Digital Business Model Innovation in Complex Environments: A Knowledge System Perspective

Abstract

Digital technologies are reshaping how firms create, deliver, and capture value, prompting growing interest in digital business model innovation (DBMI). Despite increasing scholarly attention, the existing research remains fragmented and often assumes stable environments, limiting its applicability in today’s complex and dynamic contexts. To address this gap, this study conducts a systematic literature review (SLR) to gather and critically synthesize the fragmented and evolving body of knowledge on DBMI. The review identifies key research perspectives, highlights their underlying assumptions, and reveals the limitations in addressing environmental and knowledge complexity. In response, the paper introduces the knowledge system perspective (KSP) as a novel lens that views DBMI as a knowledge-driven, adaptive process. This perspective advances the DBMI literature by integrating knowledge dynamics and contextual complexity, offering a more robust understanding of how firms navigate digital transformation. The study concludes by outlining future research opportunities and providing practical implications for managing DBMI in turbulent environments.

1. Introduction

Digital technologies (DTs) have emerged as a pivotal driver of organizational competitiveness, fundamentally reshaping how firms conceive, deliver, and capture value [1,2]. As firms are increasingly engaging in digital transformation, they face not only internal restructuring, but also fundamental shifts in their business models and innovation strategies. While DTs offer transformative potential, they simultaneously thrust organizations into highly complex, volatile, and uncertain environments marked by rapid technological change and institutional turbulence [3,4].

China’s digital economy exemplifies these dynamics. Its digital economy has expanded at an extraordinary pace, driven by strong government support, high consumer adoption, and entrepreneurial innovation. The result is a digital landscape characterized by deep technological complexity, rapid innovation cycles, and fluid regulatory conditions. For instance, Tencent’s WeChat has evolved from a messaging app into a multifunctional ecosystem that integrates communication, mobile payments, e-commerce, and public services. Emerging firms like Didi Chuxing leverage real-time data and AI algorithms to orchestrate mobility services at scale, constantly adapting their models in response to regulatory shifts and user behavior. These developments highlight the limitations of static frameworks and underscore the need to understand DBMI in highly complex and dynamic environments.

Despite such developments, prevailing business model innovation (BMI) research often remains rooted in assumptions of environmental stability. While these perspectives, such as the activity system perspective, the strategic perspective, and dynamic capability perspective, have offered valuable insights into the configuration, evolution, and alignment of business models, they struggle to explain how BMI unfolds within highly interdependent and fast-evolving environments. In response, a growing body of work has begun to conceptualize BMI through the lens of the knowledge system perspective (KSP), which emphasizes the generation, recombination, and integration of knowledge across organizational and technological boundaries as the key mechanism for navigating DBMI. However, this nascent perspective remains fragmented, lacking theoretical cohesion and empirical validation.

Therefore, the purpose of this paper is threefold. First, it systematically reviews the evolution of BMI research, identifying core assumptions and mapping paradigm shifts. Second, it builds on these findings to elaborate how DBMI evolves in complex environments through the lens of the knowledge system perspective. Third, it proposes a future research agenda that tries to bridge theoretical and practical gaps in the literature. In doing so, this paper contributes to the retheorization of BMI in the context of digital complexity and environmental turbulence.

The remainder of this paper is structured as follows. Section 2 introduces the concept of business models and BMI. Section 3 outlines the methodology used for the systematic literature review. Section 4 examines the existing conceptual frameworks of business models and BMI, assesses their limitations in light of dynamic complexity, and introduces the KSP as an alternative theoretical lens. Section 5 presents the key findings and proposes future research directions.

In sum, this study makes three main contributions: First, it offers a comprehensive review of the BMI literature, identifying key thematic areas and mapping the evolution of research perspectives in this fragmented field. Second, it introduces the KSP as a novel framework that reconceptualizes DBMI as a knowledge-driven, adaptive process. Third, by generating theoretical insights and outlining future research directions, this study advances ongoing academic and managerial discussions on DBMI in complex and dynamic environments.

2. Background

The concept of the business model (BM) was first introduced into academia by Bellman et al. [5] to describe the “multi-stage and multi-agent” commercial behaviors within enterprises. Subsequently, this concept began to be used as a “management tool” in the study of administrative science by Konczal [6]. Early research on business models predominantly emphasized their functional attributes, describing them as process models of business operations. With the advent of e-business, increasing attention has led to a discussion of the definition of BM from various angles. For example, some scholars consider “transactions” to be at the core of a business model, defining it as “the content, structure, and governance of transactions” [7]. Others view it as a method of doing business to “…offer its customer[s] better value…” [8] or “… to sustain itself…” [9,10]. To represent another perspective, some argue that a business model functions more like a “logic” that, combined with “…data and other evidence…”, constitutes the value proposition for customers (p. 4). Zott and Amit [11] introduced the activity system perspective (ASP) into business model research, defining a business model as “a system of interdependent activities that transcends the focal firm and spans its boundaries”. Conceptually, although a consensus on the definition of business models is yet to be reached in academic circles [12], the understanding of business models has gradually evolved from a tool-based perspective in operations management to a more integrated perspective in strategic management [13].

BMI has evolved from the concept of the business model and has gradually become a central focus and prominent topic in business model studies [14]. Since Schumpeter’s theory of innovation was introduced, innovation research has continuously expanded, influencing various academic disciplines and drawing significant attention to the intersection of “business model” and “innovation” [15]. Over the past decade in particular, advancements in internet technology and the rise in emerging DTs have triggered a wave of transformative changes in business practices, significantly broadening and deepening the scope of BMI. The ability of enterprises to adapt to the internal and external environmental shifts driven by this technological revolution and to create or refine business model prototypes (to achieve BMI) has become a critical determinant of their success or failure [16]. Consequently, both in theory and in practice, research on BMI has been elevated to a strategic priority [13].

3. Methodology

The aim of this study is to comprehensively gather and synthesize the existing research perspectives on BMI, with the goal of further exploring DBMI from a KSP. To this end, we employ a systematic literature review (SLR), a method that ensures the comprehensiveness and replicability of the review by systematically identifying, evaluating, and synthesizing the existing literature [17]. This approach is particularly suitable for addressing complex, interdisciplinary research questions. In fact, due to the diverse academic backgrounds of researchers (e.g., strategy, economics, and sociology, etc.), there are significant differences in how BMI is conceptualized and studied. As a result, multiple research perspectives have emerged in the literature on BMI, many of which have yet to be systematically reviewed and synthesized.

Following Tranfield et al. [17]’s three-stage procedure—planning, execution, and reporting—this study conducts an in-depth analysis of the relevant literature. By consolidating diverse research perspectives, this study seeks to bridge disciplinary divides and provide a more comprehensive and systematic conceptual framework for understanding BMI.

3.1. Selecting Databases and Keywords

Following the steps of a systematic literature review (SLR), we first defined the inclusion and exclusion criteria for the literature review, determining that the selected articles should center on BMI, with a particular emphasis on diverse research perspectives. As a result, we excluded the literature that failed to offer new theories, models, or empirical evidence, as well as studies that were purely descriptive or lacked systematic contributions. Additionally, to guarantee the quality and relevance of the selected literature, we specifically focused our search on influential, high-impact articles that clearly addressed theoretical frameworks, conceptual models, typologies, or research agendas related to BMI.

Second, we selected Scopus and Web of Science as the two primary research databases. These databases were chosen because they cover multiple disciplines (e.g., strategic management, economics, sociology, and technological innovation, etc.), ensuring a broad coverage of BMI research. Additionally, both databases provide standardized search functionalities, support citation analysis and bibliometric analysis, and facilitate systematic and reproducible research. By cross-verifying the results from these two databases, we reduced the likelihood of omissions and improved the accuracy of the literature selection process.

Third, by referring to major articles in the field, we identified the keywords to be searched: (“business model innov*” OR “innov* business model” OR “business model chang*” OR “chang* business model” OR “business model adapt*” OR “adapt* business model” OR “business model transform*” OR “transform* business model” OR “new business model” OR “novel business model” “business model evolution” OR “business model” OR “business model renewal”) AND (“theor*” OR “framework” OR “perspective” OR “lens” OR “viewpoint” OR “approach” OR “insight” OR “analysis”).

3.2. Selecting the Relevant Articles

The chosen keywords, to be found either in the title, the abstract, or the keywords list, were entered into the selected databases. To ensure the quality of the literature, we excluded conference papers and proceedings from our search, limiting our sources to peer-reviewed journals. Firstly, these articles undergo rigorous peer review, ensuring the quality and credibility of the research findings [18,19]. Secondly, they typically provide more comprehensive and in-depth analyses, including detailed methodologies, extensive data, and insightful discussions, which are indispensable for fully understanding the research [17]. Additionally, peer-reviewed journals are generally archived and indexed in databases, making them easy to retrieve and reference in the future [20]. In summary, by limiting our research sources to peer-reviewed journals, we aim to ensure that the literature we obtain is reliable, of a high-quality, and readily accessible [21]. The review covers the literature published between 1975 and 2022, spanning nearly five decades of scholarly development from the earliest conceptualization of business models to contemporary digital transformations.

