Understanding the Complex Adoption Behavior of Cloud Services by SMEs Based on Complexity Theory: A Fuzzy Sets Qualitative Comparative Analysis (fsQCA)

To survive in a competitive environment, small andmedium enterprises (SMEs) have had to adapt to the digital environment in order to adjust to customer needs globally, particularly in the post-COVID-19 world. ,e advantages of cloud computing (e.g., flexibility, scalability, and low entry cost) provide opportunities for SMEs with a restricted budget and limited resources. To understand how SMEs adopt cloud computing in a complex digital environment, this study examines how antecedents combine with each other to explain the high adoption of cloud computing. From the perspectives of holism and set theory, we draw on complexity and configuration theories, present a conceptual model including seven antecedents based on the technology-organization-environment framework, and conduct an asymmetric fuzzy-set qualitative comparative analysis. ,rough an empirical study with 123 Chinese companies, we identify nine combinations (configurations) of determinant antecedents that lead to the high adoption of cloud computing. ,e results show that none of the factors are indispensable to explain a high adoption on their own; instead, they are insufficient but necessary parts of the causal combinations that explain a high adoption. ,is study contributes to the literature on cloud computing adoption by extending current knowledge on how antecedents combine to increase the adoption and identify specific patterns of SMEs for whom these factors are essential and greatly influence their adoption.


Introduction
Small and medium enterprises (SMEs) use digital technologies in the current digital era, such as artificial intelligence, smart technologies, and cloud technology, to grow their online businesses alongside their offline operations.
e digital environment provides numerous opportunities for entrepreneurs who start and run SMEs to impact and grow their businesses. To survive in a turbulent competitive environment, SMEs have had to adapt to the digital environment to adjust to customer needs globally, particularly in a post-COVID-19 world. erefore, there is a strong need for digital technologies to improve the business performance, cost advantages, and competitive advantages with a restricted budget and limited resources.
As a representative emerging digital technology, cloud computing has become an important information technology (IT) development trend in recent years. Cloud computing is considered a transformation of IT resources and the service delivery paradigm, and promises to deliver IT resources as a utility, similar to water, electricity, gas, and telephony services [1]. An increasing number of well-known IT service providers have begun to provide cloud services, such as Amazon (Elastic Computing Cloud), Google (App Engine), and Microsoft (Azure). Cloud computing can transform business processes, reduce IT expenditures, deliver real-time applications, provide access to ubiquitous storage, and have unlimited computing power and the potential for market information mobilization. It has therefore been used in various fields such as e-commerce, e-government, and customer relationship management [2][3][4][5][6]. e advantages of cloud computing services (e.g., flexibility, scalability, and low entry cost) also provide opportunities for SMEs. ese opportunities can enable SMEs to improve IT capability with relatively low transaction and switching costs, and to formulate quality IT strategies allowing SMEs to develop within the new technological paradigm shift [7]. Small businesses can improve their operations and productivity enormously by moving to the cloud. Doing so will increase their revenue, thereby setting them ahead of the competition. In addition, the governments of many countries have issued relevant policies to encourage domestic enterprises to use cloud services. For example, the State Council of China issued its "Opinions on Promoting the Innovative Development of Cloud Computing and Cultivating New Types of Information Industry" in January 2015. Although the adoption of cloud services can bring about many benefits to SMEs, not all SMEs are eager to adopt cloud services and are still on the sidelines owing to various challenges, such as privacy and data security issues, a lack of uniform standards for the cloud services industry, and an inadequate regulation of cloud services [8][9][10][11]. In particular, the size and structure of SMEs often make them face specific challenges. Although the adoption of cloud services can increase their competitive advantage, it also involves the consumption of funds. erefore, SMEs need to be extremely cautious when making decisions on the adoption of cloud services [12]. erefore, in response to the call of the country to encourage enterprises to adopt cloud services, thereby accelerating the digital transformation of enterprises and promoting the development of the cloud computing industry, it is important to explore ways to encourage SMEs to adopt cloud computing and take relevant targeted measures.
Numerous studies in this area have used conventional regression-based quantitative methods, such as multiple regression analysis and structural equation modeling, to explore the factors that affect the adoption of cloud services by an enterprise. ese methods assume that relations between variables are symmetric, and adopt an isolated analytical perspective on variables, focusing on exploring the marginal "net effect" of a single variable or a few variables on the adoption of cloud services. It is difficult to explain the interdependence of variables and the complex cause-effect relationships on how variables' configurations affect the results. However, as in real life, it is a common phenomenon that the conditions occurring in management practice are interdependent rather than independent.
ere are few studies on the complex cause-effect relationship between each antecedent variable (e.g., interdependence) or on the different casual configurations required for enterprises to adopt cloud services.
To address such gaps in the literature, we start from the perspectives of holism and set theory, taking 123 SMEs with a high degree of informatization as the empirical research object, using the qualitative comparative analysis method initiated by Ragin [13] to study the configurations of determinants for enterprises adopting cloud services, and to explore the conditions under which enterprises are more inclined to adopt such services. e results of this research will improve theories on cloud service adoption and promote the application of qualitative comparative analysis in the field of information systems (IS). At the same time, theoretical guidance for SMEs to formulate better IT strategies and cloud service providers to provide products that better meet market needs, formulate specific sales strategies, and formulate more scientific policies to promote the development of the cloud computing industry is provided. At the same time, it provides theoretical guidance for SMEs, cloud service providers, and government to formulate better ITstrategies, provide products that better meet market needs, and make more scientific policies to promote the development of the cloud computing industry, respectively.

