Digital Transformation Characteristics of the Semiconductor Industry Ecosystem

Abstract

This study investigates the characteristics of competitive strategies, digital competencies, and management performance as digital drivers in the semiconductor industry. We constructed a conceptual research model and proposed five research hypotheses based on previous studies and analyzed the fit of the research model using IBM SPSS AMOS software. In total, 131 companies specializing in semiconductor materials, components, equipment, and fabless design comprised the sample for analysis. To improve the fit of the research model, several unnecessary paths and measurement items were removed, the final research model was selected, and the five hypotheses were tested. The results indicate that digital drivers positively affect digital competencies, which, in turn, positively affect efficiency-centered and novelty-centered business models. Regarding the relationship between these business models and management performance, only the novelty-centered business model positively affects non-financial performance. These findings could help formulate digital transformation strategies related to products/services, processes, and business models at the firm level, as well as establish policies such as regulation and support programs at the government level.

1. Introduction

The term “digital transformation” describes the expansion of existing concepts of digitalizing products and business activities to create a completely new business model using the latest information and computer technology [1,2]. Such a digital transformation has been mainstreamed throughout entire industries by improving convenience and corporate efficiency and creating new businesses using the key technological variables of the Fourth Industrial Revolution, such as artificial intelligence (AI), the IoT, big data, and cloud computing.

Most research on digital transformation as a corporate competitive strategy has focused on case analysis, the development of evaluation indicators, and implementation methods; hence, research on the suitability of management strategies is insufficient [3]. According to Barney’s resource-based view [4], digital transformation can be considered an important driver, competency, and strategy for managing the internal resources and capabilities of a company. Excellent corporate performance is attributed to a company’s unique resources and technologies that are difficult for competitors to imitate. From a dynamic competency perspective [5], digital transformation is a corporate competency for creating new products and processes; it can be considered a source of sustainable competitive advantage in a changing competitive environment for digitalizing products, business processes, and business models.

The semiconductor industry is crucial to both enabling digital transformation and growing the industry through the continuous introduction of novel digitalized products, services, and business processes. Integrated semiconductor manufacturers and global fabless design companies are actively strengthening their digitization and digitalization capabilities to differentiate their products. Increasingly important are the various sub-industries related to materials, components, and equipment in the semiconductor industry which form a growing ecosystem.

The COVID-19 pandemic triggered global supply chain issues that have hindered global economic growth. In particular, semiconductor device manufacturers are experiencing difficulties in timely production and investment due to global supply chain issues related to materials, components, and equipment. Cooperation among companies competing within the ecosystem is increasingly critical to resolving this issue. The digitization of products and digitalization of operations, which are promoted by large companies, are also required in the semiconductor ecosystem, as they are vital to companies’ core competitiveness, differentiation, and efficiency. Therefore, this study conducted an empirical analysis based on the following research question: “What are the characteristics of the digital transformation of ecosystem companies with differences in business scale and resource capability compared with large semiconductor device-centered companies?” Since the semiconductor industry comprises various companies specializing in materials, components, and equipment whose competitiveness significantly affects the entire industry, these companies were selected as the target firms.

The remainder of this paper is structured as follows. Section 2 provides a literature review examining the relationships among competitive strategies, competencies, management performance, and digital transformation, and presents hypotheses. Section 3 describes the research method used in this study. Here, we target and collect information on the ecosystem of Korean semiconductor device companies from the CEOs and CTOs of members of the Korea Semiconductor Industry Association, device companies’ partner companies, and executives who represent companies. Section 4 shows the results of the hypothesis testing using structural equation modeling. Section 5 discusses the corresponding findings of the study, including the study’s theoretical, practical, and political implications. Finally, Section 6 concludes the paper.

2. Literature Review and Hypotheses

Companies develop competitive strategies to gain advantages in an evolving competitive environment. Porter [6] divided competitive advantage into cost and differentiation advantages. Cost advantage is the power to ensure a more economical supply than that provided by competitors through various activities, including production, sales, and after-sales service. Differentiation advantage is generated by consumers’ perception of “edge” through unique design, excellent quality, and brand power. Producing a competitive strategy involves establishing a plan that links the opportunities and threats identified through external environment analysis to internal resources and capabilities. The suitability of a competitive strategy implies that a company’s external environment and internal resources, capabilities, and strategies are optimally linked [7]. Noh [3] investigated 20 studies on the suitability of digital transformation with competitive strategies. Studies of competitive strategies’ impacts on corporate performance and effectiveness were the most dominant, followed by studies suggesting alternatives to competitive strategies related to particular industries and businesses and the effects of competitive strategies. Noh’s finding of only one study on the suitability of competitive strategies indicates that research on this topic is insufficient; further studies are thus needed to examine the suitability of competitive strategies and the internal and external environments. Furthermore, case analysis was found to be the most frequent method of analysis employed in 12 studies, followed by diagnosis of digital transformation internalization levels, evaluation in index development-related research, and digital transformation promotion plan-related research. Only one study was conducted on digital transformation strategy. Nevertheless, companies must identify opportunities and threats through internal and external environmental analyses and pursue digital transformation to exploit opportunities and overcome threats.