Our search initially yielded 242 papers. To systematically narrow these down, we employed a structured three-step screening process. First, duplicate records (51) were identified and removed, leaving 191 unique articles for review. The titles and abstracts of these articles were then independently screened by all authors using the predefined inclusion and exclusion criteria. This led to the exclusion of 116 articles that were clearly irrelevant to the research scope or lacked theoretical focus. For articles with ambiguous abstracts or doubtful relevance, full-text assessments were conducted to evaluate their conceptual and theoretical contributions to BMI. This resulted in the exclusion of an additional 37 articles. Whenever discrepancies in judgment occurred, the authors discussed the inclusion status until a consensus was reached.

Ultimately, this careful selection process resulted in identifying 38 core articles (see

Table 1. Database search process and results.

Database	Number of Journal Articles
Initial search total	242
Duplications	51
Exclusion based on abstract and intro analysis	116
Exclusion based on full-text read	37
Total selected publications	38

), representing the most influential and relevant research perspectives on BMI. As shown in

Table 2. Business model innovation perspectives.

N	Authors	Title	Year	Source	Research Perspective	Core Insights
1	Amit, R., Zott, C. [7]	Value creation in e-business	2001	Strategic management journal	Activity system perspective	Business model depicts the content, structure, and governance of transactions designed to create value through the exploitation of business opportunities.
2	Zott, C., Amit, R. [11]	Business model design: An activity system perspective	2010	Long range planning	Activity system perspective	BMI is the design of a firm’s activity system to achieve a new value proposition.
3	Hedman, J., Kalling, T. [22]	The business model concept: theoretical underpinnings and empirical illustrations	2003	European journal of information systems	Activity system perspective	The business model is conceptualized as a dynamic system of interrelated activities, including resources, processes, and market interactions, where innovation emerges from reconfiguring these behavioral components.
4	Johnson, M. W., Christensen, C. M., Kagermann, H. [23]	Reinventing your business model	2008	Harvard business review	Activity system perspective	BMI is a holistic behavioral transformation achieved through the systematic redesign of four interdependent elements: customer value proposition, profit formula, key resources, and key processes.
5	Foss, N. J., Saebi, T. [24]	Business models and BMI: Between wicked and paradigmatic problems	2018	Long range planning	Activity system perspective	BMI is not a simple linear process, but rather involves complex challenges of wicked problems and paradigm shifts. It requires transcending traditional thinking frameworks and adopting holistic approaches to address uncertainty and multidimensional problems.
6	Snihur, Y., Zott, C., Amit, R. [13]	Managing the value appropriation dilemma in business model innovation	2021	Strategy Science	Activity system perspective	Firms face a behavioral tension between value creation and value appropriation in BMI, which they manage through institutional and cognitive strategies that shape stakeholder interactions.
7	Konczal, E. F. [6]	Computer models are for managers, not mathematicians	1975	Management Review	Knowledge system perspective	Business models should be designed as knowledge-based managerial tools that facilitate decision-making by translating complex data into actionable insights.
8	Stewart, D. W., Zhao, Q. [25]	Internet marketing, business models, and public policy	2000	Journal of public policy & marketing	Knowledge system perspective	Internet business models reshape how knowledge is created, shared, and applied across markets, demanding new frameworks for organizational learning and public policy.
9	Chesbrough, H., Rosenbloom, R. S. [9]	The role of the business model in capturing value from innovation: evidence from Xerox Corporation’s technology spin-off companies	2002	Industrial and corporate change	Knowledge system perspective	Business models function as cognitive structures that help firms to make sense of technological knowledge and convert it into economically valuable innovation.
10	Doz, Y. L., Kosonen, M. [26]	Embedding strategic agility: A leadership agenda for accelerating business model renewal	2010	Long range planning	Knowledge system perspective	Strategic agility in business model renewal relies on an organization’s ability to process and reconfigure knowledge rapidly, enabled by leadership and collective learning.
11	Chen, J., Wang, L., Qu, G. [27]	Explicating the business model from a knowledge-based view: nature, structure, imitability and competitive advantage erosion	2021	Journal of Knowledge Management	Knowledge system perspective	From a knowledge-based view, business models are composed of structured knowledge assets and routines that determine their imitability and the sustainability of competitive advantage.
12	Barney, J., Wright, M., Ketchen Jr, D. J. [28]	The resource-based view of the firm: Ten years after 1991	2001	Journal of management	Resource-based view	The key to BMI lies in reconfiguring corporate resources or developing new capabilities to support new methods of value creation or operational models.
13	Morris, M., Schindehutte, M., Allen, J. [29]	The entrepreneur’s business model: toward a unified perspective	2005	Journal of business research	Resource-based view	Business models reflect how entrepreneurs configure and leverage internal resources and capabilities to create and deliver value, emphasizing alignment between resources and strategic choices.
14	Demil, B., Lecocq, X. [30]	Business model evolution: in search of dynamic consistency	2010	Long range planning	Resource-based view	Business model evolution depends on the dynamic consistency between resources and capabilities.
15	Wirtz, B. W., Pistoia, A., Ullrich, S., Göttel, V. [31]	Business models: Origin, development and future research perspectives	2016	Long range planning	Resource-based view	BMI is driven by the dynamic development and reconfiguration of a firm’s resource base, highlighting the need for adaptable resource deployment in response to environmental changes.
16	Magretta, J. [32]	Why Business Models Matter	2002	Harvard Business Review	Strategy perspective	Business models are stories that explain how enterprises work and how they create value for customers.
17	Teece, D. J. [10]	Business models, business strategy and innovation	2010	Long range planning	Strategy perspective	BMI is a strategic imperative that enables firms to capture value from innovation by aligning internal activities with external opportunities in dynamic environments.
18	Teece, D. J. [33]	Dynamic capabilities as (workable) management systems theory	2018	Journal of Management & Organization	Strategy perspective	Dynamic capabilities serve as a strategic management framework through which firms continuously adapt, integrate, and reconfigure their business models to sustain their competitive advantage.
19	Markides, C. [34]	Disruptive innovation: In need of better theory	2006	Journal of product innovation management	Strategy perspective	Disruptive innovation challenges existing business models, requiring firms to develop distinct strategic logics and business model responses rather than relying on traditional competitive strategies.
20	Lanzolla, G., Markides, C. [35]	A business model view of strategy	2021	Journal of Management Studies	Strategy perspective	BMI is not only the outcome of strategy, but should also guide strategic direction, especially in highly competitive markets.
21	Casadesus-Masanell, R., Ricart, J. E. [36]	From strategy to business models and onto tactics	2010	Long range planning	Strategy perspective	A business model is a reflection of the firm’s realized strategy, highlighting the logic linking strategic choices.
22	Iansiti, M., Levien, R. [37]	Strategy as ecology	2004	Harvard business review	Ecosystem perspective	BMI must consider the overall health of ecosystems and the role the firm plays within them.
23	Adner, R. [38]	Ecosystem as structure: An actionable construct for strategy	2017	Journal of management	Ecosystem perspective	BMI occurs within ecosystems; firms must understand ecosystem structure and collaborate with ecosystem partners.
24	Gawer, A., Cusumano, M. A. [39]	Industry platforms and ecosystem innovation	2014	Journal of product innovation management	Ecosystem perspective	BMI in platform-based industries stems from a firm’s ability to orchestrate and co-evolve with a broader ecosystem, enabling value creation through shared technological architectures and coordinated interdependencies.
25	Jacobides, M. G., Cennamo, C., Gawer, A. [40]	Towards a theory of ecosystems	2018	Strategic management journal	Ecosystem perspective	Firms can achieve breakthroughs in business models by reshaping ecosystem roles or rules.
26	Tikkanen, H., Lamberg, J. A., Parvinen, P., Kallunki, J. P. [41]	Managerial cognition, action and the business model of the firm	2005	Management decision	Cognitive perspective	Managerial cognition determines the selection and evolution of business models.
27	Baden-Fuller, C., Morgan, M. S. [42]	Business models as models	2010	Long range planning	Cognitive perspective	Business models serve as cognitive instruments that help managers and organizations articulate and communicate their strategic intent.
28	Martins, L. L., Rindova, V. P., Greenbaum, B. E. [43]	Unlocking the hidden value of concepts: A cognitive approach to BMI	2015	Strategic entrepreneurship journal	Cognitive perspective	BMI stems from changes in managers’ cognitive frameworks or mental models. Through redefining business concepts or value logic, firms can uncover hidden innovation potential.
29	Sosna, M., Trevinyo-Rodríguez, R. N., Velamuri, S. R. [44]	Business model innovation through trial-and-error learning: The Naturhouse case	2010	Long range planning	Organizational learning perspective	BMI emerges through trial-and-error organizational learning.
30	Chesbrough, H. [45]	Business model innovation: opportunities and barriers	2010	Long range planning	Organizational learning perspective	Organizational learning capabilities influence a firm’s ability to identify and overcome barriers in BMI.
31	Berends, H., Smits, A., Reymen, I., Podoynitsyna, K. [46]	Learning while (re) configuring: Business model innovation processes in established firms	2016	Strategic organization	Organizational learning perspective	BMI is a learning-by-doing process through which organizations continuously adjust their configurations to understand what can generate new value.
32	Teece, D. J. [47]	Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance	2007	Strategic management journal	Dynamic capabilities perspective	Dynamic capabilities enable firms to continuously create, extend, upgrade, and protect their resource base, thus supporting BMI.
33	Augier, M., Teece, D. J. [48]	Dynamic capabilities and the role of managers in business strategy and economic performance	2009	Organization science	Dynamic capabilities perspective	Dynamic capabilities, including sensing, seizing, and reconfiguring resources, are critical to BMI.
34	Teece, D. J. [10]	Business models, business strategy and innovation	2010	Long range planning	Dynamic capabilities perspective	BMI is the core of strategy and the key for firms to maintain their competitive advantage in rapidly changing environments.
35	Teece, D. J. [49]	Business models and dynamic capabilities	2018	Long range planning	Dynamic capabilities perspective	Firms need to adjust and optimize their business models by continuously developing their dynamic capabilities to adapt to the dynamic changes in the external environment.
36	Tracey, P., Phillips, N., Jarvis, O. [50]	Bridging institutional entrepreneurship and the creation of new organizational forms: A multi-level model	2011	Organization science	Institutional perspective	Institutional entrepreneurship facilitates the emergence of new organizational forms and business models by altering institutional environments.
37	Bohnsack, R., Pinkse, J., Kolk, A. [51]	Business models for sustainable technologies: Exploring business model evolution in the case of electric vehicles	2014	Research policy	Institutional perspective	The institutional environment shapes the paths firms choose for BMI in sustainable technologies.
38	Battilana, J., D’aunno, T. [52]	Institutional work and the paradox of embedded agency	2009	Institutional work: Actors and agency in institutional studies of organizations	Institutional perspective	Firms can innovate their business models through “institutional work” to overcome barriers or reshape rules.

