Advancing Firm-Level Digital Technology Diffusion: A Hybrid Bibliometric and Framework-Based Systematic Literature Review

Abstract

The rapid proliferation of digital technologies across industries has created significant opportunities for enterprises, yet a comprehensive and integrated review of digital technology diffusion (DTD) at the firm level remains underexplored. This study addresses this gap by systematically analyzing 87 research articles published between 1993 and June 2024 in top-tier management and business journals (ranked A* or A in the ABDC-JQL 2022), sourced from the Web of Science database. Employing a hybrid review methodology that combines bibliometric and framework-based reviews, this study provides a comprehensive overview of the existing research on firm-level DTD. By leveraging the established theories, contexts, methods, antecedents, decisions, and outcomes (TCM-ADO) framework, we consolidate fragmented insights and propose future research directions. This study pioneers the application of the TCM-ADO framework to DTD, offering a novel taxonomy that advances theoretical development.

1. Introduction

The rapid advancement in digital technologies, including big data analytics, artificial intelligence, blockchain, and machine learning, has increasingly positioned digitalization as a catalyst for business model innovation [1,2]. These technologies are reshaping value propositions and fostering the creation of new business models by introducing novel elements into industrial and organizational contexts [3]. While the initial adoption of digital technologies is important, it is the widespread diffusion of these technologies across organizations that truly unlocks their transformative potential and facilitates business model innovation [4,5]. Over recent decades, the diffusion of digital technologies has accelerated, creating significant business opportunities, including innovative entrepreneurship [6], digital transformation in established firms [7], and the emergence of new business models like omnichannel strategies [8].

Research on digital technology diffusion (DTD) spans multiple levels of analysis. Some studies focus on individual-level effects, such as the role of word-of-mouth and conspiracy beliefs in app adoption [9], while others examine broader impacts at the industry, regional, or national levels, such as the influence of DTD on innovation performance in China’s high-tech sector [10]. This paper, however, centers on the firm level, a scope that balances granularity with analytical breadth. Unlike individual-level adoption, organizational diffusion involves complex strategic decisions about resource allocation, stakeholder engagement, and business model adaptation [11]. Firms are pivotal in driving economic transformation and technological shifts [3], and digital technologies differ from traditional technologies in their connectivity, complementarity [12], generativity, flexibility [1], immediacy of interaction [13], and disruptive potential [3,5,14]. These unique characteristics necessitate a distinct focus on DTD at the firm level, warranting further exploration.

The literature on DTD has evolved alongside technological advancements, shifting from early studies on e-commerce diffusion [15] to contemporary research on next-generation technologies [4]. While some scholars have examined digital technologies broadly [1,6], others have focused on specific stages of diffusion, such as adoption [16] or multi-stage implementation processes [11]. Recent studies have narrowed their focus to specific technologies, exploring topics like AI adoption [17,18,19], Metaverse usage intentions [20], B2B digital platform adoption [21], and digital transformation [22]. Despite this growing body of research, the literature remains fragmented, underscoring the need for a comprehensive systematic review to consolidate past findings and identify future research directions. Although some reviews have mapped specific areas, such as the diffusion of smart circular economy technologies [23], a gap persists in synthesizing research on general digital technology diffusion. This paper addresses this gap by focusing on DTD at the firm level, offering a broader and more integrated perspective.

This study systematically synthesizes the literature on DTD at the firm level, using a hybrid approach that integrates bibliometric analysis with a framework-based review. We analyzed 87 articles published before June 2024 in top-ranked management and business journals (ranked A* or A in the ABDC-JQL 2022) to map the intellectual structure of this field, identifying key articles, journals, scholars, countries, and themes. By employing the theory–context–method (TCM) and the antecedent–decision–outcome (ADO) frameworks [24], we organize and develop a comprehensive model and future research directions on DTD in terms of its theories, methods, contexts, antecedents, decisions, and outcomes.

The paper is structured as follows: Section 2 outlines the theoretical background. Section 3 details the bibliometric and TCM-ADO framework-based review, along with the data source and analysis. Section 4 presents findings from the bibliometric analysis, while Section 5 synthesizes the literature on DTD using the TCM-ADO framework. Section 6 identifies the future research directions, and Section 7 concludes with the theoretical and managerial implications, as well as the research limitations.

2. Theoretical Background

2.1. Understanding Digital Technology Diffusion

DTD represents a significant evolution from traditional technology diffusion, driven by rapid advancements in big data analytics, the Internet of Things, artificial intelligence, blockchain, and machine learning. These technologies exhibit distinct diffusion patterns, diverging from conventional models. This paper views DTD through the dual lenses of ‘digital technology’ and ‘diffusion’, analyzing these concepts both independently and interactively.

Digital technologies function as strategic resources and technical assets that drive organizational innovation [1,25]. They reduce transaction costs, alleviate coordination challenges, and diminish information asymmetry [26], transforming business operations and fostering sustainable growth [1]. By enhancing operational efficiency and collaboration, they enable innovative business models that improve profitability [27]. As digitalization progresses, the value of these technologies emerges not merely from their adoption but from their widespread diffusion [1], which is essential for achieving significant economic impact [28].

Rogers’s seminal work in 1995 delineates diffusion as the process by which an innovation spreads over time among members of a social system through various channels [29]. At the enterprise level, diffusion involves a dynamic process that propagates new principles and behaviors within and between firms [17,30,31]. Building on this framework, Hassan et al. [1] conceptualize digital diffusion as the strategic integration of digital technology features with a firm’s internal capabilities to generate value in dynamic environments. Drawing on these insights, this paper defines DTD at the firm level as the propagation of digital technologies within and across firms.

2.2. Features of Digital Technology Diffusion

Digital technologies exhibit distinct characteristics that differentiate their diffusion from non-digital technologies [13]. Recognized as general-purpose technologies [3,5,18], they possess broad applicability across diverse sectors, reflecting their dynamic nature [32]. This universal relevance spans industries such as electronics [33], mobility services [34], transportation, and medical technology [35], profoundly influencing their development.

A defining feature of digital technologies is modularity. Modular plug-in units enhance system flexibility and interoperability, enabling the integration of new functionalities without overhauling the existing architectures [36]. This modularity reduces adoption costs, accelerates diffusion, and simplifies upgrades, which was instrumental in the rapid spread of early digital technologies [37].

Furthermore, digital technologies are characterized by ubiquitous connectivity and real-time interaction [13]. These attributes facilitate inter-firm communication and information sharing, enhancing networking capabilities and dissolving traditional firm boundaries, potentially fostering digital ecosystems [27,38]. In a digitalized world, knowledge spillovers are less constrained by geographical proximity [18]. Network effects [29,39] amplify the value of these technologies as adoption grows, creating positive feedback loops that drive further diffusion. DTD, propelled by network effects at the micro and macro levels [5,18], represents a dynamic, self-reinforcing process [5].

2.3. Phases of Digital Technology Diffusion

Rogers [29] highlights that diffusion is a dynamic process that occurs over time. The DTD progresses through several stages [22], which include desirability, feasibility, initial trial, implementation, and sustenance, as detailed by Steiber et al. [40]. These stages can be broadly categorized into two primary phases: adoption and use. In the context of artificial intelligence, McElheran et al. [19] and Ameye et al. [17] underscore the importance of these phases in the successful diffusion of emerging technologies.

The adoption phase begins when an organization identifies a problem solvable through digital technologies [41,42]. This is followed by a comprehensive evaluation of potential benefits and risks from technical, financial, and strategic perspectives [22], informing the adoption decision [42]. Successful evaluations lead to technology adoption aimed at creating organizational value [34]. Rogers [43] delineates this process into five steps: knowledge, persuasion, decision, implementation, and confirmation. Krieger et al. [42] expand this framework to include six activities: ideation, evaluation, resource commitment, solution building, deployment, and operational use. This phase has been extensively studied across various contexts, from general digital technologies [16] to specific applications like artificial intelligence [18,44], blockchain, IoT [45], and advanced data analytics [42].