Cloud Services and eir
Adoption. Cloud computing is defined as "a model for enabling ubiquitous, convenient, ondemand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction" by the National Institute of Standards and Technology (NIST) [14]. Cloud computing refers to both applications delivered as services over the Internet and hardware and systems software in a data center that provide such services [2].
Cloud computing can bring about many benefits to an enterprise and is an innovative computing model that enables SMEs to obtain complex technologies without investing much money. Enterprises can take advantage of the flexibility of cloud computing to increase computing resources as demand increases, which enables enterprises to plan far ahead for provisioning [15,16]. Computational resources are stored in the resource pool, and users can access these scalable and adjustable resources virtually [15]. ese computing resources can be used automatically, and their use does not require any organizational interaction [14,15]. However, not all enterprises are willing to adopt cloud services owing to various challenges, such as privacy and data security issues, a lack of uniform standards for the cloud services industry, and inadequate regulations [8][9][10][11]. erefore, to accelerate the development of the cloud computing industry, numerous studies have used technology adoption theories to explore the determinants of intention and behavior to adopt cloud services. Table 1 lists the empirical studies on cloud service adoption over the past 10 years.
As shown in Table 1, the existing literature mainly uses a regression analysis and structural equation modeling to study the independent effects of factors, such as technology, organization, and environment, on the adoption and assimilation of cloud services in enterprises and medical and educational industries. Few studies have investigated the interdependence between multiple factors. However, to further reveal the influence mechanism of the adoption of cloud services by an enterprise, it is necessary to explore the complex interactions of multiple factors from a holistic perspective. To fill in this gap, we adopt a fuzzy-set 2 Complexity  [38] DOI, TAM SPSS, SEM e intent to adopt CC, actual usage of CC Awareness, upfront cost saving, running cost, risk, data security, availability of good information and communications technology infrastructure, relative advantage, compatibility, complexity, observability, trialability, results demonstrable, ease of use, usefulness, sociocultural factors, the age of the university, the size of the university, the location of the university, the age of university information and communications technology experts and decision-makers [39] TAM SEM Behavioral intention to use Perceived usefulness, perceived ease of use, top management support, training, communication, technological complexity, organization size 4 Complexity qualitative comparative analysis (fsQCA) method based on the set theory to study the main configuration of determinants of cloud service adoption by SMEs, and analyze the joint action and interaction mechanism of various factors on the cloud service adoption by SMEs both comprehensively and systematically.

TOE Framework.
e technology-organization-environment (TOE) framework was proposed to explain the innovation process in the context of an organization [44]. It considers three contexts that influence the adoption of innovation by an organization: technology, organization, and environment context [44]. e TOE framework has been applied to research on the adoption of various technologies [43,45,46]. It has been applied many times to study cloud service adoption at the enterprise level [5,25,31,47]. e technology context refers to the internal and external technologies related to the organization [5,23,36] and technologies that affect the adoption process [31]. Organizational context refers to the descriptive characteristics of an enterprise, such as enterprise size, management structure, degree of centralization, and quality of human resources [5,23,31,36,44,48]. e environment context includes market elements, competitors, regulatory environments, and government policies [5,23,36,44]. e TOE framework is consistent with the Diffusion of Innovation (DOI) proposed by Roger et al. e DOI takes the five technical characteristics of relative advantage, complexity, compatibility, observability, and trialability, as determinants of innovation adoption [49][50][51]. Some researchers in the IS field believe that the institutional (INT) theory can strengthen the environmental context of the TOE framework [48,[52][53][54]. e INT theory defines different institutional pressures, such as the coercive pressure, normative pressure, and mimetic pressure [55]. Institutional pressure has proven to be an essential determinant of innovation adoption, particularly in remote systems such as SaaS [43,56]. However, the DOI and INT theories only consider unilateral factors that affect the process of innovation adoption. Combining DOI, INT theory, and the TOE framework can compensate for this deficiency and further improve the explanatory capability of the model [23]. Many previous studies have adopted integrated models to explore the adoption of various innovations [12,23,36].