, the selected articles are systematically summarized by authorship, publication year, source, and core insights, providing readers with a clear and comprehensive overview of the foundational literature in the field.

4. Thematic Analysis

4.1. Traditional Perspectives on Business Model Innovation

In the context of increasing environmental uncertainty and digital disruption, BMI has emerged as a strategic imperative for firms seeking to adapt, compete, and thrive. As a result, BMI has attracted substantial scholarly attention across diverse theoretical traditions. To better organize and interpret these contributions, this paper draws upon Mintzberg et al.’s “ten schools of strategy” framework as a higher-order analytical lens [53] (see

Table 3. Perspectives on business model innovation under Mintzberg’s strategy schools.

The Perspectives on BMI	Associated Mintzberg School(s)	Core Focus
Activity system perspective	Design School Configuration School	Interdependent organizational activities and system design
Resource-based view	Design School	Leveraging internal resources and capabilities
Strategy perspective	Design School	Strategic alignment and market positioning
Ecosystem perspective	Power School Configuration School	Network orchestration and ecosystem coordination
Cognitive perspective	Cognitive School	Managerial cognition and mental models
Dynamic capability perspective	Learning School Design School	Sensing, seizing, and reconfiguring in dynamic environments
Institutional perspective	Power School Environmental School	Institutional norms and legitimacy pressures
Organizational learning perspective	Learning School	Learning through iteration and organizational adaptation

). These ten schools, ranging from the Design, Planning, and Positioning Schools to the Learning, Cognitive, Power, and Configuration Schools, offer distinct yet complementary perspectives on how strategy is formulated and implemented. By leveraging this framework, we can systematically examine how different theoretical views interpret the mechanisms, drivers, and dynamics of BMI.

Existing research on BMI spans a range of perspectives (see Figure 1), including the activity system perspective (ASP), resource-based view (RBV), strategic perspective, ecosystem perspective, cognitive perspective, dynamic capability perspective, institutional perspective, and organizational learning perspective. Each perspective aligns explicitly or implicitly with particular schools of Mintzberg’s strategic thought, offering unique insights into how firms design, evolve, and adapt their business models. In the following sections, we categorize and analyze these BMI perspectives within the broader structure of Mintzberg’s strategy framework. This approach enables a more integrated understanding of how BMI functions not only as a set of practices, but as a strategic process embedded in different organizational logics and environmental contexts.

4.1.1. Activity System Perspective

The activity system perspective (ASP) typically regards business models as “behavioral systems” [11,22], aiming to address the following key question: How do the main participants in a business model interact to achieve value creation, as well as the processes of delivery and distribution? This perspective views the business model as a behavioral system—one centered on the focal firm but extending to partners, suppliers, and customers—through which transactions and activities are structured to enable coordinated value creation [11,22]. Consistent with the logic of the Design School within Mintzberg’s strategic framework, the activity system perspective emphasizes the deliberate configuration of internal organizational components to achieve a strategic fit with external conditions. It assumes that effective BMI results from architecting coherent activity systems that align internal processes with market opportunities. Firms design these systems by combining and sequencing activities such as training, development, manufacturing, budgeting, sales, and service [23], often in response to changing environments. Building on this design logic, the ASP identifies four primary types of innovation: content (what activities are performed), structure (how they are linked), governance (who performs them), and value logic (how value is created and captured) [13]. These categories underscore the importance of system-level thinking and the adaptability of activity interconnections—features that also resonate with the Configuration School, which emphasizes internal consistency and alignment across strategic elements.

However, despite its structured approach, the ASP has notable limitations. It tends to focus on explicit and observable activities, potentially underrepresenting the cognitive, experiential, and symbolic dimensions of strategic behavior. As critics have noted, reducing a business model to a set of coordinated behaviors may overlook the interpretive work that managers perform in shaping strategy and meaning [29]. This gap becomes particularly salient in digital contexts, where rapid change and complexity demand more than technical configuration. Thus, while the ASP offers a robust framework for analyzing the structural aspects of BMI, its explanatory power can be enhanced by integrating complementary perspectives that address the less tangible, yet strategically vital dimensions of cognition, learning, and organizational interpretation.

4.1.2. Resource-Based View

The resource-based view (RBV) typically conceptualizes BMI as the result of reconfiguring a firm’s internal resources and capabilities [54,55]. The central question this perspective addresses is as follows: How can firms achieve a sustained competitive advantage through unique combinations of valuable internal resources, and institutionalize this advantage through business models? Rooted in strategic management theory, the RBV is grounded in the premise that resources that are valuable, rare, inimitable, and non-substitutable (VRIN) can form the basis of long-term firm performance [54]. Within this framework, BMI is not merely a response to market shifts or a structural redesign, but a strategic mechanism that enables firms to translate internal strengths into unique value propositions. This logic resonates with the Design School in Mintzberg et al.’s strategic taxonomy, which emphasizes the deliberate alignment of internal strengths with external opportunities [53]. The resource-based view assumes that effective business model design stems from leveraging distinctive resource bundles and embedding them into coherent value creation architectures. Firms intentionally build competitive positions by developing and orchestrating assets such as technological know-how, brand equity, and organizational routines. BMI, in this context, becomes the vehicle through which these resource advantages are configured, stabilized, and scaled. Building on this, Amit and Zott highlight the role of resource complementarity and recombination in shaping value systems that are difficult to imitate [7]. Teece further extends the RBV logic by stressing dynamic resource orchestration—that is, the firm’s ability to adapt and reconfigure existing resources to support evolving business model logics in dynamic environments [10]. In this sense, BMI is not only a way to exploit existing capabilities, but also to renew them.

However, despite its strengths, the RBV also faces important limitations. Its internal focus often underplays the influence of external institutional and ecosystem dynamics on innovation. As Priem and Butler [56] point out, this inward-looking orientation may render the RBV less suited to explaining BMI in open, platform-based, and highly networked digital economies, where collaboration and interdependence are essential. Furthermore, the RBV offers limited explanatory power for startups or resource-constrained firms that succeed not through internal resource superiority, but through agile experimentation and ecosystem positioning. Such cases have prompted scholars to extend the RBV by integrating complementary theories such as ecosystem thinking, social capital theory, and institutional perspectives [14], thereby enhancing its explanatory reach.