The usage phase focuses on broad deployment and routinization within the organization, where significant business value is realized [22]. This involves integrating digital technologies into the value chain and aligning them with organizational needs and structures [34]. It enhances productivity; drives innovation in products, services, and business models; and fosters collaboration with digitally capable firms or requiring digital support, promoting industry upgrades and convergence [17,46]. Over time, these technologies become embedded in routine operations, losing their novelty, and paving the way for new innovation cycles [42].

Although adoption may occur at a specific moment, technology use is an ongoing process that stabilizes within the firm [17]. The transition from initial adoption to widespread diffusion can experience significant delays [47]. Legitimacy is crucial throughout these stages: during adoption, organizations legitimize technological change internally and externally, while in the usage phase, they consolidate decisions by gathering positive external feedback [18].

In conclusion, DTD is a multifaceted and dynamic process that integrates digital technologies with organizational strategies to drive sustainable innovation and growth.

3. Methodology

3.1. Hybrid Reviews

Systematic reviews represent a scientific approach to secondary research that utilizes reproducible methods to identify, select, and evaluate prior studies in response to specific research questions [48]. These reviews are typically categorized into four broad types: domain-based, theory-based, method-based, and meta-analytical reviews [24,48]. Domain-based reviews, which concentrate on specific topical areas, can be conducted through structured, framework-based, bibliometric, hybrid, or theory development-based approaches [24,48]. Hybrid reviews, which combine elements of two or more review types, offer a comprehensive examination of the research landscape. Our study employs a hybrid review approach, integrating bibliometric and framework-based reviews.

A bibliometric review involves the statistical analysis of published articles, providing an objective and quantitative assessment of the research corpus [49]. This method enhances data impartiality in academic research and facilitates detailed, rational analysis through visualization techniques [50,51]. In contrast, a framework-based review is structured around a specific conceptual framework, which is often widely accepted due to its robust structured approach [48]. Numerous frameworks are available for such reviews, including the TCM framework (theories, contexts, and methods) [52], the TCCM framework (theories, constructs, characteristics, and methods) [53], the ADO framework (antecedents, decisions, and outcomes) [54], and the integrated TCM-ADO framework [24].

Combining bibliometric analysis with a framework-based review enhances the comprehensiveness, rigor, and contribution of the research [49]. Scholars frequently employ such hybrid approaches. For instance, Ben Jabeur et al. [55] combined bibliometric analysis with a 4Ws framework (what, where, why, and how) to explore AI and machine learning research in fake review detection. Similarly, Sahaf and Fazili [56], and Chakma et al. [57] integrated bibliometric analysis with the TCCM framework to, respectively, explore service failure and recovery, and organizational ambidexterity. Kacprzak and Hensel [58] employed bibliometric analysis alongside the ADO framework to investigate online customer experience. This study employs a hybrid review that merges bibliometric analysis with the integrated TCM-ADO framework.

3.2. Data Source

We utilize the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement to guide our data search and refinement processes (see Figure 1). The PRISMA framework enhances the reliability and reproducibility of systematic reviews and has been widely adopted in numerous review studies [24,59,60].

This study adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement to ensure methodological rigor and reproducibility in data search and refinement processes (see Figure 1). PRISMA is widely recognized for enhancing the reliability of systematic reviews and has been extensively applied in prior research [24,49,59].

For our database selection, we chose the Social Sciences Citation Index (SSCI) and Science Citation Index Expanded (SCIE) from the Web of Science (WoS) Core Collection. Based on recommendations by Paul et al. [61] and García-Lillo et al. [59], we opted for a single database to minimize inefficiencies caused by overlapping records and avoid the complexity of consolidating journal titles. In addition, WoS is regarded as the benchmark database for evaluating scholarly performance.

The initial search focused on firm-level DTD publications, with a cut-off date of June 2024. Using the TOPIC field, three search strings were employed: (1) ‘technology diffuse*’, (2) ‘digital’, and (3) ‘firm’. The search began with ‘technology diffuse*’ within the Management and Business categories, limited to English-language papers. Building on methodologies from Cho et al. [4] and Kohtamäki et al. [38], the search was expanded to include terms such as digitization, artificial intelligence (AI), Internet of Things (IoT), smart solutions, smart products, 5G, 6G, and connectivit* to refine the focus on DTD. Firm-level relevance was ensured by incorporating ‘firm’, ‘enterprise’, and ‘company’, resulting in an initial pool of 161 documents.

During screening, only journals rated A* or A in the ABDC-JQL 2022 ranking were retained to ensure scientific rigor, reducing the dataset to 117 articles. In the eligibility phase, a full-text review was conducted, excluding papers misaligned with the research topic or lacking firm-level focus. For example, some studies mentioning firms primarily addressed country- or individual-level analyses. Ultimately, 87 documents were included in the review.

3.3. Data Analysis

This study employs a hybrid review approach, integrating a bibliometric analysis with a TCM-ADO framework-based review to examine DTD at the firm level. First, we conducted a bibliometric analysis to obtain quantitative insights into the research field, such as the annual publication trends. Using the Bibliometrix and Biblioshiny R-packages [62], we mapped the intellectual structure of this field by identifying key articles, journals, scholars, and countries, and performing co-citation and co-occurrence analyses.

We then applied the TCM-ADO framework to explore the theories, contexts, methods, antecedents, decisions, and outcomes related to DTD, offering a comprehensive and structured review. The TCM framework organizes foundational research elements by categorizing theories that guide research perspectives, contexts that describe the specific circumstances under investigation, and methods that provide empirical evidence [24,52]. Complementarily, the ADO framework structures past research findings into antecedents that explain the causes of behavioral engagement or avoidance, decisions that outline the actions taken or not taken, and outcomes that evaluate the consequences of these actions [24,54]. This integrated approach leverages the strengths of both frameworks, offering a robust structure for reviewing DTD. Its growing popularity is evident in various research domains, including studies on tacit knowledge in organizations [63], cynical consumer behavior [64], and value co-creation behaviors [60].

During data analysis, the research team (the authors and two collaborators) systematically extracted and organized key elements from the full texts of 87 academic papers, including theories, contexts, methods, antecedents, decisions, and outcomes. To ensure accuracy and consistency, we adopted a cross-validation approach, conducting multiple rounds of independent coding, mutual verification, and in-depth discussions until a full consensus was reached.

4. Findings from Bibliometric Analysis

In this section, we employed the Bibliometrix and Biblioshiny R-package [40] to conduct a comprehensive bibliometric analysis. This analysis provides descriptive results, including publication years, leading scholars, key countries, influential journals, and articles. Additionally, we performed a co-citation analysis to generate historiographic mapping and a co-occurrence analysis to identify clusters.

4.1. Descriptive Findings

4.1.1. Overview of Main Information

Table 1. Main information about literature.