2.1. Changes in Digital Transformation

Regarding early product digitization, Bloomberg [1] mentioned that analog information is converted to “0” and “1” to be stored, processed, and moved to computers. Savić [2] defined digitization as the process of converting analog form to digital form. Typical examples of product digitization are digital TVs and smartphones. Moreover, Savić [2] defined digitalization as the automation of business processes in three steps: (i) automate a single operation and process, (ii) automate or link related processes, and (iii) enable various systems to support or link to the management process. Savić [2] thus considered digital transformation to be a means of creating completely new business models using the latest information and computer technology. Furthermore, he mentioned that mobile applications, AI, cloud computing, big data, and other digital services could augment current businesses without changing their essence. According to the International Data Corporation [8], an information research institute, digital transformation is the application of recent technology that rapidly changes processes, customer experience, and value. Numerous studies have also assessed the performance of platform businesses such as Google, Amazon, Facebook, and Apple via digital transformation [9,10,11].

2.1.1. Case of Consumer Goods Industry

In this study, we introduce cases of product digitization, management process digitalization, and digital transformation by citing case studies from the Samsung Economic Research Institute [12,13]. Johnson & Johnson Inc. [12] transformed from a giant consumer goods company providing children’s products, band-aids, and contact lenses into a digital company, and its market capitalization doubled within a decade. There are three main reasons for this success, the first being product digitization based (primarily) on digital technology. In 2015, Ethicon, a Johnson & Johnson subsidiary company, established Verb Surgical with Google Verily; in 2018, it acquired Orthotaxy, a company specializing in knee surgery robots; and in 2019, it acquired Auris Health, a surgical robot company, and developed AI technology, while promoting active open innovation internally and externally. Today, Ethicon is developing a Remote Assessment in Rheumatoid Arthritis app that assists users in treating diseases and managing their health through links with various wearable devices.

The second reason is the digitalization of product development processes. New medicine development involves substantial costs and time, but Johnson & Johnson has been actively using digital technology to dramatically reduce the cost and time needed to develop new medicines. In 2016, Janssen Pharmaceuticals, another subsidiary of Johnson & Johnson, signed an MOU with Benevolent, a UK-based new medicine development company, agreeing to cooperate on AI to reduce costs and save time. Furthermore, they announced the world’s first digital technology-based “virtual clinical trial” plan in 2019, wherein all processes—from the recruitment of clinical patients to data screening and collection—were conducted via wearable devices and smartphone apps, reducing reluctance to participate in clinical trials; consequently, they secured a greater number of clinical participants.

Lastly, digital technology is used to track and monitor manufacturing and distribution processes. This example is connected to the aforementioned digital transformation. Tracking manufacturing and distribution processes via sensors and IoT technology has led to the development of smart supply chains that can connect all places of use such as manufacturing plants, distributors, and hospitals. Hence, a new business model was promoted that facilitates high-quality product management by collecting real-time data via digital technology and enabling decision making related to manufacturing (e.g., inventory management and demand forecasting).

Domino’s Pizza [13] is an example of a traditional consumer goods company that transformed into an IT-based company by adopting digital technology to digitalize its business processes. It has invested heavily in the digitalization of all its processes, from ordering to delivery. In fact, approximately half of the company’s employees are software engineers and data analysts. In 2003, it started the industry’s first online ordering system and introduced a pizza tracker system that allows consumers to check their order status in real time. In 2015, the company established a digital platform called “Domino’s Anywhere” to allow customers to place orders in multiple and diverse ways.

Furthermore, Domino’s has applied diverse and new digital technologies in the delivery sector. For example, its unmanned delivery vehicle (i.e., Domi-No Driver) was introduced in 2015. Autonomous driving technologies were applied in the form of a two-wheel-drive motorcycle. In 2016, in Australia, the company commercialized the world’s first Domino’s robotics unit, which had been converted from a military robot, and found delivery locations using Google Maps and GPS technology. Customers can retrieve their products by entering a security code sent to their mobile phones. As this example demonstrates, consumer goods companies have been promoting changes in corporate strategies, capabilities, and business models through digitalization, and business-to-business (B2B) companies can thus follow suit.

2.1.2. Case of Semiconductor Industry Ecosystem

In memory semiconductors, closed innovation occurs that differentiates each company’s design and process capabilities based on standardized specifications, whereas the manufacturer provides customers with a marketplace by securing process capabilities and IP in advance. The semiconductor industry comprises various design and manufacturing-related sectors [14]; this ecosystem helps reduce costs as well as achieve technology learning, innovation, and growth [15].

From a design perspective, companies’ ecosystems (i.e., their main customers, systems companies, and IC design (fabless) companies) determine the necessary functions and performance. The companies are in the design stage of implementation regarding semiconductors and the provision of tools and IP for IC design. They are also stakeholders in the value chain by performing design services for the manufacture of semiconductors. From a manufacturing perspective, companies’ ecosystems include stakeholders in the facilities and materials processes related to manufacturing to create the design results. While memory and system semiconductors are similar, the latter require various processes such as a dielectric constant and high voltage to match the characteristics of IC. The increasing number of products using advanced processes rather than memory has recently generated fierce competition among manufacturers.

The digital features of the B2B industry, such as the ecosystem of semiconductor companies, are similar to those of the business-to-consumer industry in terms of product digitization, the digitalization of management activities, and digital transformation. However, such digital transformation is being digitalized to secure the innovation and capabilities required by customers.

The first innovation is the digitization of products through the internalization of AI. In 2020, Applied Materials, Advanced Semiconductor Materials Lithograph, and Lam Research, respectively, ranked first, second, and third place in global sales of semiconductor equipment. Applied Materials [16] is equipped with a complex facility capable of complex processes and AI and promotes not only preventive maintenance but also stable performance maximization. Advanced Semiconductor Materials Lithography [17], a representative lithography company that supplies equipment such as extreme ultraviolet (EUV) photoresists to semiconductor factories, has a predictive maintenance solution that informs the status of equipment abnormalities in real time by analyzing vibrations or currents to prevent equipment failure in factories. Lam Research [18] has announced its intention to promote advancements and a shorter learning cycle by developing equipment design and processes in the virtual world through digital twins.