4.1.3. Strategy Perspective

The strategic perspective approaches BMI as a purposeful mechanism through which firms realize long-term strategic goals and maintain competitive advantage [10]. Rather than viewing BMI as a byproduct of technological innovation, this perspective treats it as a central component of strategy formulation—used by firms to respond to market shifts, reconfigure resources, and address innovation-related uncertainties [11]. It reflects a view of strategy as intentional design and decision-making at the organizational level. This understanding aligns closely with the Design School logic within Mintzberg’s strategy framework, where strategy is conceptualized as a rational process of aligning internal capabilities with external opportunities [53]. From this standpoint, BMI functions as a structural expression of strategic intent. Firms design or revise their business models to strengthen strategic fit—redefining how value is created and captured in response to shifting technological, market, or institutional conditions. For instance, Chesbrough [57] highlights how firms use BMI to integrate internal and external resources, improving the commercialization of innovation. Markides similarly notes that BMI can unlock new revenue models and reduce strategic risk [34]. This perspective is further supported by insights from the Planning School, particularly in its emphasis on deliberate planning and resource allocation. BMI, in this view, is often linked to executive-level choices about long-term direction, ecosystem engagement, and capability deployment. As Casadesus-Masanell and Ricart suggest, effective BMI requires a close coupling between strategic decision-making and business model configuration [36]. In today’s global and digital environments, such adaptability becomes critical for firms to stay competitive [58,59].

However, the strategic perspective’s focus on top-down planning may underemphasize the emergent, iterative, and experimental nature of many real-world innovation processes [44,60]. BMI often unfolds through cycles of learning and revision rather than linear execution [30]. While the strategic lens highlights intentionality and alignment [10], it benefits from complementary perspectives that account for how strategies evolve under uncertainty and change. Overall, the strategic perspective offers a valuable view of BMI as a means of translating strategic goals into organizational systems. Yet, to fully capture its complexity, it must also accommodate the dynamic and adaptive characteristics of innovation in practice.

4.1.4. Ecosystem Perspective

The ecosystem perspective views BMI as a process integrated within a wider network of interdependent actors, including suppliers, partners, customers, and platform providers, rather than a firm-centric or internally focused activity [38]. From this viewpoint, BMI is not only about internal capabilities, but also about coordinating across organizational boundaries to create, deliver, and capture value collectively [37,61]. Firms engage in value co-creation by integrating and aligning resources, roles, and interests within the ecosystem. This perspective reflects the logic of Mintzberg’s Power School, which views strategy formation as a political and negotiated process involving multiple stakeholders. In the ecosystem context, firms must continuously navigate power dynamics, manage interdependencies, and negotiate mutual gains among actors with diverse goals [38]. BMI here depends not only on firm-level planning, but also on the ability to shape and influence the strategic architecture of the entire ecosystem. As ecosystems grow more complex, especially in digital platforms and networked industries, such coordination becomes a key source of competitive advantage. Moreover, the ecosystem perspective resonates with the Configuration School in its emphasis on systemic fit and inter-organizational coherence. Firms do not operate in isolation, and successful BMI requires configuring an adaptive and stable structure across multiple participants. This includes clarifying ecological niches, fostering complementarities, and building symbiotic relationships that enable ongoing innovation and value exchange [39,62].

The ecosystem perspective of BMI shifts the unit of analysis from the individual firm to the ecosystem level. It recognizes that innovation outcomes increasingly depend on the collective alignment and performance of all involved actors [40]. For instance, platform-based strategies often enable upstream and downstream integration, allowing for more effective resource mobilization and value flow across the network. However, coordinating innovation in ecosystems presents inherent challenges. Firms must manage divergent interests, resolve potential conflicts, and avoid value capture asymmetries. Ecosystem-level BMI requires not only strategic foresight, but also governance mechanisms that promote fairness, reciprocity, and trust among actors.

4.1.5. Cognitive Perspective

The cognitive perspective conceptualizes BMI as a process shaped by the mental models, interpretive frameworks, and sensemaking patterns of organizational decision-makers. Rather than viewing innovation as solely a response to external conditions or internal resources, this view emphasizes the central role of managerial cognition in recognizing opportunities, constructing meaning, and initiating change [41,63]. Cognitive structures influence how firms perceive their environment, evaluate alternatives, and define the logic of value creation—particularly in dynamic and uncertain contexts. This line of thinking aligns closely with Mintzberg et al.’s Cognitive School, which frames strategy formation as a mental process grounded in the subjective perceptions and knowledge structures of managers [53]. Within this framework, business models themselves are understood not just as structural configurations, but as cognitive artifacts that guide strategic reasoning and collective understanding [42]. Based on these, cognitive perspective research highlights that BMI often emerges from shifts in cognitive frameworks, such as reinterpreting value logic, reframing customer needs, or questioning taken-for-granted assumptions [43,64]. These shifts enable firms to break free from the dominant logic and unlock previously overlooked strategic options. Moreover, knowledge sharing and collective cognition also play a critical role, as business model transformation frequently depends on shared learning, internal dialog, and cross-functional interpretation [65]. Furthermore, leaders’ cognitive abilities significantly influence the innovation process. Their capacity to integrate information, navigate ambiguity, and reshape team cognition becomes crucial for fostering organizational adaptability [66]. In some cases, internal cognitive conflict that arises from diverse interpretations or competing logics can spark creative tension that drives new business model experimentation [67].

The cognitive perspective offers a distinct analytical lens that explains BMI as a product of strategic sensemaking and cognitive transformation, rather than solely environmental adaptation or resource deployment. However, it may understate the material and structural constraints that shape what is cognitively feasible [68,69], or the influence of ecosystem dynamics on perception [38].

4.1.6. Dynamic Capabilities Perspective

The dynamic capabilities perspective approaches BMI as a continuous process of resource adaptation and organizational learning in response to external change. Rather than focusing solely on existing capabilities, this view emphasizes the firm’s ability to sense, seize, and reconfigure resources in dynamic market and technological environments [47,67]. BMI, in this context, represents dynamic capabilities, serving as a method for firms to restructure internal assets and knowledge to create new value generation approaches. This perspective aligns closely with the logic of Mintzberg’s Learning School, which views strategy not as a fixed plan, but as an emergent process shaped by experience, experimentation, and ongoing feedback [53]. Firms engaging in BMI often iterate through cycles of trial-and-error, progressively refining their models based on environmental signals and organizational learning [70]. Especially in fast-paced digital environments, firms must build new capabilities, including data analytics, platform coordination, and networked collaboration, to enable agile adaptation and business model transformation [14,71]. Moreover, the Design School also underpins this perspective. Dynamic capabilities are not entirely emergent, but involve purposeful processes of capability development and resource alignment to support strategic renewal. Effective BMI thus depends on both responsive learning and the intentional reconfiguration of business models over time [4,71].

However, implementing dynamic capabilities poses challenges. Firms often struggle to balance short-term operational efficiency with long-term innovation goals, especially in resource-constrained settings [72]. Developing these capabilities demands technical investments, cultural openness, leadership support, and cross-functional coordination, which may not be easily sustained [47]. Nonetheless, the dynamic capabilities perspective offers important insights into how firms build the strategic flexibility and adaptive routines needed to navigate uncertainty. It reframes BMI as a learning-driven and capability-based process, where competitive advantage stems not just from what resources a firm has, but from how well it can transform them in response to change.

4.1.7. Institutional Perspective

The institutional perspective views BMI as a strategic response to external institutional pressures. Rooted in institutional theory, this perspective emphasizes that firm behavior is shaped not only by economic rationality, but also by social, legal, and cultural environments [73]. Firms must address both formal institutions, including policies, laws, and regulations, and informal ones, such as cultural norms and industry expectations [74]. From this standpoint, BMI often arises as firms seek legitimacy, compliance, or strategic advantage within evolving institutional frameworks. This perspective aligns with Mintzberg’s Environmental School, which views strategy as a reactive and adaptive process shaped by environmental forces beyond the firm’s control [53]. In dynamic institutional environments, BMI becomes a mechanism through which firms adjust their structures, offerings, and governance models to accommodate new institutional logics. For example, regulatory shifts or changes in market norms may compel firms to redesign their models to meet new compliance standards, or to leverage emerging institutional incentives [75].

In addition to adaptation, the institutional view also recognizes the constraining effects of path dependence and institutional inertia. Firms often remain embedded in established routines and legacy structures, making BMI difficult to implement without disrupting deeply institutionalized practices. This inertia reflects the cognitive and normative pressures exerted by the institutional environment, which can reinforce stability at the expense of flexibility. However, significant institutional disruptions, like those caused by digitization or globalization, can create opportunities for change, prompting firms to challenge existing norms and adopt innovative strategies that break from tradition. These dynamics reflect insights from Mintzberg’s Power School, where strategic behavior emerges through negotiation, influence, and contestation. Firms may not only adapt to institutions, but also seek to influence them through lobbying, shaping narratives, or engaging in institutional entrepreneurship [52]. In doing so, they actively participate in redefining the boundaries and logics that govern industry norms.