Description	Results
Main information about data
Timespan	1993:2024
Sources (Journals, Books, etc.)	39
Documents	87
Annual Growth Rate %	9.57
Document Average Age	6.8
Average Citations per Doc	72.4
References	6527
Document contents
Keywords Plus (ID)	350
Author’s Keywords (DE)	405
Authors
Authors	259
Authors of Single-Authored Docs	6
Authors collaboration
Single-Authored Docs	7
Co-Authors per Doc	3.11
International Co-Authorships %	45.98

provides a comprehensive overview of the literature on DTD at the firm level, spanning from 1993 to June 2024, with specific publication trends depicted in Figure 2. This review identifies 87 articles across 39 distinct sources, each averaging 72.4 citations and a document age of 6.8 years. The literature has experienced an annual growth rate of 9.57%. Collectively, these articles cite 6527 sources. The research, contributed by 259 authors, demonstrates extensive collaboration, with an average of 3.11 authors per article and a 45.98% rate of international co-authorship. Notably, the annual distribution of publications remained below five per year until the last five years, when there was a significant surge, now representing 62% of the total publications. This recent rise underscores the increasing scholarly interest in DTD and its growing relevance as a research area.

4.1.2. Most Influential Authors and Countries

As shown in Figure 3, influential authors are identified based on the number of publications and local citations. The top contributors in this area are Kraemer KL, Zhu K, Li SL, Shaheer N, and Lyytinen K. According to Lotka’s Law, only 12 authors have published two papers, while the remaining authors, who account for 95.4% of the total, have written just one paper each. This trend reflects the relatively new nature of the DTD field.

Table 2. Production by countries.

Country	Freq.	SCP	MCP
USA	25	19	6
Italy	10	5	5
Germany	8	5	3
China	7	3	4
France	4	1	3
Finland	3	1	2
India	3	1	2
New Zealand	3	3	0
United Kingdom	3	3	0
Australia	2	0	2
Canada	2	0	2
Hong Kong	2	2	0
Korea	2	0	2
Netherlands	2	0	2
Switzerland	2	1	1

Note: Freq. denotes frequency, SCP denotes single country publication, and MCP denotes multiple countries publication.

presents the most influential countries based on total publications, single-country publications, and multiple-country collaborations. The USA leads with 25 papers, followed by Italy with 10, Germany with 8, and China with 7. These four countries, each having more than five publications, also demonstrate a high level of international collaboration, as indicated by multiple-country publications. Further details on the collaboration between these four countries are illustrated through the co-authorship analysis in

Table 3. Worldwide collaboration between countries.

From	To	Freq.	From	To	Freq.
USA	China	4	Italy	Russia	1
USA	Australia	2	Italy	Slovakia	1
USA	Canada	2	Italy	Sweden	1
USA	Israel	2	Italy	United Kingdom	1
USA	Italy	2	Total		9
USA	Korea	2	China	Korea	3
USA	United Kingdom	2	China	Australia	1
USA	Denmark	1	China	Denmark	1
USA	Finland	1	China	France	1
USA	France	1	China	India	1
USA	India	1	China	Romania	1
USA	Singapore	1	Total		6
USA	Sweden	1	Germany	Netherlands	2
Total		13	Germany	Switzerland	2
Italy	France	2	Germany	Belgium	1
Italy	China	1	Germany	Denmark	1
Italy	Denmark	1	Germany	Estonia	1
Italy	India	1	Germany	Finland	1
Italy	Netherlands	1	Total		6

Note: Freq. denotes frequency.

.

4.1.3. Most Influential Journals and Articles

Table 4. Most influential journals.

Journal	h-Index	Freq.	TC
Technological Forecasting and Social Change	9	13	652
Industrial Marketing Management	6	6	235
Technovation	5	5	116
Information Systems Research	4	4	1176
IEEE Transactions on Engineering Management	3	6	152
Management Science	3	4	612
Research Policy	3	4	291
Small Business Economics	3	3	22
Journal of Enterprise Information Management	2	4	61
Organization Science	2	2	674
Journal of Product Innovation Management	2	2	245
Information & Management	2	2	130
Journal of Management Information Systems	2	2	72
International Journal of Accounting Information Systems	2	2	49
Electronic Markets	2	2	46
Journal of Economics & Management Strategy	2	2	7

Note: Freq. denotes frequency and TC denotes total citation.

presents the most influential journals, sorted by h-index, number of articles, and total citations. Technological Forecasting and Social Change ranks highest with an h-index of nine and 13 published papers. Industrial Marketing Management follows, with an h-index of six and 6 publications. Technovation has an h-index of five, while Information Systems Research stands out with an h-index of four and the highest total citations. Additionally, IEEE Transactions on Engineering Management, Management Science, Research Policy, and Small Business Economics each have an h-index of three.

As shown in

Table 5. Most influential papers.

Document	LC	NLC	GC
Innovation diffusion in global contexts: determinants of post-adoption digital transformation of European companies [65]	4	1.00	390
Post-adoption variations in usage and value of e-business by organizations: Cross-country evidence from the retail industry [15]	3	2.25	874
Artificial intelligence and industrial innovation: Evidence from German firm-level data [66]	2	10.00	57
The international penetration of ibusiness firms: Network effects, liabilities of outsidership and country clout [67]	2	3.00	118
What’s driving the diffusion of next-generation digital technologies? [4]	1	10.00	28
Managing Digital Transformation: Scope of Transformation and Modalities of Value Co-Generation and Delivery [68]	1	3.50	75
Revisiting Location in a Digital Age: How Can Lead Markets Accelerate the Internationalization of Mobile Apps? [69]	1	3.50	30
Applied artificial intelligence and trust-The case of autonomous vehicles and medical assistance devices [35]	1	2.00	334
The role of users and customers in digital innovation: Insights from B2B manufacturing firms [46]	1	2.00	116
Network Interconnectivity and Entry into Platform Markets [70]	1	1.00	475

Note: LC denotes local citations, NLC denotes normalized local citations, and GL denotes global citations.

, the sample papers are sorted according to local citations, normalized local citations, and global citations. Two papers authored by Zhu K hold the top positions in local citations, with four citations for the paper by Zhu et al. [65], and three citations for the paper by Zhu and Kraemer [15]. Both the papers by Rammer et al. [66] and Chen et al. [67] have received two local citations each.

4.2. Co-Citation and Co-Occurrence Analysis

4.2.1. Historiographic Mapping

Figure 4 illustrates the historiographic mapping of 27 nodes based on co-citations. Each node represents a document cited by other documents within the analyzed sample, while each edge indicates a direct citation. This visual representation facilitates the identification of historical paths, which highlight distinct research topics and their foundational authors or documents [62]. Notably, the works of Adner and Levinthal [70], Forman [37], Zhu and Kraemer [15], Boland et al. [71], Baird et al. [72], Hengstler et al. [35], Cennamo et al. [68], and Rammer et al. [66] have contributed to the development of new research themes.

4.2.2. Co-Occurrence Analysis

Figure 5 depicts the co-occurrence network generated from the abstract using bigrams to capture detailed insights. The network is constructed based on the Louvain clustering algorithm, with the number of nodes set to 200. Following the methodology of Shi and Shen [50], commonly used but non-informative phrases (e.g., ‘future research’, ‘study examines’, and ‘extant literature’) are removed, and synonyms are consolidated to unify keyword variants, such as ‘artificial intelligence’ and ‘AI’.

The analysis identifies seven distinct clusters. The largest, shown in red, emphasizes digital technologies, while the second-largest, in blue, focuses on digital transformation and supply chain management. The green cluster centers on artificial intelligence, and the brown cluster explores communication technologies. The orange cluster highlights innovation diffusion and its influencing factors, the pink cluster investigates adoption intentions, and the purple cluster examines the technology adoption process.

The nodes in these clusters can be categorized into several groups: theories (e.g., ‘DOI theory’ and ‘TOE framework’ in the pink cluster), contexts (e.g., ‘international business’ in the brown cluster), methods (e.g., ‘structural equation modeling’ in the blue cluster), antecedents (e.g., ‘adoption intention’ in the pink cluster), processes (e.g., ‘adoption process’ in the purple cluster), and outcomes (e.g., ‘sustainability performance’ in the green cluster). In Section 4, we will conduct a detailed analysis of these nodes based on the TCM-ADO framework through content analysis following a comprehensive review of the full text.