Research Model and Antecedents of Cloud
Service Adoption  understanding of what they provide, an increasing number of cloud service providers have been adjusting and improving their related technologies to attract enterprises to adopt such services [40]. Moreover, with an increasing number of enterprises adopting cloud services, cloud service providers are emerging in the market. e characteristics of cloud service providers will also influence other antecedents to a certain extent. For example, the technical and management service capabilities of cloud service providers can influence the relative advantages of cloud services. erefore, provider factors are also essential factors for enterprises to consider when making cloud service adoption decisions [41]. Based on the above analysis, combining the characteristics of cloud services and expanding on the three dimensions of technology, organization, and environment that have been clearly defined in the TOE framework, we propose a cloud service adoption model of SMEs with four different dimensions, namely, technology-organization-environment-provider, and empirically test the relevant antecedents. e research model is shown in Figure 1.

Technology Context.
e two-factor theory of technology adoption recently appeared in the IS field, which is the inheritance and development of Herzberg's two-factor theory of motivation [57][58][59][60]. e two-factor theory divides the factors that affect the user's will into promoting factors (motivators) and inhibiting factors (hygiene) [59]. In the IS field, promoting factors refer to factors that promote users to adopt a specific product or service, whereas inhibiting factors refer to factors that hinder users from adopting them [59,60]. e two-factor theory states that promoting and inhibiting factors are not opposed to each other, and independently affect the user's will [59,61]. e adoption of cloud services by enterprises is an act of adopting new technologies or services, and technical characteristics are the key factors that influence the adoption of innovative technologies by organizations. As an innovative technology, the relative advantage of cloud services can bring competitiveness to an enterprise. However, because of security issues, enterprises can also be exposed to certain risks. erefore, based on the two-factor theory, enterprises need to consider both the relative advantages and perceived security risks of cloud services in terms of technical characteristics when making cloud service adoption decisions.
(i) Relative Advantage. e relative advantage is defined as "the degree to which an innovation is perceived as being better than its precursor" [62]. Existing studies suggest a positive correlation between relative advantage and IT adoption [28,29,31]. When enterprises perceive that an innovative technology has relative advantages, the possibility of adopting this technology will increase [63,64]. Cloud services provide many advantages for adopters, including flexibility, scalability, a low entry cost, pay-per-use, portability, ease of installation and maintenance, and run analyses through the Internet [2,3,25]. Enterprises are usually willing to adopt new technologies with the advantage of improving the business performance and efficiency. erefore, the relative advantages of cloud services are considered essential variables for their adoption by an enterprise.
(ii) Perceived Security Risk. Perceived security risk refers to the risk that users perceive when they believe that cloud service adoption will hinder an organization's performance [42,65]. Cloud services are the integration of storage and computing in a shared multi-user environment. Because they cannot fully understand and determine potential security risks, enterprises add security considerations when deciding whether to adopt a cloud service [66][67][68]. e main security risks perceived by an enterprise in using a cloud service come from data security issues, including a malicious manipulation of private data, a risk of disclosure of private information, and a reduced auditability of the enterprise data. ese security risks will cause economic losses to an enterprise and lead to a decline in its prestige and trust from the public to a certain extent. erefore, we believe that the perceived security risks that cloud services bring to an enterprise will affect an enterprise with its intention to adopt a cloud service.

Organization Context.
Organizational factors include the organizational characteristics and the support of available resources within the organization, which is an internal situational condition that promotes or restricts the adoption of innovative technologies within the organization. As an innovative technology, whether an enterprise adopts a cloud service and includes it in its management decision depends mostly on the top management's attitude and actions inside the enterprise. At the same time, IT competency of an enterprise also promotes or restricts its interest to adopt cloud services to a certain extent. erefore, this study takes top management support and IT competence as essential variables that influence the adoption of cloud services by an enterprise at the organizational level.
(i) Top Management Support. Top management support has proven to be a necessary condition for overcoming organizational resistance to change and ensuring a successful IT implementation [69]. Top management support is the vision, support, and commitment provided by top management to create an ideal environment for adopting a cloud service [70]. Top management support is particularly crucial for the adoption of technology in SMEs, where the CEO or equivalent personnel often has the final say over the organization's information and communications technology strategy and resulting investments [39,71]. When top management recognizes the advantages of cloud services, it may allocate necessary resources for adopting a cloud service and encourage its use by the employees; conversely, when top management does not recognize the benefits of cloud services to the enterprise, it will refuse to adopt such services [23]. erefore, top management support is also considered an essential factor influencing the adoption of cloud services by an enterprise.