4.1.8. Organizational Learning Perspective

The organizational learning perspective conceptualizes BMI as a dynamic and iterative process driven by the continuous acquisition, integration, and application of knowledge within the firm. Rooted in organizational learning theory, this perspective holds that sustained competitive advantage arises from the firm’s ability to generate, disseminate, and leverage knowledge through experience and interaction [76]. Rather than being a one-time structural redesign, BMI is viewed as an ongoing process of experimentation, reflection, and adaptation. This logic is closely aligned with Mintzberg’s Learning School, which sees strategy formation not as a product of deliberate planning, but as an emergent outcome of organizational experience and learning over time [53]. BMI, in this perspective, evolves through feedback loops as firms respond to environmental shifts, test new configurations, and refine their value creation logic based on learning outcomes. Firms gather insights from both internal operations and external environments, allowing them to iteratively adapt their models to changing market demands and technological conditions [77,78].

In highly uncertain and rapidly changing contexts, such as digital markets or innovation-intensive industries, learning becomes a critical organizational capability [76]. Firms that successfully institutionalize learning mechanisms, such as after-action reviews, knowledge-sharing practices, and boundary-spanning teams, are better equipped to manage complexity and maintain innovation momentum [77]. Moreover, organizational learning also facilitates cross-functional knowledge integration, which is essential for innovation that spans multiple domains of the business [79]. Through shared learning processes, firms promote knowledge exchange across departments, encouraging experimentation and enabling the recombination of diverse insights in new business model designs [80]. This integration of varied perspectives enhances both the novelty and adaptability of innovation efforts.

Figure 1. Timeline of evolution of the traditional perspectives on BMI [6,7,9,28,37,41,44,47,52].

Figure 1. Timeline of evolution of the traditional perspectives on BMI [6,7,9,28,37,41,44,47,52].

4.2. Three Core Issues of Business Model Innovation

Generally speaking, the research on BMI has entered a period of rapid development. Based on the retrospective studies of Foss and Saebi [14] and Zott et al. [12] in this field, we divided the existing research into three core issues: connotations, process, and outcome (see Figure 2).

4.2.1. The Multiple Connotations of Business Model Innovation

The existing research on BMI reveals substantial variation in how its core concepts are defined and interpreted. These conceptual inconsistencies hinder efforts to establish clear, operationalizable constructs that can support the empirical examination of causal relationships and underlying mechanisms [81]. As a result, the theoretical conceptualization of BMI remains a central focus in current academic discourse [82]. Two key themes frequently emerge in this stream of research. The first concerns the degree of novelty in BMI. Building on Schumpeter’s (1934) foundational work, scholars have categorized innovation based on its novelty, distinguishing whether it is “new to the firm”, “new to the industry”, or “new to the world” [15]. Given that business models involve multiple stakeholders, debates around novelty are particularly salient in the BMI literature. Some researchers contend that a business model is innovative if it is new to the focal firm [23,83], while others argue that true innovation requires industry-level novelty [84]. The second theme relates to the scope of BMI. This pertains to the scope and thoroughness of the changes made to the original model. Some studies suggest that even altering a single component, such as redefining the value proposition, qualifies as innovation [85]. In contrast, other scholars argue that BMI must involve changes to at least two components, or require a holistic reconfiguration of all elements and their interdependencies [86]. In sum, there is no scholarly consensus on what level or scope of change qualifies as BMI. This lack of agreement likely stems from two interrelated factors: the evolving nature of business practices, which continuously reshapes the meaning and boundaries of BMI, and the absence of a unified theoretical framework to anchor conceptual clarity [14].

4.2.2. Business Model Innovation as a Process

This stream of research adopts a dynamic perspective, focusing on the process of organizational innovation and transformation in BMI. Three main areas of attention have emerged:

(1)

Stages and Process of Business Model Innovation

Building on strategic management research, specifically the perspectives of strategic positioning, organizational learning, and cognitive behavior, studies on the process of BMI can be broadly classified into three main approaches: rational positioning, evolutionary learning, and the cognitive perspective. The rational positioning approach conceptualizes BMI as a rational optimization process, in which managers proactively redesign the business model in response to external environmental changes [36]. This process is typically structured into sequential phases, including design (or development), testing (or implementation), and iterative adjustment. The evolutionary learning approach emphasizes BMI as a dynamic process of organizational exploration, driven by trial and error [29,44]. It includes stages such as identification, optimization, adaptation, modification, and restructuring [87]. The cognitive perspective emphasizes the vital role of managerial cognition, especially mental models and cognitive schemas, in steering the innovation process. It views BMI as a process through which managers identify opportunities, compare alternatives, integrate new insights, and modify existing models based on their cognitive interpretations of internal and external cues [43].

(2)

Internal and External Factors Influencing Business Model Innovation

Prior research has identified a range of internal and external factors that significantly shape the process of BMI. Internally, three categories of factors are commonly emphasized: ① Resource and Capabilities. For instance, whether a firm possesses the dynamic capabilities to sense opportunities, seize them, and reconfigure resources to support BMI [10]. ② Organizational Activities. This includes organizational learning processes that enable firms to absorb experience and iteratively refine their business models over time [44]. ③ Managerial Characteristics and Capabilities. This includes managers’ cognitive capacity and decision-making abilities [88], their ability to identify opportunities and anticipate environmental threats [9], and the heterogeneity of the management team. Externally, key factors involve the institutional environment, technological environment [10,89], market environment [90], and industry competition [91]. Notably, due to the strong link between BMI research and the rise of digital and e-business, a significant body of literature has explored how advances in communication technologies stimulate BMI. Central to this discussion is the recognition that effectively leveraging technological change is a key driver of BMI [92,93]. Overall, while both internal and external factors are acknowledged, the extant literature continues to place particular emphasis on external enablers [14], especially the pivotal role of technological advancement in driving BMI.

(3)

The Role of Experimentation and Learning in Business Model Innovation

Experimentation and learning have been repeatedly emphasized as critical mechanisms in the process of BMI [93,94]. In uncertain environments, enterprises can pursue BMI through various modes of experimentation and learning, such as commitment, incremental experimentation, and radical experimentation [95]. The ability to adapt learning modes over time is considered essential for fostering new business models. Despite this recognition, research on the broader set of factors influencing the BMI “process” remains limited. Few studies provide comprehensive analyses that connect these factors to specific mechanisms within the innovation process. In addition, most existing studies are retrospective, descriptive, and based on case study methodologies, with relatively few predictive investigations that test theoretical assumptions and causal relationships using empirical evidence.

4.2.3. Business Model Innovation as an “Outcome”

Innovation can be understood as both a process and an outcome. From an outcome-oriented perspective, BMI is viewed as the result of an organizational transformation process, with emphasis placed on its outputs and strategic impacts. In recent years, the concept of the business model has garnered significant attention in both academic and managerial circles, primarily due to its critical role in enabling organizations to create, sustain, and enhance competitive advantage [9,23,29,31,96]. Research in this domain has primarily evolved in three key directions:

(1)

A New Source of Performance Difference

The first direction emphasizes that BMI represents a distinct and increasingly important source of performance heterogeneity among firms. A central question in strategy and innovation research is why firms differ in performance and how they sustain competitive advantage. Traditional explanations have drawn on various theoretical perspectives, including industry structure and strategic positioning, the resource-based view, and core competencies, as well as dynamic capabilities. Recent developments in business model research have contributed to this discourse by highlighting the strategic importance of BMI as a driver of both competitive advantage and firm performance [97]. This has reinforced the view that BMI holds unique strategic value within the broader landscape of innovation and strategy scholarship [35,98].

(2)

Exploration of Contingency or Moderating Variables

The second direction focuses on identifying potential contingency or moderating variables that influence the competitive advantage derived from BMI. While the process perspective emphasizes the various internal and external factors that shape BMI, the outcome perspective is concerned with how these factors interact with contextual contingencies to generate differentiated innovation outcomes. At the macro-level, formal and informal institutional participation, along with the pursuit of legitimacy, have been shown to play critical roles in shaping the outcomes of BMI [99]. At the meso-level, organizational factors such as values and culture [100], as well as organizational design and structure [101], significantly influence the effectiveness of BMI. At the micro-level, managerial traits and capabilities, combined with employees’ skills, motivation, and commitment, are potential influences on how BMI initiatives are implemented and sustained.

(3)

Limited Attention on Negative Effects

The third direction highlights the overlooked negative effects of BMI. Studies adopting the outcome perspective frequently assume that BMI inherently leads to positive results and enhanced competitive advantage. However, this assumption does not always hold. For example, empirical evidence suggests that BMI efforts aimed at disrupting industry structures, revenue models, or organizational boundaries do not necessarily translate into superior financial performance [102]. Moreover, even when BMI is initially successful, it is often subject to rapid imitation by competitors, which can quickly erode any temporary competitive advantage gained [103].