5. TCM-ADO Framework-Based Review of DTD

In this and the following section, we apply the TCM-ADO framework to analyze its theories, contexts, methods, antecedents, decisions, and outcomes based on the existing literature. An overview of these elements is presented in Figure 6. Our analysis is confined to the firm level, excluding individual, regional, or national perspectives to focus on organizational research.

5.1. Theories

Theory is a cornerstone of academic progress [60], guiding scholars in addressing research questions and fostering communication through a unifying language [73]. The theoretical foundations of innovation adoption research differ between individual and firm levels [74]. At the individual level, the technology acceptance model (TAM) [75] and the unified theory of acceptance and use of technology (UTAUT) [76], along with their extensions, are widely used to analyze innovation adoption behaviors [42,77]. In contrast, at the firm level, the diffusion of innovation (DOI) theory [78] and the technology–organization–environment (TOE) framework [79] dominate [17,74]. Our review identifies 42 theories in the business and management literature on DTD, with DOI and TOE being the most prevalent (see

Table 6. List of theories used in the literature [1,3,6,7,11,18,20,21,22,23,26,28,31,35,37,41,42,44,45,46,72,77,80,81,82,83,84,85,86,87,88,89,90].

Theories	Samples	Freq.	Theories	Samples	Freq.
Diffusion of innovation	[20,31,41]	19	Intellectual capital theory	[85]	1
Technology-organizational-environment framework	[45,80,81]	11	Knowledge spillover theory of entrepreneurship	[6]	1
Contingency theory	[44,72]	5	Lead user theory	[37]	1
Dynamic capability theory	[28,83]	4	Neo-institutional theory	[82]	1
Knowledge-based view	[18,46]	3	Network theory	[84]	1
Institutional theory	[23]	2	Organizational information processing theory	[22]	1
Internalization theory	[26]	2	Organizational or structure theory	[11]	1
Resource based view	[11]	2	Organizational resilience theory	[85]	1
Resource dependence theory	[84]	2	Plural governance theory	[86]	1
Social network theory	[21]	2	Signaling theory	[87]	1
A mid-range process theory	[42]	1	Social capital theory	[18]	1
Absorptive capacity theory	[83]	1	Stakeholder theory	[23]	1
Adaptive structuration theory	[88]	1	Structuration theory	[88]	1
Adoption of innovations	[88]	1	Task-technology fit theory	[88]	1
Competence-based management	[3]	1	Technological learning and catching-up	[89]	1
Critical success factors theory.	[44]	1	Technology acceptance model	[35]	1
Customer experience theory	[90]	1	Technology affordances and constraint theory	[1]	1
Effectuation theory	[77]	1	The geographic concentration theory	[37]	1
Hoffman and Novak’s flow model	[90]	1	Theory of institutional entrepreneurship	[7]	1
Industrial-organizational view of the firm	[11]	1	Triple bottom line theory	[44]	1
Information systems success model	[90]	1	Trust commitment theory	[90]	1

Note: Freq. denotes frequency.

).

The DOI theory, also known as the innovation diffusion theory (IDT) [91] or the theory of diffusion of innovations [3], explains what, how, why, and at what rate new ideas, technologies, and process innovations spread over time [78,92]. It addresses four key questions: what (innovation itself), how (innovation process and communication channels), why (reason for the adoption or rejection), and at what rate (time to critical mass) [74,90,92]. The theory highlights four key elements: perception of innovation attributes (such as relative advantage, compatibility, complexity, trialability, and observability), communication channels, the period over which diffusion occurs, and the social system [29,78]. These innovation attributes, particularly the relative advantage, compatibility, and complexity, significantly influence organizational adoption and diffusion rates [80]. For example, Kumar et al. [20] applied DOI to explore metaverse adoption, while others integrated it with the organizational information processing theory [22], contingency theory [72], triple bottom line theory and critical success factors theory [44], trust commitment theory, information systems success model, Hoffman and Novak’s Flow Model, and customer experience theory [90].

The TOE framework, introduced by Tornatzky and Fleischer [79], examines technological, organizational, and environmental contexts influencing innovation adoption. Recognized as a generic theory [15,39,81], it is particularly appropriate for use in new technology studies [11]. It consists of three dimensions: firstly, the technological context includes the technologies that organizations are currently using, those available in the market but not yet adopted, and the related technological capabilities. Secondly, the organizational context examines internal organizational characteristics such as size, behavior, top management support, and slack resources. Finally, the environmental context encompasses external factors such as industry characteristics, competitors, trading partners, and regulatory requirements [11,18,41,42,65,79]. This framework provides a flexible lens for analyzing adoption strategies [16,74,80]. For instance, Ameye et al. [17] applied TOE to AI adoption, while Yeh and Chen [81] expanded it to include cost dimensions for 3D printing adoption.

The TOE framework and DOI theory are often combined to study organizational technology adoption [42]. While both emphasize internal and external organizational factors, TOE uniquely incorporates the business environment [15,21,41]. This integration offers a comprehensive approach to understanding innovation adoption, as seen in studies on e-business [65], blockchain and IoT [45], digital reporting [41], and AI [80]. Given the dominance of these two theories, we utilize them to organize the antecedents of DTD, as illustrated in Section 5.1. Additionally, some scholars have combined the DOI model, the TOE framework, and other theories. For example, Marzi et al. [21] combined the DOI, TOE, and social network theory to analyze B2B digital platform adoption.

In addition to these two theories, the contingency theory and dynamic theory are mentioned several times. The contingency theory provides a perspective that explains how varying conditions impact organizational behavior, bridging the gap between the ‘one best way’ approach and the ‘it all depends’ viewpoint [93]. This theory is founded on two key insights: firstly, there is no singular, optimal way to manage a company; secondly, not all management methods are equally effective [94]. Generally, a better alignment between organizational practices and contingency variables enhances an organization’s ability to innovate and, ultimately, its overall effectiveness [72,95]. For example, Baird et al. [72] used it to study patient portal adoption in outpatient clinics, while Rauniar et al. [11] integrated it with TOE to examine digital technology adoption and implementation.

The dynamic capability theory emphasizes firms’ capabilities to acquire and allocate resources, enabling them to adapt to market changes and implement new strategies [96,97]. Through sensing, seizing, and reconfiguring, firms enhance responsiveness and flexibility, achieving sustainable competitive advantages [28,50,98]. In the digital era, capabilities such as digital transformation and innovation are critical. Chaudhuri et al. [83] note that these capabilities enable organizations to continually update their innovation processes and rapidly introduce market-aligned products. They also enhance data acquisition and analysis, improving performance in dynamic environments. Truant et al. [23] apply this theory to identify drivers and barriers in smart circular economy implementation, while Cumming et al. [83] highlight the role of digital sensing, capture, and transformation in enabling cross-border e-commerce.

In summary, DOI and TOE are central to understanding DTD at the firm level, often complemented by the contingency and dynamic capability theories to provide a holistic perspective on innovation adoption and diffusion.

5.2. Contexts

Contexts refer to the conditions under which research is conducted [24,52,73,99]. Building on prior studies (e.g., [63,73]), this review examines two primary contexts across the 87 articles: industries and countries. As shown in

Table 7. List of industries studies in the literature [11,13,15,16,21,22,27,28,33,34,35,42,45,46,67,71,72,81,83,86,90,100,101,102].