In 2019, the items that the Japanese government regulated for export in the field of semiconductor materials were high-purity hydrogen fluoride (etching gas), EUV photoresists (photoresists for semiconductors and displays), and polyimide fluoride (fluorinated polyimide film). This situation heightened the sense of crisis across the ecosystem of the Korean semiconductor industry. However, the government and companies used this crisis as an opportunity to reduce their dependence on Japan for materials, components, and equipment, and the public and private sectors collaborated to localize semiconductor materials and secure a supply chain. Consequently, South Korea’s 50% dependence on Japanese hydrogen fluoride was lowered to 10%, and the adoption of alternative materials virtually eliminated polyimide fluoride imports from Japan. In the case of EUV photoresists, dependence on Japan has been reduced by less than half as Belgian imports have increased. However, to accelerate the rise in competitiveness in materials and components by securing high-quality core materials faster than competitors, the government has also promoted the development of customized materials for companies using data on materials such as raw materials and properties and AI. Furthermore, it has encouraged the digitization of the materials industry to secure new materials.

Second, we examined the digitalization characteristics of business processes. According to Kwon and Song [19], the promotion of platform-based digital transformation for manufacturing innovation is commonly found in an ecosystem consisting of large firms and several partners in Korea. By organically sharing information between partners or between partners and companies through collaboration platforms, they should achieve “mutual growth, strengthen partnerships and global manufacturing competitiveness, and improve global manufacturing quality competitiveness”. This concept can be identically applied to global materials, components, and equipment companies. To this end, it is important to improve product quality by using a system, strengthen the capacity for timely supply, and induce changes for survival when components are discontinued. Smart factories should not aim for factory automation; instead, they must focus on the digital transformation of manufacturing industries based on a collaboration platform that enables reasonable decision making through the production of competitive products and implementation of appropriate operating processes.

Among our cases, Lam Research [18], a global equipment company, runs a manufacturing execution system that can detect and act on manufacturing status as well as solve problems in real time by integrating the operation of the manufacturing system with the systems of partner companies. Moreover, it has introduced AR/VR to train staff and customer support services for new facilities and technologies. Korean companies also promote digitalization through supply chain management and manufacturing execution systems. For example, Tokyo Electron [20] has increased sales of remote maintenance services based on digital technologies by securing a stable source of revenue by performing more equipment maintenance projects and maintaining certain factory operation rates. This has increased profits because of the need for fewer customer support personnel and overseas business trips.

2.2. Factors in Digital Transformation

As digitalization has evolved into product digitization, business process digitalization, and digital transformation, its importance has been highlighted as a core competency and differentiation factor for companies. Digitalization as a key competitive advantage has been expanding for both business-to-consumer and B2B companies. This study derived a research model from empirical studies of the resource-based view and dynamic capabilities perspective frameworks, as well as recent digitalization and business models. Regarding the resource-based view, Barney [4] mentioned that having unique resources and technologies that are difficult for competitors to imitate leads to excellent corporate performance; hence, he created a research model based on environmental analysis/strategy/resource capabilities and competencies/performance. Teece et al. [5] defined dynamic capabilities as helping companies create new products and processes and respond to a changing market environment; they also mentioned dynamic capabilities as a source of sustainable competitive advantage in a changing competitive environment. In this study, we examine the characteristics of the drivers of corporate digitalization, digital competencies, digital orientation, business models, and management performance in line with the new wave of the Fourth Industrial Revolution and environmental changes such as those brought about by COVID-19.

2.2.1. Digital Drivers and Digital Competencies

A driver is defined as “a factor that triggers a specific phenomenon, such as causing or developing”. Hrustek et al. [21] defined digitalization as influential factors and ideas for organizational innovations that occur because of external environmental changes and internal innovation, ideas, and will. Hrustek’s team used strategic environmental maps as an analytical framework to study the impacts of digital transformation on business model creation. This method is useful for identifying and analyzing business models; prompting innovation; and analyzing changes in technology, regulation, customers, competitors, and the economic environment.

According to Digitrans strategic landscape maps [22], technological change is defined as a new condition wherein novel digital technologies in an industry are already widely used in other industries, or as changes in laws and regulations that define new conditions applicable to all organizations. Changes in customers indicate shifts in expectations for products and services; changes in competitors are shifts in other companies within the same business; and changes in the economic environment are changes in customers’ purchasing habits and practices. In terms of innovation-related ideas within a company, it is crucial to invest in human capital, equipment, and infrastructure to stimulate creativity and ideas in the value creation process [23]. Thus, the idea of digitalization within a company affects its capacity for investment to realize digitalization.

Technical competencies, important resources for innovation processes [24,25], can be defined as the ability to formulate and develop new products and related processes [26]. For digital products, digital competencies can be defined as the skills, talent, and expertise of a company that manages digital technologies for new product development. Carcary et al. [27] indicated that successful digital transformation requires the development of several functions in diverse domains and that such functions may differ depending on the requirements of specific sectors and organizations. Westerman et al. [28] clarified that digital competencies are a basic building block that enables companies to improve consumer experiences, operational processes, and business models, and mentioned that technological gaps could disrupt digital innovation.

According to Teece et al. [5], external linkages (i.e., relationships with other businesses within the same industry) help companies acquire complementary information and knowledge as well as formulate and implement innovative strategies and use current resources in novel ways through new insights and ideas. Furthermore, forging relationships with heterogeneous industries allows companies to understand the competitive environment at the macroeconomic level and access novel resources and ideas to develop novel strategies. As previously discussed, digital drivers affect digital competencies based on the logic that technical competencies are influenced by external information and knowledge as well as internal ideas. Considering this point, we posit Hypothesis 1 as follows:

Hypothesis 1. 