Whether approached from a process-oriented or outcome-oriented perspective, research on how BMI shapes, enhances, and sustains competitive advantage remains in its nascent stages. Further theoretical development and empirical investigation are needed to build a more comprehensive and integrated framework.

4.3. The Rise of Digital Business Model Innovation

To better situate BMI within the digital era, scholars have introduced the concept of DBMI. In its broad sense, DBMI refers to BMI occurring in the context of digital transformation, regardless of whether DTs are the direct driving force. In contrast, a narrower definition conceptualizes DBMI specifically as BMI that is enabled or driven by DTs. This perspective focuses on how firms leverage emerging DTs to transform or redesign their business models [16,49,89,104]. To emphasize the distinct changes and characteristics introduced by DTs, this study adopts the narrower definition. Accordingly, DBMI in this paper primarily refers to BMI that is directly driven by emerging DTs. As an emerging research domain, DBMI continues to face challenges such as ambiguous conceptual boundaries and fragmented theoretical development. Nevertheless, existing studies have demonstrated that the embeddedness of DTs introduces new dimensions and characteristics into the BMI process (see

Table 4. The new connotation, characteristics, and dilemma of DBMI.

Research Topics	Traditional Research on BMI	DBMI
Connotation	Discusses the changes in business activity systems and business behaviors in terms of value creation, value transformation, and value capture arising from the novelty or scope of business models, through perspectives such as “transaction activities” [7] and “decision-making behaviors” [36].	Emerging DTs, characterized by their technological knowledge attributes, have become new “innovation elements”, enabling [104] or empowering [105] the evolution of business models through functions such as automation, augmentation, and transformation. This reshapes firms’ business cognition, logic, and modes related to value creation, value transformation, and value capture.
Characteristics of process	Focuses on critical process activities and behaviors within BMI, such as research and development (design), implementation (testing), and adjustment (reconfiguration). Particularly emphasizes how external factors, such as institutional and technological environments, influence the innovation process of business models. Highlights the role and mechanisms of experimental actions and learning approaches, including “trial-and-error” and “iteration”, within the innovation process.	The focus is on the surge in innovation elements such as “scale” and “connectivity” driven by digital technology embedding, and the resulting systematic changes induced by innovation processes. Special attention is paid to digital technology features such as “imitability” and “transferability”, significantly lowering the barriers to external knowledge acquisition and expanding its boundaries. The integration of massive user data drastically increases the complexity of business model knowledge, making the utilization of “big data” a central issue in BMI. Traditional innovation models based on physical resources and specific behaviors are increasingly replaced by knowledge creation models relying on digital twins and machine learning.
Performance results	Mainly discusses the sources of performance differences in business models: first, the direct impacts by exploring how firm strategies, resources, and capabilities affect BMI performance; second, the indirect impacts by examining the moderating effects of contingent factors such as institutional and organizational contexts.	The primary concern is how embedding DTs leads to performance differences in firms’ BMIs, emphasizing the dual-edged impact of DTs. On the one hand, DTs facilitate access to relevant external knowledge through replicability and standardization; on the other hand, these same characteristics lead to the “competitive imitation” dilemma, making it challenging for innovating firms to prevent imitation, thus easily eroding their competitive advantage.

).

4.3.1. “New Connotations” of Digital Business Model Innovation

The embeddedness of DTs has enriched the conceptual connotations of BMI. In contrast to traditional forms of BMI, DBMI places greater emphasis on leveraging the capabilities and potential of DTs to create and capture value, often resulting in the reconfiguration of core business model components. Accordingly, understanding the BMI driven by DTs requires paying close attention to the functional characteristics of these technologies, particularly how they bring new dimensions and connotations to the innovation process.

DTs are understood to perform two fundamental functions in BMI: enabling and empowering [105]. The enabling function refers to the capacity of DTs to make previously unfeasible business activities and operations possible within the structure of a business model. In contrast, the empowering function relates to the enhancement of performance and value across key components and architectures of BMI through the application of DTs. Moreover, the impact of DTs on BMI can be categorized into three forms of change: automation, extension, and transformation [104]. Automation involves the use of digital tools to streamline or improve existing activities and processes, such as information dissemination or communication facilitation. Extension captures cases where DTs support new modes of operation that complement, rather than replace, the existing practices. Transformation denotes the application of DTs to fundamentally replace traditional business activities with novel operational models. DBMI can be further classified into two critical dimensions: the type of bottleneck it seeks to overcome (integration vs. independence) and the nature of the firm’s response to digitalization (transformational vs. incremental). Based on this framework, DBMI can be grouped into four functional categories: extension, empowerment, disruption, and market creation.

4.3.2. “New Characteristics” of the Digital Business Model Innovation Process

The inherent characteristics of DTs, such as enabling functionality, integration capacity, and self-evolving capabilities, facilitate value creation involving a broader set of participants, stronger interconnections, and more intensive interactions. These technological features have redefined the scope and boundaries of business model prototypes. Compared to insights from earlier research on traditional BMI, the embeddedness of DTs introduces several new characteristics into the innovation process, fundamentally altering how business models are designed, implemented, and evolved.

First, the vast volume of data generated by emerging DTs has made the experimentation and learning processes in BMI significantly more feasible through the use of “digital twin technology”. As previously discussed, BMI is inherently linked to experimental learning. In traditional settings, such experimentation is both essential and resource-intensive, requiring the deployment and testing of multiple business model components under real-world constraints. The embeddedness of DTs enables large-scale, high-resolution data collection, thereby rendering the BMI process “super-scalable” [106]. When supported by digital twin technology, experimentation and learning can be conducted in a virtualized, data-rich environment. This substantially lowers the costs of decision-making and trial-and-error, while simultaneously enhancing innovation efficiency and the likelihood of success.

Second, DTs reduce barriers to knowledge sharing, enabling greater consumer participation in the innovation process. These technologies facilitate real-time knowledge exchange and collaboration among producers, partners, and large consumer groups, effectively integrating consumer needs into production and innovation workflows. Moreover, digital tools support the personalization of demand by enabling knowledge transfer and establishing structures and coordination mechanisms within innovation ecosystems. Through these processes, DTs enhance the potential for value co-creation within business models.

Third, increasing attention has been paid to understanding the dynamic role of DTs throughout the BMI process. A critical component of analyzing DBMI lies in identifying how DTs function across different stages of the innovation lifecycle. Existing studies have proposed preliminary analytical frameworks. For instance, DTs have been found to support activities such as data generation, acquisition, processing, aggregation, analysis, visualization, and distribution during the innovation process [107]. Further work has simplified these functions into three core mechanisms: representation, connectivity, and aggregation [106]. While these contributions provide useful starting points, the current research has yet to fully integrate these technological functions into a cohesive innovation process framework. As a result, the development of a unified and comprehensive analytical model for DBMI remains an open challenge.

4.3.3. “New Dilemma” of Enhancing Performance Through DBMI

There is currently no clear consensus in the literature regarding whether the embeddedness of DTs into BMI consistently yields a sustainable competitive advantage or improved firm performance. From a positive perspective, it is argued that innovative business model designs enable firms to convert digital technology investments into tangible value [12]. This value can be realized through enhanced internal and external efficiency via cost reduction and process optimization [108], the exploration of new business models and cross-industry collaborations [109], and the strengthening of digital R&D capabilities to increase market share [104], ultimately contributing to improved organizational performance. Conversely, a more critical view suggests that despite widespread enthusiasm for digital transformation, firms engaging in digitalization-driven BMI often exhibit inconsistent performance outcomes, with some even experiencing performance declines. Two primary explanations for this have been identified. First, DTs require substantial capital and human resource investments to support organizational learning, which can elevate cost pressures and reduce competitive flexibility. Second, the excessive use of digital tools can result in an information overload, thereby increasing both managerial complexity and the costs associated with learning and decision-making [110].

The conflicting findings of the current research may be attributed to the underexplored issue of competitive imitation driven by the replicability of DTs. Replicability, a core characteristic of DTs, enables BMIs to be rapidly deployed and scaled at minimal cost. However, as business processes, information, and operational methods become increasingly digitalized, imitation not only becomes easier, but also more accurate. The codification and standardization of knowledge undermine the scarcity of previously unique competitive resources, thereby increasing firms’ exposure to imitation risks from competitors [27]. In this context, the ability to build robust competitive barriers and enhance organizational resilience becomes critically important. Nevertheless, this challenge remains insufficiently addressed in the current body of research.