Industry Types		Freq.	Samples
Secondary sector of industry (n = 13)
Mining industry	Oil and gas industry	1	[11]
Manufacturing	Manufacturing firms	7	[21,22,28,46,81,86,100]
	Firms in industrial districts	1	[16]
	Food industry	1	[27]
	Thermoelectric generators	1	[101]
Electricity, heat, gas and water production and supply industries	Electronics sector	1	[33]
Construction industry	Architecture, engineering, and construction	1	[71]
Tertiary sector (n = 9)
Services industries	Logistics and supply chains	1	[45]
	Last-mile delivery	1	[102]
	Social media platforms	1	[90]
	Audit firms	1	[42]
	Publisher	1	[13]
	Health care	1	[72]
	Mobility services	1	[34]
	I-business firms	1	[67]
	Retail Industry e-business	1	[15]
Mixed (n = 2)
Services industries and manufacturing	Service organizations and manufacturing organizations	1	[83]
	Transportation and medical technology industries	1	[35]

Note: Freq. denotes frequency.

, industries are classified into secondary (n = 13) and tertiary sectors (n = 11) based on China’s National Economic Industry Classification (GB/T 4754-2011), with manufacturing being the most studied subsector (n = 7). Geographically, as outlined in

Table 8. List of countries studied in the literature [1,3,4,18,21,27,44,66,81,83,91,100,103,104,105,106,107].

Country	Freq.	Samples
United States	10	[103,104]
China	7	[81,91]
Germany	4	[1,66]
Italy	2	[105]
United Kingdom	2	[21]
Austria	1	[18]
Danish	1	[44]
Finland	1	[106]
Greece	1	[27]
New Zealand	1	[107]
Indian	1	[83]
Irish	1	[100]
Russian	1	[3]
South Korean	1	[4]
Switzerland	1	[18]

Note: Freq. denotes frequency.

, the United States (n = 10), China (n = 7), and Germany (n = 4) dominate in research volume. Additionally, studies have focused on firms of varying sizes, including small- and medium-sized enterprises [1,41,85], startups [6], and large firms [3].

5.3. Methods

Methods refer to the systematic processes for collecting and analyzing data in empirical research [60]. Drawing on previous studies [24,60], this review examines 87 articles from two perspectives: research approach and data type. Research approaches in DTD studies are diverse, including quantitative methods, qualitative analyses, literature reviews, modeling and simulation, and multiple methods (see

Table 9. List of research approaches used in the literature [3,4,6,13,16,18,20,21,33,44,45,70,74,81,82,87,88,90,105,106,107,108,109,110,111,112,113,114,115].

Methodologies	Methods	Freq.	Samples
Quantitative (n = 41)	Regression analysis (e.g., methods = Cox proportional hazard model, cross-sectional two-period difference, discriminant, event study, logistic, multinomial probit, negative binomial, ordinary least squares, panel, parametric hazard model, Poisson, probit, proportional hazards, time series, tobit, and weighted probit)	29	[4,6,105]
	Structural equation modeling (e.g., method = covariance-based, partial least squares)	9	[16,108]
	A fuzzy analytic hierarchy process	1	[81]
	Diffusion model (e.g., Bass model, one-way effects model, and two-way effects model)	1	[109]
	Machine learning algorithm (e.g., DBSCAN algorithm)	1	[110]
Qualitative (n = 16)	Case study	11	[33,111]
	Qualitative approach with interview	2	[82]
	Thematic analysis	1	[74]
	Quasi-ideal natural experiment	1	[13]
	Fuzzy-set qualitative comparative analysis (fsqca)	1	[21]
Literature review (n = 11)	System literature review	10	[112,113]
	Research Commentary	1	[88]
Multiple/Mixed method (n = 9)	SEM and artificial neural network (ANN) analysis	1	[20]
	Transformer language model and regression analysis	1	[18]
	Fuzzy-Delphi technique, fuzzy-decision-making trial and evaluation laboratory tool, graph theory matrix approach, and sensitivity analysis	1	[45]
	Gray DEMATEL and case study	1	[44]
	Discrete-time, stochastic dynamic program, and numerical study	1	[102]
	Explorative vector autoregression analysis and content analysis	1	[3]
	Content analysis and lasso and ridge regression	1	[90]
	Qualitative content analysis, cluster analysis, ordinal logistic, and negative binomial regression analyses	1	[115]
	OLS linear regression, unsupervised topic modeling, and deep learning	1	[87]
Modeling and Simulation (n = 2)	A formal computer simulation model	1	[70]
	Computational simulation	1	[114]

Note: Freq. denotes frequency.

). Nearly half of the studies (n = 47) employed quantitative methods, primarily regression analysis (n = 29) and structural equation modeling (n = 9). Qualitative approaches were used in 16 articles, with case studies (n = 11) being the most common. Additionally, 11 articles conducted literature reviews, 9 applied multiple or mixed methods, and 2 focused on simulation techniques.

Regarding data types, DTD studies balance primary and secondary data (see

Table 10. List of research data used in the literature [1,3,7,8,16,18,20,31,66,67,69,87,101,102,110,111,116,117,118].

Data Types	Data	Freq.	Samples
Primary (n = 38)	Survey	20	[16,20]
	Interviews (including semi-structured interviews, online interviews, and expert interviews)	17	[101,111]
	Firm transaction data	1	[102]
Secondary (n = 39)	Publications/literature	12	[8,116]
	Government data and report	11	[66,117]
	Research institution data	5	[1,118]
	Annual reports of listed companies	5	[7,31]
	Patents data	2	[87,110]
	Self-collected open web data sources	2	[3,18]
	Apple’s app store Publicly available sources	2	[67,69]

Note: Freq. denotes frequency.

), with 38 and 39 articles, respectively. For primary data, 20 articles used surveys, such as Likert-scale questionnaires, while 17 relied on interviews (e.g., semi-structured, online, and expert) for exploratory research. Secondary data sources included publications/literature (n = 12), government data and reports (n = 11), research institution data (n = 5), and annual reports of listed companies (n = 5).

5.4. Antecedents

Antecedents are factors that influence the decision to adopt or reject a behavior, directly affecting decisions or indirectly shaping outcomes [54,60,99]. These factors are critical as they establish path-dependent relationships, shaping early diffusion patterns and the distribution of technological opportunities [18]. As outlined in Section 4.1, the DOI theory and TOE framework are widely used to identify these antecedents. This study integrates these perspectives to explore factors influencing DTD, focusing on innovation attributes from the DOI theory, such as relative advantage, compatibility, and complexity.

5.4.1. Technological Antecedents

Technological antecedents drive enterprises to adopt and promote innovation [28]. As summarized in

Table 11. Technological antecedents [1,3,5,11,12,15,16,17,18,20,21,27,35,37,40,41,42,45,65,66,74,80,87,119,120,121].

Types	Antecedents	Articles
Attributes of digital technology	Relative advantage	[20,21,65,80,119]
	Compatibility	[20,35,65,80,119]
	Trialability	[20,35,65]
	Observability	[20]
	Technology complementarity	[12,37]
	Technology standardization	[40,87]
	Codified knowledge	[18,120]
	Technology trust	[35,74]
	Usability	[35]
	Complexity (-)	[20,41,80,119,121]
	Costs (-)	[5,27,37,41,65]
	Technology uncertainty (-)	[17,66]
	Security concern (-)	[65]
Technology capabilities	Technology competence	[12,15,41,42,65]
	Digital technology maturity	[1,3,16,17,27,45]
	Digital agility	[20,45]
	Absorptive capacity	[11]
	Digital infrastructure	[1]

Note: A negative sign in parentheses indicates an inhibitor.

, 18 technological antecedents are categorized into two groups: attributes of digital technology (n = 13) and technology capabilities (n = 5).

