Figure 5 summarizes the main findings.

The results of this study derive several implications for practitioners and decision-makers. In the following, several prominent aspects are highlighted.

First, relating to sensing capabilities, the results reveal that several challenges are attributed to information or knowledge deficits. Mainly found in the micro-foundation of recognizing market dynamics and trends at an early stage, this challenge is also present concerning the interaction with customers and partners. Countermeasures include securing knowledge, skills, and tools for sensing opportunities. To enhance market, customer, competitor, and technology-related knowledge, internal networks, interdisciplinary teams, cross-industry or region-spanning search represent possible options. Simultaneously, tools and method include design thinking, value stream analysis, co-creation workshops, feedback sessions with customers, or the creation of user stories. Finally, although not exclusive to the sensing dimension, cultural challenges must be solved at the outset. Foremost, management should create an open communication culture, involving employees in decisions and changes at an early stage.

Second, the study’s findings show that building DCs in the seizing dimension leads to knowledge-related challenges at both individual and process levels. Hence, firms must effectively manage, orchestrate, and coordinate knowledge processes including a culture that encourages knowledge sharing and transparency among all actors internally and externally. Integrating external IT expertise or combining market knowledge with IT expertise can assist to develop digital skills. Likewise, investments in IT tools, forming heterogeneous development teams, or cross-departmental knowledge sharing can facilitate the creation of interdisciplinary knowledge. To overcome the challenges related to external actors (e.g., service providers, customers, offshore developers), close information exchange and generating mutual understanding is advised. Regarding data handling and security, additional investments in technologies and competencies is required while running penetration tests that involve partners and customers from early on.

Third, the dimension of transforming shows most challenges and should thus be regarded in particular. The results reveal that focusing on individuals, in addition to process-related issues such as coordination and communication difficulties, promises to be an important lever to overcome challenges. The results clearly show that individuals are pivotal for the success of transforming activities. While specialists with new skillsets are required, they are difficult to recruit and retain, and they must be able to support the development of DCs within the existing team. Traits such as flexibility, independent learning, communication, curiosity and openness must be connected within teams in this context. In particular, incumbent firms with pronounced path dependencies must support cultural change and path-breaking activities. Further, management must balance coordinating and motivating employees to create suitable framework conditions and resources. Hence, top management must create a congruence between strategic orientation and organizational design. On the one hand, digital BMI must be integrated into corporate strategy as a key element. On the other hand, organizational design must resemble strategic projects. Structural options include establishing separate digital units for BMI or a central unit for cross-departmental support and coordination. Especially for incumbents, digital BMI can unfold separately complex established processes and routines of the core business. Finally, communication management must resemble internal and external levels, most notably improvement of communication channels and structures across teams, projects, and companies.

Several limitations are to be mentioned based on the multiple case study design that limits generalizability. Moreover, the case studies are limited to two business model types, i.e., SaaS and IoT platforms by incumbent industrial firms from German-speaking countries. This regional focus enhances comparability among our cases due to cultural similarities, but also limits the generalizability of results to other geographical and cultural contexts.

Hence, future studies should include further variables to obtain more far-reaching insights. For instance, considering further business model types, firm sizes, regions, or industries would enhance theory-building. Further, companies with less prominent path dependencies than the incumbent firms analyzed, such as start-ups or hidden champions may uncover different challenges. Potentially, the DCs required for digital BMI are then affected by distinct micro-foundations, as well as underlying challenges associated. Hence, additional qualitative studies should attempt to understand the micro-foundations of digital BMI in different contexts. Moreover, longitudinal studies could investigate how DCs emerge over time and during changing environmental conditions. In addition, how firms manage challenges that arise while building capabilities could be of additional value for future research. Likewise, detailed strategies towards the challenges investigated would allow to enhance the rather superficial results of this study.

Further, we acknowledge that our findings are limited to specific cases of digital BMI. While this allows to focus on the required DCs for BMI, some DCs might not be specific for digital BMI as such. Thus, we recommend for future studies to investigate the differences of required DCs between digital BMI and BMI, for instance by investigating both forms of BMI in the same companies.

Despite these limitations, the presented study serves as a starting point for understanding how and where challenges in building DCs for digital BMI occur. Further, it provides guidelines for practitioners to identify critical issues in building DCs proactively.