4.4. Rediscovering Digital Business Model Innovation: The Emergence of the Knowledge System Perspective

4.4.1. The Emergence of the Knowledge System Perspective in Digital Business Model Innovation

The inherent complexity of business models and their innovation processes has been widely acknowledged in academic research [13,14]. The embeddedness of DTs has further amplified this complexity. From the perspective of internal innovation processes, digitalization blurs the boundaries between innovation processes and outcomes, rendering them more intricate and dynamic in nature [27]. Moreover, the technical sophistication of DTs imposes greater demands on firms’ capabilities to acquire and apply digital knowledge, including competencies in data mining, analytics, and digital system integration. From the perspective of the external competitive environment, digitalization has introduced heightened dynamism and uncertainty into competitive environments, resulting in an increased emergence of disruptive actors. Firms are thus compelled to continuously innovate their business models to maintain competitiveness. However, digital business models are inherently replicable, which accelerates imitation and erodes Schumpeterian rents, making it more difficult to sustain long-term competitive advantage. From the interplay between internal and external environments, innovators are required to incorporate feedback from the digital environment in real time and iteratively refine their business models through ongoing experimentation and trial-and-error processes. Such adaptability is essential to meet the evolving and heterogeneous demands of markets and stakeholders [111]. Simultaneously, firms must maintain a certain level of complexity within their business models to counteract the risks of competitive imitation and preserve innovation defensibility.

In summary, the integration of DTs not only exposes firms to increasingly complex and dynamic environments, but also raises demands regarding their ability to process large volumes of data and integrate such data with existing innovation components. As a result, traditional research perspectives on BMI, reviewed in the preceding sections, have evolved in response to these emerging challenges. One of the most significant developments in this evolution is the emergence of the KSP, which complements and extends the traditional activity system perspective by offering new explanatory power in digitally complex contexts (see Table 5).

The KSP offers insights into the characteristics and dynamics of the innovation process by conceptualizing business models as structured “knowledge systems”. Early foundations of this view can be traced back to studies that defined business models as managerial tools embedding organizational knowledge and guidance [6]. Subsequent research has elaborated on this foundation by framing business models as forms of “logic” [9], “narrative” [25], and “cognitive schema” [43]. From this perspective, a business model is understood as a cluster of knowledge that guides how a firm defines its boundaries, creates value, and structures internal activities [26]. BMI, therefore, is seen either as a process of creatively integrating and applying knowledge across diverse domains [14], or as an ongoing effort in cognitive framing and meaning construction [43]. This view highlights the cognitive and epistemological dimensions underpinning how business models evolve and adapt over time.

In recent years, as DBMI has become more widespread, research adopting the KSP has also evolved. The focus has shifted from isolated knowledge elements within BMI toward the reconstruction of a more holistic analytical framework based on the concept of “complex knowledge systems” [27]. One such framework is the Business Model Iceberg Theory, which conceptualizes the business model as a complex knowledge system embedded within industry and social and technological environments, and aligned with a firm’s internal conditions [112]. This framework highlights the importance of analyzing the tacit knowledge that lies beneath the observable elements of the business model. Furthermore, business models have been conceptualized as structured knowledge clusters composed of both explicit and tacit components [27]. The explicit components include organizational activity patterns [113], customer value propositions [9,114,115], revenue models [116,117], and key resources and processes [23]. In contrast, the tacit components encompass elements such as practical skills [118], cognitive maps and schemas [119], emerging meanings, and processes of meaning construction [120]. This line of research suggests a growing movement toward an integrated systems perspective that accounts for both the structural and interpretive dimensions of BMI. From this standpoint, a business model can be defined as “a structured knowledge cluster that describes how a firm integrates business-related components (e.g., factors, resources, institutional conditions) to form business themes, and how it coordinates these themes (e.g., core strategies, meaning construction processes, representative sub-models) to create and deliver value” [27,103].

Table 5. Activity system perspective and knowledge system perspective on business model innovation.

Research Perspective	Activity System Perspective	Knowledge System Perspective
Connotation	The activity system perspective conceptualizes business models as “a set of interdependent activities or actions centered around a focal firm, involving the focal firm itself, its partners, suppliers, or customers” [11].	The knowledge system perspective conceptualizes the business model as a structured knowledge cluster, defined as “cognitive or mental representations regarding how the firm sets boundaries, creates value, and organizes its internal governance structure” [26].
Research focus	This perspective focuses on the behavioral activities related to BMI—particularly, how companies create value through transactions and key activities within their business models.	The knowledge system perspective focuses on how firms coordinate business-related components (such as factors, resources, and conditions) to form knowledge themes (or knowledge subsets), and how they further coordinate these knowledge themes to create and deliver value.
Key innovation elements	The activity system perspective identifies critical activities such as training, development, manufacturing, budgeting, planning, sales, and service, which significantly influence the effectiveness of BMI [23].	The knowledge system perspective pays attention to the creative combination and utilization processes of component elements during BMI, emphasizing the specific experimental and learning mechanisms involved.
The main differences	The activity system perspective emphasizes how these activities, their interdependencies, and cross-boundary interactions collectively enable value creation.	The knowledge system perspective focuses on how knowledge facilitates mechanism creation, highlighting that BMI encompasses not only explicit knowledge components like value propositions and process architectures, but also implicit knowledge components such as skills, schemas, and sense-making.
Representative studies	Zott and Amit, 2010 [11]; Snihur et al., 2021 [13]; Tykkyläinen and Ritala, 2021 [121].	Chen et al., 2021 [27]; Sosna et al., 2010 [44].

Table 5. Activity system perspective and knowledge system perspective on business model innovation.

Research Perspective	Activity System Perspective	Knowledge System Perspective
Connotation	The activity system perspective conceptualizes business models as “a set of interdependent activities or actions centered around a focal firm, involving the focal firm itself, its partners, suppliers, or customers” [11].	The knowledge system perspective conceptualizes the business model as a structured knowledge cluster, defined as “cognitive or mental representations regarding how the firm sets boundaries, creates value, and organizes its internal governance structure” [26].
Research focus	This perspective focuses on the behavioral activities related to BMI—particularly, how companies create value through transactions and key activities within their business models.	The knowledge system perspective focuses on how firms coordinate business-related components (such as factors, resources, and conditions) to form knowledge themes (or knowledge subsets), and how they further coordinate these knowledge themes to create and deliver value.
Key innovation elements	The activity system perspective identifies critical activities such as training, development, manufacturing, budgeting, planning, sales, and service, which significantly influence the effectiveness of BMI [23].	The knowledge system perspective pays attention to the creative combination and utilization processes of component elements during BMI, emphasizing the specific experimental and learning mechanisms involved.
The main differences	The activity system perspective emphasizes how these activities, their interdependencies, and cross-boundary interactions collectively enable value creation.	The knowledge system perspective focuses on how knowledge facilitates mechanism creation, highlighting that BMI encompasses not only explicit knowledge components like value propositions and process architectures, but also implicit knowledge components such as skills, schemas, and sense-making.
Representative studies	Zott and Amit, 2010 [11]; Snihur et al., 2021 [13]; Tykkyläinen and Ritala, 2021 [121].	Chen et al., 2021 [27]; Sosna et al., 2010 [44].

4.4.2. Applicability of the Knowledge System Perspective in the Digital Context

As an emerging theoretical lens for examining BMI within the context of digital transformation, the KSP offers a novel analytical framework for interpreting and evaluating DBMI. In comparison to traditional perspectives, the knowledge system perspective provides distinctive advantages in addressing the inherent complexity of DBMI, particularly across three critical dimensions.

First, the KSP provides a suitable framework for analyzing the complex collaboration among innovation actors in digitally intensive environments. Within the digital innovation landscape, innovation actors have evolved beyond traditional human agents to include a hybrid constellation of humans and knowledge units. The vast amount of user data generated by DTs encapsulates latent demands that are difficult to identify using conventional human-centered analytical methods. In practice, non-human “knowledge units”, powered by cloud computing and artificial intelligence, are increasingly handling data processing and decision-making tasks with growing autonomy and efficiency. For instance, AI-driven recommendation algorithms can autonomously detect user needs, filter relevant information, and make decisions without direct human input. This shift marks the emergence of a new paradigm characterized by human–knowledge unit collaboration, in which knowledge autonomously generates new knowledge [122,123], thereby creating new pathways for value discovery, creation, and capture. Traditional perspectives tend to assume that human actors are the sole decision-makers in innovation, which constrains their explanatory power in digital contexts. By contrast, the KSP incorporates a unified NK modeling approach to examine the complex interactions between human agents and knowledge units. This enables a more comprehensive understanding of innovation dynamics in digital ecosystems, addressing the analytical limitations of prior frameworks that overlook the agency of artificial knowledge units.

Second, the KSP offers effective analytical tools for examining the experimental and learning processes of “virtualized” innovation under conditions shaped by DTs. In digital innovation environments, experimentation and learning have progressively shifted from traditional physical world practices to virtual approaches, including digital twins, machine learning, and metaverse-based simulations [98]. The embeddedness of DTs creates a parallel virtual realm that not only mirrors but also dynamically interacts with the physical world in real time. This dual-world configuration has led to the replacement of conventional business model experimentation with sophisticated, simulation-driven processes operating in digital environments. These virtual mechanisms, fueled by continuous data inflow, significantly increase the complexity of experimentation and learning, rendering traditional, physically grounded approaches increasingly impractical. As a result, innovation efforts supported by advanced digital infrastructures have become indispensable. While traditional theoretical perspectives often fall short in capturing the intricacies of such digitally mediated experimentation, the KSP naturally accommodates these complexities. By recognizing the integration of data-rich simulations and continuous learning cycles, this perspective provides a more suitable framework for analyzing the evolving nature of BMI in virtualized settings.