In the semiconductor industry ecosystem, digital drivers positively affect digital competencies.

2.2.2. Digital Competencies and Business Models

Yu et al. [29] found that researchers regard business models as interconnected and interdependent systems that enable a company to determine how to conduct business with customers, partners, and vendors [30]. Another study found that companies use business models to provide customers with something of value and customers offer financial rewards for that value, thereby converting valuable assets into profits [31]. However, Martins et al. [32] indicated that most researchers consider a business model to be an activity system, with behaviors developing differently according to the company’s internal characteristics.

Hrustek et al. [21] posited another view. They categorized nine elements of a business model and introduced a three-dimensional business model option matrix. The first dimension is optional and comprises the elements related to a company’s internal business structure (e.g., key activities, key resources, key partners, and cost structure). The second dimension is a commercial model related to customer elements (e.g., customer segments, customer relationships, channels, and revenue streams). The last dimension is a value proposition in terms of the value of products and services that use them. Researchers typically agree that business models are a key factor behind competitive advantage in the knowledge-based economy [29].

Zott and Amit [33] suggested two business models based on the activities of a company to discover and create value. While the efficiency-centered business model focuses on achieving transaction efficiency by reducing transaction costs, the novelty-centered business model emphasizes a new method of performing economic exchange [33,34]. First, the essence of the efficiency-centered business model aims for a continuous decline in transaction costs [33], for which the possession of sufficient alternative resources for any company is an important prerequisite [35]. Furthermore, digital competencies affect the efficiency-centered business model, as discussed earlier, since competencies can improve the bargaining power of a company with resource holders and reduce information asymmetry. Therefore, we posit Hypothesis 2 as follows:

Hypothesis 2. 

In the semiconductor industry ecosystem, digital competencies positively affect the efficiency-centered business model.

Second, compared with the efficiency-centered business model, the novelty-centered business model has rarely been applied because of high market uncertainty [29]. However, when used, it can create new markets and new trading opportunities in existing markets. Furthermore, the essence of the novelty-centered business model is the conceptualization and adoption of new methods of conducting economic transactions by connecting and establishing relationships with existing transaction participants in new ways. However, when a current business model changes, the level of novelty rises, leading to higher conversion costs, and such a conversion process may bring about significant confusion. Owing to the nature of B2B relations, coordination with customers is important; nevertheless, new products and services provided by companies’ ecosystems may affect their continuous cooperative relationships with existing customers. Therefore, we propose Hypothesis 3 as follows:

Hypothesis 3. 

In the semiconductor industry ecosystem, digital competencies positively affect the novelty-centered business model.

2.2.3. Business Models and Management Performance

Westerman et al. [28] revealed that companies with above-average digital innovation values are more profitable and earn higher profits. Weill and Woerner [36] also indicated that companies that accept digital technologies and operate within digital ecosystems enjoy higher profits. Conversely, Chae et al. [37] found no relationship between IT functions and organizational performance. They claimed that future studies should identify and integrate other variables that may affect the relationship between IT functions and business performance. However, an efficiency-centered business model may affect financial performance and a novelty-centered business model may have a partial impact. Additionally, novelty-centered business models may affect non-financial performance in areas such as customer satisfaction and market share. As efficiency-centered and novelty-centered business models are both considered to impact financial and non-financial management performance, we posited Hypothesis 4 as follows:

Hypothesis 4-1. 

In the semiconductor industry ecosystem, an efficiency-centered business model positively affects financial performance.

Hypothesis 4-2. 

In the semiconductor industry ecosystem, an efficiency-centered business model positively affects non-financial performance.

Hypothesis 4-3. 

In the semiconductor industry ecosystem, a novelty-centered business model positively affects financial performance.

Hypothesis 4-4. 

In the semiconductor industry ecosystem, a novelty-centered business model positively affects non-financial performance.

The research model for evaluating those hypotheses is shown in Figure 1.

This research model has been derived based on the following topics of recent empirical studies: (a) the external environment (e.g., industry and government), resource combination, and business models [29]; (b) continuous recombination capability and stakeholder management, strategic orientation, and destructive business model innovation [38]; (c) digital orientation and digital competencies, digital innovation, and financial/non-financial performance [39]; and (d) digital transformation drivers and business model creation [21]. Based on the model above, we investigate the following effects: (a) the impact of internal and external digital drivers on digital competencies (Hypothesis 1), (b) the impact of digital competencies on the efficiency-centered business model (Hypothesis 2), (c) the impact of digital competencies on the novelty-centered business model (Hypothesis 3), and (d) the impact of the efficiency- and novelty-centered business models on financial and non-financial performance (Hypotheses 4-1, 4-2, 4-3, and 4-4).

Recent studies on digital transformation can be largely divided into three categories. The first is conceptual studies on digital transformation. After Hrustek et al. [21]’s conceptual study on the impact of digital transformation drivers on business model creation, empirical studies on the impact of these digital drivers have been difficult to find. So, in this study, an empirical study on digital drivers was conducted. Bican and Brem [40] proposed a conceptual framework for digital terms and inter-relations. A case study was conducted with 12 experts from German high-tech companies. Rof et al. [41] conducted a case study on how business models are innovated due to the digital transformation of higher education institutions. The second is empirical studies of digital transformation. Khin et al. [39] conducted an empirical study on relations among digital capability, digital innovation, and management performance for Malaysian SMEs IT companies. Kim [42] conducted an empirical study on causality among a company’s ability (management and technological), competency (technology innovation and technology marketing), and performance (financial and technical) in the era of digital transformation. Lastly, there are case studies on the economic, social, and environmental impacts of digital technology. Peron et al. [43] defined procedures and emerging technologies for facility layout planning using digital and introduced digital tools. The impact of the introduction of these new technologies on reducing costs related to sustainability, improving social aspects, and protecting the environment was discussed.