Attributes of Digital Technology

These attributes, derived from Rogers [29], include relative advantage, compatibility, trialability, observability, and complexity. Relative advantage refers to the perceived benefits of new digital technologies, such as artificial intelligence and the Metaverse, over the existing solutions [20,80]. High compatibility with the existing systems, values, and practices facilitates adoption [35,65]. Additional factors like technology complementarity, standardization, codified knowledge, trust, and usability enhance adoption appeal, while complexity may hinder adoption if the technology is perceived as difficult to use [41,121]. Implementation costs and organizational adjustment expenses can also negatively impact adoption [27,65], though declining costs over time accelerate the diffusion of new technologies [5].

Technology Capabilities

This category includes technology competence, digital technology maturity, digital agility, absorptive capacity, and digital infrastructure. Technology competence reflects a firm’s internal technology resources for adopting new technologies [41,79], significantly influencing e-business adoption [15]. Digital technology maturity refers to an organization’s accumulated digital expertise [3], which shapes digital affordances and learning trajectories, even among firms using similar technologies [1]. Firms with higher digital maturity or early adoption are more adept at leveraging AI [17]. Digital agility denotes an organization’s ability to swiftly identify and capitalize on emerging opportunities through digital technologies [20], enhancing its intent to adopt and exploit competitive advantages [20].

5.4.2. Organizational Antecedents

Organizational contexts significantly influence technology adoption and outcomes [19]. As outlined in

Table 12. Organizational antecedents [4,11,13,16,17,18,19,23,26,27,37,40,41,42,72,80,90,119,122].

Types	Antecedents	Articles
Organizational characteristics	Firm size (bigger organizations)	[4,17,18,19,27,42,72]
	Firm age (younger organizations)	[4,18,19]
	Decentralized structure	[37]
Manager traits and behaviors	Manager characteristics (younger, more educated, and more experienced manager)	[19]
	Management conviction and support	[40,41,72,80,119]
	Managerial cognitive capability	[13]
	Resistance to innovation by managers (-)	[16]
	Leadership change (-)	[40]
Business operations	Brand trustworthiness	[26,27,90]
	Inter-functional collaboration	[23,27]
	Innovation and growth-oriented firm strategy	[19]
	Internal storytelling	[40]
	Corporate venturing and acquisition	[40]
	Resource commitment	[11]
	Data assets	[4]
	Financial distress (-)	[40,122]

Note: A negative sign in parentheses indicates an inhibitor.

, 16 organizational antecedents are categorized into three groups: organizational characteristics (n = 3), manager traits and behaviors (n = 5), and business operations (n = 8).

Organizational Characteristics

Organizational characteristics encompass firm size, firm age, and organizational structure. Firm size is a critical determinant of innovation diffusion [39], with larger firms more likely to adopt next-generation digital technologies due to greater access to technological resources and skilled personnel [4,42]. These firms often leverage new technologies to sustain long-term competitive advantages [27]. In contrast, younger firms exhibit greater agility and are more likely to adopt digital technologies early, as their newer assets are typically more compatible with emerging innovations [4,19].

Manager Traits and Behaviors

Managers significantly influence organizational performance and digital technology adoption. Key factors include managerial characteristics, management conviction and support, and cognitive capabilities, which positively drive digital diffusion. Conversely, resistance to innovation and leadership changes can hinder this process [50]. Younger, well-educated, and experienced managers are more likely to advocate for digital adoption [19], while strong top management support is crucial for successful digital transformation [40,41]. However, managerial aversion to innovation or leadership instability can impede technological progress [40].

Business Operations

Business operations encompass factors such as brand trustworthiness, inter-functional collaboration, innovation-oriented strategies, corporate venturing, resource commitment, and data assets. Strong brand trustworthiness reduces customer uncertainty, encouraging them to accept digital solutions even if these solutions do not fully meet their expectations [26,90]. Brand reputation also motivates firms to integrate digital technologies, such as in food supply chains to mitigate waste [27]. Collaboration among stakeholders further promotes digital adoption [23], though financial distress [40], such as high debt levels, can hinder technological deployment [122].

5.4.3. Environmental Antecedents

Environmental contexts, including country, culture, and industry, significantly influence technology adoption [39]. As summarized in

Table 13. Environmental antecedents [11,15,16,17,18,19,20,21,23,26,27,31,37,40,42,47,65,72,80,90,123].

Types	Antecedents	Articles
Partners	Readiness of partners	[65,80]
	Digital capabilities of suppliers	[31,90]
	Partner congruence and cognitive proximity	[18,27]
	Trust in partner	[16,27]
	Quality and intensity of business linkages	[18]
	Flexibility and fluidity of supply network	[21]
	Enterprise network structure	[123]
	Geographic proximity (-)	[18,37]
	Perceived risk by supply chain partners (-)	[27]
Customers	Client characteristics	[42]
	Customer technical capability	[90]
	Customer satisfaction	[23]
	State-owned customers	[31]
	Perceived risks by customers (-)	[90]
Industry and market	Industry characteristics (industry types, structure)	[11,18,19,26]
	Technology penetration/stock in the industry	[17,72]
	Competitive pressure	[65,72]
	Market externalities	[17]
	Information transparency	[27]
Government	Effectiveness of government policy	[20,23,80]
	Basic information infrastructure	[15]
	Subsidies	[47]
	Universities and other institutions providing human resources	[23,40]

Note: A negative sign in parentheses indicates an inhibitor.

, 23 environmental antecedents are classified into four categories: partners (n = 9), customers (n = 5), industry and market (n = 5), and government (n = 4).

Partners

Partner-related factors play a critical role in inter-firm technology diffusion. Partner readiness, reflecting the preparedness of value chain partners to adopt digital technologies, significantly drives firms’ digital adoption [65]. The digital capabilities of partners, particularly suppliers proficient in advanced applications, further encourage technology adoption [31]. Supply chain congruency, including alignment on targets, compatibility, and cognitive proximity—defined as shared knowledge bases—facilitates the diffusion of complex technologies like AI [18,27]. Trust among partners enhances data and information sharing [16,27], while the quality and intensity of business linkages, along with network flexibility and fluidity, promote DTD [18,21,123]. However, geographic proximity and perceived risks by supply chain partners can act as barriers to this process [18,27,37].

Customers

Customer characteristics and technological capabilities significantly influence firms’ adoption of digital technologies, with studies demonstrating their positive impact on technology integration [42,90]. Customer satisfaction further drives the diffusion of technologies aligned with the smart circular economy [23]. However, perceived risks associated with new technologies among customers can hinder adoption rates [90].

Industry and Market

Industry characteristics significantly shape technology adoption patterns [26]. For instance, firms in the manufacturing and information sectors have been early and intensive adopters of AI technologies [19]. The existing stock of digital technology within an industry positively influences adoption rates, as industry maturity fosters broader uptake [17]. Competitive pressure, driven by peer influence, acts as a catalyst for innovation [65,72]. Furthermore, market externalities and information transparency enhance the visibility and perceived benefits of new technologies, accelerating DTD [17,27].

Government

Disruptive digital technologies, such as AI, pose complex challenges related to security, privacy, and social ethics, underscoring the need for robust legislative and regulatory frameworks to address these issues [80]. Effective policies can amplify benefits, reduce complexities, and create an enabling environment for technology adoption [20]. Government support, including subsidies and investments in basic information infrastructure, alleviates cost barriers to digital innovation [15,47]. Educational institutions, particularly universities, play a pivotal role in knowledge dissemination and human resource development, equipping managers and employees with the expertise required for successful digital implementation [23,40].

5.5. Decisions

Decisions are central to behavioral performance or non-performance, serving as both responses to antecedents and precursors to outcomes [24,54]. They determine the actions to be taken or avoided [73]. This study categorizes firm decision characteristics related to DTD into two types: intra-firm and inter-firm diffusion.