Third, the KSP offers a novel analytical lens and appropriate methodological tools for addressing the critical issue of competitive imitation in DBMI. One of the most significant challenges in DBMI lies in the ease of replication. Due to the high replicability of digital business models, innovations that previously generated Schumpeterian rents are now subject to rapid imitation, leading to the erosion of competitive advantages at an unprecedented rate [91]. A notable example is the OFO bike sharing business model in China, a leading bike sharing platform that lost its market position due to its inability to sustain a competitive edge. Consequently, firms must ensure that their digital competitiveness combines value rarity and inimitability [124]. The traditional perspectives struggle to link the competitive imitation dilemma to the inherent traits of digital business models, limiting their explanatory power. In contrast, the knowledge system perspective allows for a structured analysis of DBMI by examining the degree of knowledge complexity. This perspective provides not only theoretical explanations for why different business models exhibit varying degrees of vulnerability to imitation, but also practical methods for measuring the knowledge complexity of digital business models. This analytical superiority further reinforces the knowledge system perspective as a powerful tool for studying DBMI.

5. Discussion and Conclusions

5.1. Findings

DBMI has become a central topic in contemporary management and innovation research, reflecting its critical role in helping firms to adapt to technological disruptions and environmental complexity. Research on this topic has been published across a wide range of academic journals, reflecting its interdisciplinary relevance to fields such as innovation studies, strategic management, information systems, and entrepreneurship. Despite the growing body of literature, the field remains highly fragmented, with diverse theoretical perspectives and inconsistent terminologies that challenge cumulative knowledge building.

This study addresses this fragmentation and contributes to the literature on DBMI in three important ways. First, we found that the extant literature on BMI has evolved into three major thematic areas: (1) the conceptual foundations of BMI, including definitions, key elements, and underlying logics; (2) the dynamic mechanisms and processes of innovation, especially concerning learning, experimentation, and adaptation; and (3) the performance outcomes and strategic implications of BMI across various industries. Mapping these themes reveals both the progress and persistent conceptual divergence in the field. Second, this study identifies the growing need for new theoretical lenses capable of capturing the complexities of DBMI. Traditional perspectives, such as the activity system or value logic view, offer useful foundations, but fall short in addressing the distributed, adaptive, and knowledge-intensive nature of innovation in digital contexts. In response, we propose the KSP as a novel integrative framework that conceptualizes business models as dynamic, interdependent knowledge architectures. This perspective accommodates both explicit and tacit knowledge contributors, highlighting how knowledge creation, transfer, and recombination underpin DBMI. Third, by synthesizing insights from the reviewed literature and identifying key theoretical tensions, this study proposes an integrative framework and a future research agenda. This framework links DTs, knowledge mechanisms, and innovation outcomes, providing a basis for comparative research on and empirical exploration of DBMI in complex, fast-evolving environments.

Together, these findings help to structure a fragmented field, introduce a forward-looking theoretical approach, and open new avenues for both scholarly inquiry and managerial action in the era of digital transformation.

5.2. Implications for Policy and Practice

This study offers several actionable insights for both policymakers and practitioners seeking to foster DBMI in complex and dynamic environments. For policymakers, enabling DBMI in today’s volatile and interconnected environments requires more than providing infrastructure or funding isolated technological upgrades. It demands a systemic and adaptive policy approach that reflects the dynamic complexity of DBMI and actively supports the cross-boundary flow of knowledge. Unlike traditional innovation trajectories, DBMI often entails rapid reconfigurations of value logic, driven by data, algorithms, and platform ecosystems. For instance, while national strategies—such as the “Digital China” initiative—have laid important groundwork for digital transformation, future policies must go further to cultivate open and knowledge-intensive innovation environments. Policymakers should focus on supporting regulatory sandboxes, cross-industry data platforms, and institutional connectors that facilitate the agile recombination of knowledge assets that foster collaborative learning, data sharing, and digital experimentation.

For Managers, the KSP underscores the strategic value of cultivating rare, inimitable knowledge assets. Firms should embed continuous learning, experimentation, and cross-functional integration into their core business models. This includes developing internal mechanisms for knowledge integration, leveraging AI and data analytics, and engaging in external collaborations with startups, universities, and ecosystem partners. At the same time, digital scalability should be aligned with a differentiated, customer-centric value proposition, ensuring that innovation efforts are both adaptable and market-relevant.

At the ecosystem level, both public and private actors should promote collaborative environments by establishing open standards, encouraging platform interoperability, and developing trust-based data-sharing protocols. Participation in digital ecosystems allows firms to access external knowledge, reduce uncertainty, and co-create innovation in ways that would be difficult to achieve independently. Taken together, these implications highlight that successful DBMI depends on complementary roles: governments must build the enabling environment, while firms must strategically cultivate and apply complex knowledge.

5.3. Future Agenda

As DTs continue to reshape business activities and organizational structures, their integration into BMI presents new challenges and opportunities that merit further academic inquiry. While the existing frameworks have laid important foundations for understanding BMI, the evolving digital landscape calls for more diverse theoretical perspectives and robust empirical studies. In particular, how DTs influence the design, implementation, and transformation of business models—under conditions of complexity, uncertainty, and interdependence—remains an area ripe for exploration. To advance the field, future research can focus on the following three directions:

• Developing an integrated analytical framework: Despite increasing scholarly attention, current research on DBMI remains fragmented and lacks a shared conceptual foundation. Future studies should aim to build an integrated analytical framework that clearly defines key constructs, explicates their interrelations, and outlines the boundaries of DBMI. A unified framework should incorporate both structural and processual elements, reflecting the interplay between digital capabilities, organizational learning, and strategic adaptation;
• Unpacking BMI mechanisms in the digital context: The role of DTs in shaping BMI processes remains underexplored. Existing research often treats tools as static enablers, neglecting the dynamic, iterative processes through which they interact with organizational routines and innovation behaviors. Future studies should examine how DTs co-evolve with experimentation, organizational learning, and systemic adaptation. Multi-level and longitudinal research designs are especially valuable for capturing these interactions over time. Emerging scholarship also suggests the importance of behavioral and team-level dynamics, such as creativity and sustainability orientation, which mediate the relationship between digital affordances and innovation outcomes under uncertain conditions [125]. These insights point to the need for integrating cognitive, social, and technological dimensions into the study of DBMI;
• Developing the knowledge system perspective in complex environments: The KSP offers a promising lens through which to understand DBMI in the context of increasing complexity and dynamic environments. By conceptualizing business models as structured, interdependent knowledge architectures, the KSP emphasizes how knowledge—both human and artificial—is generated, recombined, and applied in innovation processes. This approach captures emerging phenomena such as human–machine collaboration, algorithmic coordination, and distributed learning within digital ecosystems [126,127]. As DBMI becomes more integrated into digital infrastructures, innovation processes are no longer solely driven by humans, but arise from the interplay of distributed knowledge units, both human and artificial, operating within complex digital ecosystems. In this context, the transition from Industry 4.0 to Industry 5.0 highlights the integration of human creativity into AI-enabled systems. Future research should explore how collaborative intelligence and distributed cognition drive business model transformation. Recent research highlights that artificial intelligence greatly improves organizational knowledge processing, especially in the externalization and combination stages, but struggles to replicate the tacit, context-sensitive judgment unique to human cognition [128]. This reinforces the relevance of Nonaka’s SECI model, which emphasizes the dynamic interaction between tacit and explicit knowledge [129]. Future research should examine how AI-based systems and human decision-making jointly shape BMI, and how firms orchestrate human–AI collaboration to co-create adaptive, value-generating business models. This approach extends the KSP by including artificial knowledge agents (e.g., AI systems and algorithmic processes) as active participants in digital business model transformation. This is particularly relevant in the emerging era of human–AI collaboration, where innovation is increasingly shaped by distributed, heterogeneous knowledge sources.

More broadly, the emergence of the KSP offers a more adaptive theoretical lens for understanding DBMI under conditions of growing complexity. By conceptualizing business models as structured, interdependent knowledge systems, the KSP enables the analysis of both the characteristics of individual knowledge components and the interdependencies among them. As DTs become increasingly embedded in business processes, activities such as configuring, integrating, and reconfiguring business model elements demand more nuanced and dynamic forms of knowledge coordination. The KSP enables the exploration of these emerging innovation dynamics—such as human–AI collaboration, distributed cognition, and algorithmic orchestration—highlighting the growing importance of hybrid knowledge systems in shaping the future of DBMI.