This research differs from previous studies in two respects. The first is to conduct an empirical study on the characteristics of digital transformation’s competitive strategies, competencies, and management performance in connection with external environmental characteristics (digital drivers). Previous studies have mainly focused on the internal characteristics of companies in digital transformation, but this research is trying to analyze the path by establishing an integrated research model for this. Second, it is meaningful to study the characteristics of ecosystem companies in the semiconductor industry, which has become more important following the Fourth Industrial Revolution and COVID-19. In the growth of the semiconductor industry, the competitiveness of ecosystem companies will be very important. This study focused on this ecosystem. This study is significant as it can be confirmed as an ecosystem study of other industries.

3. Research Methods

3.1. Data Collection

The semiconductor industry, which consists of various ecosystems, was selected as the target industry because it was deemed capable of representing the characteristics of a B2B company in digital transformation research. Hence, we targeted member companies of the Korea Semiconductor Industry Association and partner companies of a Korean integrated device manufacturer. We sent them questionnaires and received their responses via e-mail. The survey was conducted for approximately four months, from May to August 2021. The questionnaire was administered to approximately 250 people and 140 responses were received. Among them, 131 samples were used for the final analysis after excluding 6 samples because of incomplete responses and 3 samples because of inappropriate responses (i.e., using one answer repeatedly for several questionnaire items).

Since this study investigated the features of companies, it mostly targeted CEOs but also included the responses of other executives such as CTOs and heads of sales divisions. Increasing the number of respondents further was limited because we targeted experts. As shown in

Table 1. Distribution of the respondent companies and respondents.

Respondent Companies		Frequency	%	Respondents		Frequency	%
Age	1–10 years	30	23	Tenure	1–10 years	0	0
	11–20 years	45	34		11–20 years	22	17
	21–30 years	36	27		21–30 years	77	59
	More than 31 years	20	15		More than 31 years	32	24
Number of employees	1–100 employees	75	57	Position	CEO	71	54
	101–500 employees	38	29		CTO	19	15
	>501 employees	18	14		Others	41	31
Total		131	100	Total		131	100

, 77% of the respondent companies had been operating for at least 11 years. Such companies were assumed to be appropriate for analyzing digital transformation features since they have long been performing semiconductor ecosystem projects. Altogether, 57% of the respondent companies had fewer than 100 employees. More than 80% of the respondents were experts with more than 21 years of work experience and the proportion of CEOs was 54%. These percentages suggest that our respondents could accurately portray the characteristics of the sample companies.

3.2. Definition of the Variables

3.2.1. Digital Drivers

As shown in

Table 2. Measurement items of digital drivers.

Category	Measurement Item	Measurement Method	Reference(s)
Digital drivers	ED1. Customers’ demand for product/service digitization ED2. Recognition of new external digital technology ED3. Competitors’ recognition of digital technology adoption ED4. Spread of the digitalization of sales patterns and methods ID1. Recognition of the need to adopt new digital technology ID2. Willingness to challenge differentiation through new digital technology ID3. Efficient organizational operations through digitalization	5-point scale	Hrustek et al. [21], Digitrans [22]

, in this study, we derived the measurement items from Hrustek et al. [21] and Digitrans [22], as well as ideas around the internal value creation process in a company. We derived the following measurement items for external digital drivers: customers’ demand for digitization, recognition of new external digital technology, competitors’ recognition of digital technology adoption, and spread of the digitalization of sales patterns and methods. We derived the following measurement items for internal digital drivers: recognition of the need to adopt new internal digital technology, willingness to challenge differentiation through new digital technology, and efficient organizational operations through digitalization.

3.2.2. Digital Competencies

As shown in

Table 3. Measurement items of digital competencies.

Category	Measurement Item	Measurement Method	Reference(s)
Digital competencies	DC1. Acquisition of important digital skills DC2. Identification of new opportunities using digital technology DC3. Ability to respond to digital transformation DC4. Internalization of cutting-edge digital technology DC5. Development of products/services/processes using digital technology	5-point scale	Khin and Ho [39], Zhou and Wu [44]

, this study used the measurement items transformed from a digital perspective based on the measurement items of technical competencies of Zhou and Wu [44] that were adopted by Khin and Ho [39]. The measurement items encompass the acquisition of digital technology, identification of new opportunities, ability to respond to digital transformation, internalization of cutting-edge digital technology, and development of innovative products/services/processes using digital technology.

3.2.3. Business Models

As shown in

Table 4. Measurement items of business models.