5.5.1. Intra-Firm Digital Technology Diffusion

Intra-firm diffusion focuses on the adoption and integration of digital technologies within an organization. This process typically unfolds in three stages: initiation, adoption, and implementation [124]. Initiation involves scanning for digital opportunities, often driven by change pressures. Adoption entails committing resources to acquire the technology, while implementation includes development, installation, usage, and maintenance to enhance organizational effectiveness [119]. Steiber et al. [40] further break down this process into five steps: desirability, feasibility, first trial, implementation, and sustaining, representing the firm’s journey in exploring, adopting, and embedding technological innovations.

5.5.2. Inter-Firm Digital Technology Diffusion

Inter-firm diffusion resembles a large-scale process of ‘imitation’ in the spread of technological innovations [10]. It involves two key participants: diffusers (initial adopters) and followers (subsequent adopters) [31]. The diffusion stage focuses on spreading the innovation among potential users, with deployment as the central activity. Successful deployment requires resource mobilization and persuading other firms to adopt the innovation, while assimilation occurs when firms fully integrate it into their operations [125]. Feng and Zhu [31] identify three prerequisites for effective diffusion: (1) followers may adopt voluntarily or passively, (2) diffusers typically possess greater influence or strength, and (3) diffusion is driven by external pressures and the need for communication between diffusers and followers.

5.6. Outcomes

Outcomes represent the effects of behavioral actions stemming from antecedents [49,54]. As shown in

Table 14. Outcomes [1,11,13,17,18,22,26,27,28,36,46,83,85,100,111,116,126,127].

Levels	Outcomes	Articles
At intra-firm level	Digital innovation	[1,28,46,100]
	Business model innovation	[111,126]
	Productivity	[36,127]
	Sustainable performance	[83]
	Competitive advantages	[17]
	Uncertainty (-)	[11,46]
At inter-firm level	Epidemic effects	[17,18]
	Supply chain connectivity	[22,27]
	Prevent overproduction and waste	[27]
At industry level	Industrial disruption	[13,46]
	Industrial convergence	[17]
At the macro level	Internationalization	[26,85,116]
	Productivity gains in economies	[17]
	Digital divide (-)	[18]

Note: A negative sign in parentheses indicates an inhibitor.

, the review identifies 13 outcomes across three levels: intra- and inter-firm, industry, and macro.

5.6.1. Outcomes at Firm Level

Intra-Firm Outcomes

Within firms, innovation outcomes are closely linked to technology diffusion [4]. DTD transforms firm boundaries, processes, structures, roles, and interactions [68], creating opportunities for digital innovation [1,46]. The flexibility and adaptability of digital technologies enable their integration into diverse environments, fostering innovation and amplifying organizational innovation potential [1]. The complementary adoption of advanced technologies generates learning-by-use effects, enhancing innovation gains from subsequent adoptions [100]. Digital technologies are driving profound transformations in how value is created, captured, and delivered both within and across organizations, fundamentally reshaping business operations [121,127]. For instance, the diffusion of the Internet of Things is reshaping corporate business models and driving innovation in production processes, customer interactions, and other operational domains [126]. Key factors of digital transformation—such as digital technologies, ecosystem integration, and competitive positioning—fundamentally alter the business model dimensions of firms [111]. Beyond innovation, DTD improves total factor productivity, productive efficiency [36,127], sustainable performance [83], and competitive advantages [17]. However, the adoption process can be protracted and uncertain, often marked by delays, failures, and unclear trajectories [11,46].

Inter-Firm Outcomes

DTD reshapes value chains and inter-firm relationships, particularly within digital ecosystems [68]. Digital and interactive technologies unify participants in social and economic ecosystems under a shared developmental vision, fostering collaboration [14]. This integration often triggers epidemic effects—self-propagating adoption processes driven by increased prevalence, mimetic pressures, and knowledge spillovers [18]. For instance, AI adoption exemplifies these effects through learning and network externalities [17]. Additionally, DTD enhances supply chain connectivity and reduces the risks related to overproduction and waste [22,27]. Digital tracking systems, for example, improve transparency among suppliers and customers, optimizing processes and predictability [27].

5.6.2. Outcomes at Industry Level

The widespread diffusion of digital technologies has disrupted industries, transforming business landscapes through novel customer interactions and the digitalization of products and processes [7,46]. Rapid adoption has driven significant structural changes across sectors in a short time [13]. Advances in digital technology have accelerated shifts in market demand, particularly as innovations from other sectors permeate manufacturing [46]. Historically, industries relied on specialized technologies, but the pervasive adoption of digital technologies has transcended these traditional, symbiotic relationships, fostering industrial convergence [13].

5.6.3. Outcomes at Macro Level

At the macro level, the diffusion of general-purpose technologies (GPTs) by firms has spurred substantial productivity gains in economies [17]. Digitalization enables international expansion, creating cross-border opportunities and facilitating faster, broader internationalization [26]. The affordability and accessibility of digital technologies have particularly promoted the internationalization of Small and Medium Enterprises (SMEs), easing their transition from domestic to global markets [85]. Additionally, digital multinational enterprises enhance the efficiency and competitiveness of host countries across manufacturing and non-manufacturing sectors [116]. However, the diffusion of GPTs, especially in information and communication technology, has been uneven across industries and regions, exacerbating the digital divide compared to earlier GPTs like electricity [18].

5.7. DTD Framework at the Firm Level

Building on the TCM-ADO framework, we propose an integrated firm-level framework for DTD, as illustrated in Figure 7. The framework comprises two main components. The first section integrates antecedents with the TOE framework, offering a comprehensive analysis of enablers and inhibitors of DTD within and across organizations.

The second section outlines decisions and outcomes in two contexts: intra-firm and inter-firm diffusion. Intra-firm DTD is characterized by incremental innovation, focusing on technology adoption and utilization within organizations. This process drives transformations in organizational processes, structures, cultures, management patterns, and business models, enhancing efficiency and fostering digital innovation. The unique attributes of digital technologies, such as increased connectivity and diminishing marginal costs, further support cross-national diffusion, enabling firms to expand internationally and seize new profit opportunities.

When internal resources are insufficient, organizations often collaborate with digitally capable firms, leading to inter-firm DTD. Such collaborations, particularly across industries, reflect the democratization of innovation [128] and enable disruptive innovations. By integrating strengths from diverse sectors—such as manufacturing and services—these partnerships drive significant innovative changes, create emerging business models, and promote industry convergence, heralding a new paradigm of economic development.

6. Avenues for Future Research Based on the TCM-ADO Framework

6.1. Theories-Contexts-Methods

6.1.1. Theories

This literature review reveals that research on DTD is primarily rooted in two established frameworks: the diffusion of innovations theory and the technology–organization–environment framework. While these theories have historically offered valuable insights, the rapid evolution of digital technologies calls for a reevaluation of the existing frameworks and the exploration of new theoretical perspectives across diverse contexts. Our review identified 44 theories, with 34 applied only once, highlighting a fragmented theoretical landscape. Emergent opportunities exist for leveraging underutilized conceptual frameworks such as the stakeholder theory, network theory, and intellectual capital theory, which offer robust analytical tools to address DTD’s complexities. The stakeholder theory elucidates how organizational decisions regarding digital adoption are influenced by diverse stakeholder groups encompassing both internal (e.g., employees and managers) and external (e.g., suppliers and customers) actors who exert divergent pressures on strategic objectives [129,130]. The network theory extends this analysis by emphasizing not only the role of inter-organizational partnerships but also the dynamic interplay between network configurations and organizational agency in value-creation processes [84]. Concurrently, the intellectual capital theory can examine how the strategic adoption of digital technologies serves as a catalyst for enhancing human capital, relational capital, and innovation capital [85]. Furthermore, while most studies adopt a positive analytical perspective, there is a critical need to examine barriers to technology adoption. For instance, integrating innovation resistance theory could provide deeper insights into adoption challenges, enriching the understanding of DTD [20].