Category	Measurement Item	Measurement Method	Reference(s)
Efficiency-centered Business Model	ECBM1. Reduction in inventory costs ECBM2. Reduction in transaction costs ECBM3. Decision making based on information ECBM4. A transparent transaction with information flow/utilization ECBM5. Reduction in knowledge asymmetry ECBM6. Fast transaction speed ECBM7. Higher transaction efficiency	5-point scale	Yu et al. [29], Zott and Amit [33], Yasuda [35]
Novelty-centered Business Model	NCBM1. Provision of new combinations of products, services, and information NCBM2. New incentives for trading participants NCBM3. Provision of unprecedented participants and diversity NCBM4. Connection of participants in a creative way NCBM5. Introduction of continuous innovation by the business model owner NCBM6. Competition with new business models within a company NCBM7. Originality of the business model	5-point scale	Yu et al. [29], Zott and Amit [33], Teece [31]

, for the efficiency-centered business model, we used the measurement items of Zott and Amit [33] and Yasuda [35], which were also adopted by Yu et al. [29]. The measurement items included reduction in inventory costs, reduction in transaction costs, decision making based on information, transparent transactions, reduction in knowledge asymmetry, fast transaction speed, and transaction efficiency. For the novelty-centered business model, we used the measurement items of Zott and Amit [33] and Teece [31], which were also adopted by Yu et al. [29]. The measurement items comprised the extent to which new combinations of products/services and information are provided, incentives for transaction participants, level of diversity, level of connection with original participants, level of innovation continuity, competition with new business models within a company, and originality of the business model.

3.2.4. Management Performance

As shown in

Table 5. Measurement items of management performance.

Category	Measurement Item	Measurement Method	Reference(s)
Financial performance	FP1. Satisfaction of recent increase in sales FP2. Satisfaction of recent operating profit increase FP3. Satisfaction of recent cash flow	5-point scale	Khin and Ho [39]
Non-financial performance	NFP1. Recent increase in customer satisfaction NFP2. Recent increase in market share NFP3. Recent reduction in employee turnover	5-point scale	Khin and Ho [39]

, this study used the following measurement items adopted by Khin and Ho [39]. Satisfaction with sales, profits, and cash flow was measured for financial performance, while customer satisfaction, market share, and employee turnover were measured for non-financial performance.

3.3. Research Method

The data were statistically analyzed using IBM SPSS AMOS software. First, the goodness-of-fit of the research model was verified. The indicators used to measure the fit of the research model included: the goodness-of-fit index (GFI), which is a criterion for judging the goodness-of-fit based on the similarity between the covariance matrix of the data and covariance matrix of the estimated model; the normed fit index (NFI), incremental fit index (IFI), and relative fit index (RFI), which judges the goodness-of-fit based on the discrepancy between the model and data; and the Akaike information criterion (AIC), which evaluates the size of the discrepancy between all the data of the statistical model and the model.

Second, to improve the fit of the research model, the Wald test [45] was used and unnecessary correlations or paths were removed to derive the final modified model, which showed an excellent fit. Model 1 is the baseline research model and measurement items. Model 2 includes several modified paths. Model 3 is modified by gradually removing low squared multiple correlations of the measurement items from Model 2 [45].

Third, the reliability and validity of the measurement items were measured. The average variance extracted (AVE) and concept reliability (CR) were measured to verify validity. The model’s explanatory power was evaluated based on a threshold value of the AVE 0.5 or higher [46]. CR was evaluated based on a threshold value of 0.7 or higher [47].

Fourth, after evaluating the reliability, validity, and model fit, this study used IBM SPSS AMOS software to perform a path analysis of the correlations between each factor to determine their significance.

4. Results

4.1. Model Fit and Reliability and Validity of the Measurement Tool

As shown in

Table 6. Model fit.

Category	χ2 Test			GFI	AIC	RMSEA	CFI	IFI	RFI
	χ2	Degrees of Freedom	Significance Level
Model 1	929.050	457	0.000	0.701	1071.050	0.089	0.840	0.842	0.708
Model 2	930.399	459	0.000	0.702	1068.399	0.089	0.841	0.843	0.709
Model 3	359.073	184	0.000	0.806	453.073	0.086	0.912	0.913	0.814

, the GFI value of Model 1 was 0.701. We attempted to improve the model fit by removing the paths between the efficiency-centered business model and non-financial performance and between the novelty-centered business model and financial performance; however, the resulting improvements in the GFI values and other index values were insignificant. Thus, Model 3 was selected as the final model, as it had a GFI of at least 0.8. Model 3 also had the lowest AIC among the three models, a CFI of at least 0.9, an IFI of at least 0.9, and an RFI of at least 0.8.

As shown in

Table 7. Reliability and validity of the measurement tool.

Category	Non-Standardized Coefficient	Standard Error	Critical Ratio	Standardized Coefficient	AVE	CR
Digital drivers → ED1	1.070	0.123	8.625	0.758	0.735507	0.917482
Digital drivers → ED2	1.139	0.125	9.141	0.803
Digital drivers → ID1	1.070	0.120	8.928	0.784
Digital drivers → ID2	1			0.783
Digital competencies → DC1	1			0.867	0.776651	0.945606
Digital competencies → DC2	1.026	0.075	13.677	0.878
Digital competencies → DC3	1.047	0.078	13.397	0.868
Digital competencies → DC4	1.197	0.093	12.809	0.848
Digital competencies → DC5	1.131	0.085	13.343	0.866
Efficiency-centered business model → ECBM6	1			0.933	0.872065	0.931586
Efficiency-centered business model → ECBM7	0.836	0.110	7.606	0.871
Novelty-centered business model → NCBM2	0.879	0.095	9.277	0.772	0.705953	0.923029
Novelty-centered business Model → NCBM3	0.970	0.097	10.037	0.824
Novelty-centered Business model → NCBM4	0.948	0.093	10.160	0.833
Novelty-centered business model → NCBM5	0.963	0.096	10.034	0.824
Novelty-centered business model → NCBM6	1			0.778
Financial performance → FP1	1			0.931	0.827261	0.934785
Financial performance → FP2	1.020	0.052	19.453	0.955
Financial performance → FP3	0.820	0.055	14.965	0.852
Non-financial performance → NFP1	1			0.822	0.642955	0.781531
Non-financial performance → NFP3	1.042	0.203	5.146	0.697

, the AVE and CR were evaluated to verify the reliability and validity of the model. We found that the AVE values were above the threshold value of 0.5 (digital drivers = 0.735507, digital competencies = 0.776651, efficiency-centered business model = 0.872065, novelty-centered business model = 0.705953, financial performance = 0.827261, and non-financial performance = 0.642955). In addition, all the CR values were above the threshold value of 0.7 (digital drivers = 0.917482, digital competencies = 0.945606, efficiency-centered business model = 0.931586, novelty-centered business model = 0.923029, financial performance = 0.934785, and non-financial performance = 0.781531). Therefore, the validity and reliability of the model were confirmed.

4.2. Results of the Hypothesis Testing

In the research concept model in Figure 1, the final model as shown in Figure 2 was confirmed while removing variables and paths to improve model fit. Hypothesis testing for this model is as follows.

4.2.1. Hypothesis 1: Digital Drivers → Digital Competencies

As shown in

Table 8. Results of the hypothesis testing.

Category	Hypothesis	Standard Error	p-Value		Supported?
Digital drivers → Digital competencies	H1	0.597	0.130	***	Yes
Digital competencies → Efficiency-centered business model	H2	0.504	0.099	***	Yes
Digital competencies → Novelty-centered business model	H3	0.643	0.095	***	Yes
Efficiency-centered business model → Financial performance	H4-1	0.141	0.119	0.239	No
Novelty-centered business model → Non-financial performance	H4-2	0.497	0.091	***	Yes

*** p < 0.001.

, digital drivers were confirmed to positively affect digital competencies (0.597, p < 0.001), supporting Hypothesis 1. It was thus confirmed that external digital drivers such as customers’ demand for the digitization of goods and services, awareness of new external digital technologies, awareness of competitors’ adoption of digital technology, and spread of the digitalization of sales forms and methods influence digital competencies. In addition, internal digital drivers such as recognizing the need to adopt new digital technology, acquisition of digital technology and identification of opportunities, internalization of digital technology, and products/services using digital technology are important for the tendency to challenge differentiation through digital technology and enhance operational efficiency through digitalization.

4.2.2. Hypothesis 2: Digital Competencies → Efficiency-Centered Business Model

Digital competencies were confirmed to positively impact the efficiency-centered business model (0.504, p < 0.001), supporting Hypothesis 2. The acquisition of important digital technology as well as its internalization and the digitalization of the process significantly affected fast transaction speed with customers and the improvement of transaction efficiency.

4.2.3. Hypothesis 3: Digital Competencies → Novelty-Centered Business Model

We found that digital competencies positively affected the novelty-centered business model (0.643, p < 0.001), supporting Hypothesis 3. The acquisition of important digital technology as well as its internalization, the identification of new opportunities using digital technology, and digitization of products and services significantly impacted the promotion of unique and differentiated businesses.

4.2.4. Hypothesis 4: Business Models → Management Performance

The novelty-centered business model was found to positively affect non-financial performance (0.497, p < 0.001), supporting Hypothesis 4-2. We thus confirmed that B2B companies showed better non-financial performance, such as higher customer satisfaction and higher market share, when they provided new products, services, and diversity to customers. However, the effect of the efficiency-centered business model on financial performance was insignificant and, therefore, Hypothesis 4-1 was rejected.

5. Discussion

This study aimed to clarify the concepts and definitions of digital drivers, digital competencies, a digital orientation or business model, and management performance. We also aimed to quantify and analyze the correlations among the factors to determine the suitability of digital transformation. The summary and implications of the hypothesis testing results are as follows. First, both internal and external digital drivers were found to positively impact digital competencies, which could then help develop products, services, and processes. This finding extends the conceptual study of Hrustek et al. [21], which only confirmed the impact of digital drivers through empirical research based on company data.

Second, digital competencies positively affected both the efficiency-centered business model and the novelty-centered business model. Yu et al. [29] found that relationships within the same industry, relationships within heterogeneous industries, relationships with the government, and resource combination activities positively affect efficiency-centered business models. Furthermore, relationships within the industry, relationships with the government, and resource combination activities negatively impact novelty-centered business models. Start-ups in China adopt an efficiency-centered business model that relies on mutual relationships rather than a transformative business model as the level of mutual relationships increases. However, the results confirm that firms’ digital competencies impact both the efficiency-centered and the novelty-centered business models.

Third, in terms of the relationship between such business models and management performance, the novelty-centered business model positively affected non-financial performance. Khin and Ho [39] analyzed the effect of digital orientation and digital competencies on financial/non-financial performance and the mediating effect of digital innovation. Their results showed that digital orientation impacted financial and non-financial performance via digital innovation and that digital competencies also impacted financial and non-financial performance. In this study, we found that the novelty-centered business model positively affected non-financial performance. In Korean semiconductor materials, components, and equipment-related companies and fabless design companies, there was no significant difference between the efficiency-centered business model and financial performance, as sales, operating profits, and cash flow fluctuated greatly depending on customer investment patterns. However, we confirmed that the novelty-centered business model positively affected non-financial management performance, such as higher customer satisfaction, higher market share, and lower employee turnover.