6.1.2. Contexts

This literature review reveals that DTD has been extensively researched across various industries, particularly within manufacturing. The versatility and general applicability of digital technologies have spurred their widespread adoption and even facilitated industrial convergence. However, the similarities and differences in DTD between manufacturing and service industries remain underexplored and require further investigation. The review also highlights the growing focus on inter-firm DTD, especially in supply chain contexts, where researchers examine relationships between upstream and downstream firms [31]. Additionally, there is a need to expand the scope of business cooperation to encompass ecosystem-level dynamics and explore DTD within broader business ecosystems [131]. Further research is also warranted to distinguish between intra-firm and inter-firm DTD at the organizational level and to assess their respective impacts on organizational outcomes.

6.1.3. Methods

This review identifies several methodological gaps in the study of DTD. First, both qualitative and quantitative research would benefit from longitudinal data to enable more comprehensive analyses. While quantitative studies often rely on cross-sectional data [20], and only a limited number of case studies adopt longitudinal approaches, the neglect of temporal dynamics overlooks evolving diffusion patterns, including potential increases in diffusion rates over time [16]. Second, the literature predominantly employs quantitative methods, particularly regression analysis, underscoring the need for greater integration of qualitative approaches, such as case studies, to better understand inter-firm diffusion processes. Additionally, although sample surveys are widely used for data collection, most research focuses on developing scales for digital technology adoption at the individual firm level, with limited attention to measuring DTD between firms, representing a critical area for future research.

6.2. Antecedents—Decisions—Outcomes

6.2.1. Antecedents

The review highlights that research on the antecedents of technology diffusion is well developed, encompassing technological, organizational, and environmental factors. However, several gaps remain. First, while positive enablers are extensively studied, barriers to diffusion are underexplored, limiting insights into overcoming adoption challenges. Second, cultural dimensions, particularly organizational and national culture, require deeper investigation. For instance, applying Hofstede’s framework—such as power distance, uncertainty avoidance, and individualism versus collectivism—could elucidate how cultural factors shape DTD. Third, antecedents within ecosystems, involving suppliers, customers, competitors, and complementary providers, are complex and underexamined. Exploring these interactions could reveal new insights into the multifaceted nature of diffusion.

6.2.2. Decisions

The existing research predominantly focuses on the antecedents and consequences of organizational adoption decisions [77], with limited attention to the evolutionary stages of DTD. While early adoption stages, such as the initial decision to adopt technology, are well studied [18], the subsequent phases and their interconnections remain underexplored. Additionally, while technology acceptance is widely examined, resistance to diffusion—its causes, mechanisms, and mitigation strategies—requires further investigation. Finally, differences and similarities between inter- and intra-organizational diffusion patterns are poorly understood, representing a critical area for future research to enhance our understanding of digital technology spread.

6.2.3. Outcomes

Although the impacts of digital technologies on businesses are widely studied, their long-term effects and potential negative consequences, such as the digital divide and security risks, remain underexplored. The current research primarily focuses on adopters, examining how firms integrate and benefit from digital innovations while neglecting the role of technology diffusers—firms that actively promote and disseminate digital technologies. Future studies should expand their scope to include both adopters and diffusers, providing a more comprehensive understanding of the broader implications of DTD.

As summarized in

Table 15. Future research directions.

Dimension	Research Question for Future Research
Theories	What novel theoretical frameworks can be proposed to explain the diffusion of emerging disruptive digital technologies, such as artificial intelligence? How do theories like stakeholder theory and social network theory explain DTD amidst increased inter-organizational cooperation? What theoretical frameworks can be developed or adapted to explain resistance to DTD?
Contexts	What are the key similarities and differences in DTD between the manufacturing and service sectors? How does DTD manifest within corporate partnerships, particularly within business ecosystems? How do intra-firm and inter-firm DTD processes differ, and what impact do they have on organizational innovation?
Methods	How does DTD evolve over time within and between firms? What new qualitative frameworks or models can be developed to better understand the processes of DTD between firms? What measurement scales can be developed to assess the effectiveness of inter-firm DTD?
Antecedents	What are the critical barriers to DTD? How do organizational and national culture influence the rate and effectiveness of DTD? How do business ecosystems or networks impact DTD between firms?
Decisions	What are the key dimensions and stages of the DTD model, and how do these stages differ from and relate to each other? How does resistance to DTD occur, and what measures can be taken to prevent it? What are the differences between inter-firm and intra-firm diffusion?
Outcomes	What are the long-term effects of DTD on various aspects of business operations? What are the potential negative consequences of DTD, particularly in terms of the digital divide? What are the impacts of DTD on the entities that actively promote its dissemination?

, we frame these dimensions of the TCM-ADO framework as research questions to guide future investigations into DTD.

7. Conclusions

7.1. Theoretical Implications

This review represents the first comprehensive consolidation of DTD at the firm level, addressing recent calls for further research in this area [11]. This study employs a novel hybrid review that combines bibliometric analysis with the TCM-ADO framework-based review, making two key theoretical contributions. First, it provides a systematic mapping of the field through bibliometric insights, highlighting trends in annual research output, key contributors (authors and countries), influential publications (journals and articles), and co-citation and co-occurrence analyses. Second, it advances the theoretical understanding of DTD by applying the TCM-ADO framework, which elucidates various theories, contexts (across industries and countries), and methodologies. Additionally, the study categorizes antecedents on the TOE framework and provides a multi-level analysis of decisions and outcomes, encompassing intra-firm, inter-firm, cross-industry, and macro-level perspectives. Furthermore, the study identifies promising research directions based on the TCM-ADO framework, offering a foundation for future scholarly exploration in this field.

7.2. Managerial Implications

This research delves into DTD at the firm level, examining both internal and external transfer. By uncovering the key antecedents, decisions, and outcomes associated with DTD, this study offers targeted strategic recommendations to facilitate seamless adoption and effective deployment. At the antecedent level, our findings direct diffusers towards factors that enhance the uptake of digital technologies, providing a solid starting point and a clear pathway for subsequent adoption efforts. In terms of decision making, this study equips managers with a deeper understanding of the stages of technology adoption, enabling them to devise more informed and effective implementation strategies. Regarding outcomes, the identified positive impacts serve to encourage managers to embrace digital innovations, while highlighted negatives alert them to potential risks, guiding them to mitigate adverse effects during the adoption phase. For instance, Tesla Motors revolutionized the automotive industry by leveraging digital technology to create software-centric vehicles and a vertically integrated energy ecosystem. Innovations such as over-the-air updates and autonomous driving showcase how digital transformation can redefine business models [132]. Companies can also leverage big data and machine learning to derive customer insights, personalize experiences, and develop scalable platforms that drive network effects. This enables firms to redefine their value propositions, enhance value creation efficiency, and expand their revenue models.

7.3. Research Limitations

This study presents two primary limitations. First, regarding data sources, we restricted our selection to journals ranked as A* or A on the ABDC list to ensure high-quality research. However, this restriction might introduce publication bias, potentially narrowing the generalizability of our findings and causing us to overlook important studies despite our rigorous search criteria. Second, concerning data analysis, while the TCM-ADO framework provides a robust foundation for categorizing information, it might inadvertently exclude critical data that does not conform to its predefined elements, leading to a potentially less comprehensive analysis of the available data.