5.1. Theoretical Implications

To measure the model, it is necessary to consider its theoretical significance. Previous studies on digital transformation [21,29,39] have derived limited research findings in specific fields, as they have focused on the internal features of companies and included factors related to digital competencies, digital orientation, digital innovation, business models, and management performance separately. By contrast, this research has significance, as it conducted empirical research on the extent to which competitive strategies, competencies, and management performance affect digital transformation in connection with external environmental characteristics. This integrated model concurs with that of Kim and Ko [48], who studied the differences in companies and the government’s perceptions of the priority of digital transformation success factors. They used key factors and detailed items for transforming an analog company into a digital company. They suggested establishing a progressive digital transformation strategy and policy direction in Korea through mutual understanding and cooperation based on the differences in companies and the government’s perception of importance. However, there was a limit to the generalizability of their findings because of the limited number of research targets. Moreover, their analysis focused on an integrated model highlighting the internal factors of a company. By contrast, our study adopted an integrated approach to digital transformation that integrates and generalizes the external and internal features of a company and targeted semiconductor ecosystem companies undergoing digital transformation in various industries. Digital transformation, a significant recent trend in society, provides an essential competitive edge for future survival and sustainable management in the semiconductor industry and others. In this sense, this study expanded the research scope for applying digital transformation features in other industries.

5.2. Practical Implications

The second significant aspect of this study is the practical implications. In the semiconductor industry, several types of businesses have strengthened their differentiation capabilities rapidly, including those producing devices, such as for memory storage, and global companies in the industry ecosystem that support the device industry, such as materials, components, and equipment-related firms as well as fabless design companies. The findings of this study have implications for the digital transformation of companies in the Korean semiconductor ecosystem and those in the field of capacity building. As mentioned above, global materials, components, and equipment companies have strengthened their differentiation and sustainable management capabilities through the digitization of products, digitalization of operation methods, and digitalization of business models. According to the internal and external digital drivers of digital transformation, small and medium-sized enterprises in Korea are aware of the importance of digital competencies and digital orientation. They have tried to link these to business models but have not yet connected them to financial performance. As discussed earlier, this is attributable to the sizable difference in sales, operating profits, and cash flow in line with customer investment patterns. Ecosystem companies should increase acceptance of new digital technologies and further strengthen their relationships with customers through active utilization and differentiation. In addition, it is necessary to further enhance the competitiveness (quality, cost, and service) of the main businesses to establish the basis for an efficiency-centered business model through interaction of digital transformation. The efforts of the government are important, but the active promotion of individual companies is absolutely necessary. This can be seen as the same meaning as the economic impact mentioned in the study by Peron et al. [43]. However, it is assumed that they have less sufficient digitization levels, maturity, and connections to business models than global companies due to their number of employees and sales volume. It is thus necessary to check the robustness of the measurement items used in this study by analyzing different sized firms. Differences may arise depending on the products/services provided by semiconductor ecosystem companies. This study is significant, as it provides a framework for future research on the characteristics of these ecosystem industries.

5.3. Political and Social Implications

The third significant aspect is the study’s political significance. Kim and Ko [44] indicated that digital transformation must be thoroughly prepared and discussed at the national level rather than the firm level. Transparent and horizontal cooperation between the government and companies must be emphasized and government support must be provided (e.g., legislation, regulation, systems, and support programs). This research is significant, as it clarified the policy support that should be followed for materials, components, and equipment-related firms as well as fabless design companies. This study also recommended that companies and the government cooperate by introducing digital technology such as digital twins to improve the efficiency of operations, raising the security of AI to differentiate products, creating supplementary institutions that connect to business models, and considering deregulation.

In the social aspect, as shown in Peron et al. [43], the acceptance of new technologies based on digital and changes in the business models can contribute to inducing innovation within companies, integration scattered knowledge, and making decisions in new directions. In addition, it is important for employees to reduce fatigue caused by inefficient work, shorten work hours, and enable real-time management.

5.4. Generalizability of the Findings

The results of the study on digital transformation characteristics (digital drivers, digital competencies, business models, and management performance) in semiconductor industry ecosystem companies can be applied to other industries and ecosystems. Novelty-centered business model companies through digital competencies positively affected non-financial performance. However, a company’s financial performance is greatly influenced by the market and customer environment, and competitive characteristics in addition to digital competencies and efficiency-centered business model. In other industries, whether they pursue two business models(efficiency-centered and novelty-centered) at the same time or focus on one specific business model, there will be differences in the business model that companies pursue depending on the characteristics of the industry and company. Further research is needed to study these characteristics of companies.

6. Conclusions

Although this study presented an integrated model of digital transformation and conducted an empirical study, it has the following limitations. The first limitation is the number of respondents. Since this study targeted experts such as CEOs and CTOs to understand the characteristics of companies, it was constrained from further increasing the number of research targets. Second, it did not analyze the characteristics of semiconductor ecosystem industries. Since there are differences in the product characteristics, R&D and manufacturing processes, and customer requirements of materials, components, and equipment-related firms as well as fabless design companies, additional research on the links among digital drivers, digital competencies, business models, and management performance is required. Third, additional research is needed on the internal characteristics of companies such as the number of employees and firm age.

Additional research is also necessary to examine the correlations among the internal digital transformation capability variables within companies. According to Correani et al. [49], the rapid growth of digital technology and enormous amount of data collected daily through devices and applications are rapidly changing corporate structures; however, companies may fail to create value from digital transformation due to the disconnect between the establishment and implementation of digital strategies. They suggested an implementation framework based on case studies of the digital transformation projects of ABB, CHN, and Vodafone. The variables are data source (internal and external), data platform, human competencies, partners, AI, information and knowledge, processes, delivery methods, and frameworks with customers. Hence, analyzing the correlations among such internal competency variables in detail would be worthwhile.

Finally, this study provided a novel perspective on various aspects of digital transformation in business ecosystems. We empirically examined and revealed the characteristics of competitive strategies, competencies, and management performance related to digital transformation as well as their connection with the characteristics of the external environment.