Complexity
(ii) IT Competence. IT competence refers to the technologies available in the organization, such as IT infrastructure and IT professionals [72]. IT infrastructure refers to installed platforms that can complement or be replaced by a new technology solution [23,37]. IT professionals are employees who have the expertise required to implement cloud services [36]. Existing studies have found that it is an essential factor influencing IT adoption [36,37,43]. To improve the IT performance, SMEs with weaker IT competence may have more incentives to adopt cloud services. erefore, we believe that the IT competence of an enterprise will affect its willingness to adopt a cloud service.

Environment Context.
Environmental factors refer to external constraints and feasible conditions for organizations to adopt innovative technologies. Any decision made by an enterprise will be affected by the social environment in which the enterprise is located. We believe that two factors at the environmental level-competitive pressure and government support-affect the cloud service adoption decisions of an enterprise.
(i) Competitive Pressure. Competitive pressure refers to the pressure that enterprises experience from competitors in the industry [5,73,74] and has been proven to be an essential factor in technology diffusion in the literature on innovation diffusion [23]. SMEs are largely available, have weak funds, and have poor risk resistance, and thus are more likely to feel pressure from competitors and follow them in the use of new technologies. By adopting cloud services, enterprises can better understand market visibility, collect data more accurately, and achieve a higher operational efficiency, which will lead to more profits for the enterprise [5,75]; therefore, enterprises can better continue to operate in industry competition. We believe that competitive pressure is an essential variable for enterprises adopting cloud services.
(ii) Government Support. Government support means that the government plays a promotional role in the strategic plan and business operations of an enterprise by providing political, economic, and social resources [76]. It includes direct investment, the establishment of relevant funds, the formulation of privilege policies, and the provisioning of specific guidance. SMEs have weak funds, poor fund-raising capabilities, and poor risk resistance, whereas government support can be regarded as a form of social capital that can help enterprises overcome problems such as resource shortages [77]. To accelerate the development of science and technology, many countries have issued relevant policies to promote the development of the cloud computing industry and provide financial support for some cloud computing projects. For example, the State Council of China issued the "Opinions on Promoting the Innovative Development of Cloud Computing and Cultivating New Types of Information Industry" in January 2015, and the Ministry of Industry and Information Technology issued the " ree-year Action Plan for Cloud Computing Development (2017-2019)" in April 2017. Some enterprises that have successfully adopted cloud computing will be rated as benchmark enterprises, with improved social reputation and social status, as well as receiving more political, economic, and social resources [78]. erefore, with the government's support, companies can establish a good corporate image by adopting cloud services, and to a certain extent, gain more economic and social benefits. Given this, government support is considered an essential factor that influences enterprises to adopt cloud services.

Provider Context.
With the development of cloud services and an understanding of cloud services of an enterprise, an increasing number of cloud service providers have been adjusting and improving their related technologies to attract enterprises to adopt cloud services [40]. Furthermore, as an increasing number of enterprises adopt cloud services, cloud service providers are emerging in the market. erefore, we also consider the factors of the cloud providers and believe that their capability is also an essential factor for enterprises to consider when making cloud service adoption decisions.
(i) Provider Capability. Cloud provider capabilities refer to the technical service capabilities of the provider and the management service capabilities required to virtually complete work tasks in specific areas, such as computing capabilities, the ability to reasonably take security measures and disaster recovery measures, and the ability to standardize service standards [41,63,79,80]. Cloud service providers with higher capabilities can usually better protect the private data of an enterprise and achieve higher technological advancement and service standardization. In the  Complexity face of such cloud service providers, enterprises will be more inclined to adopt cloud services to improve the operational efficiency, the security of enterprise data, and the standardization of business processes, and reduce costs in all aspects of the enterprise. erefore, we believe that the capabilities of cloud providers affect the adoption of cloud services by an enterprise.

Data Collection.
e variables in this study were measured using a 5-point Likert scale. To ensure the reliability and validity of the scale, the measurements of the variables should refer to existing mature scales as much as possible, while comprehensively considering the context of Chinese enterprises adopting cloud services. Because some of the reference items were originally in English, according to the translation and back-translation procedures, the students of this major translated the English items into Chinese, and then several master's students back-translated the items into English.
is process ensured the accuracy of the questionnaire translation through a comparison. We contacted organizations such as the Shandong CIO Union to invite their members to help us fill out the questionnaires. ese enterprises have a relatively high degree of informatization, and 240 questionnaires were distributed to enterprises in China through an online survey by Wenjuanxing (http:// www.wenjuan.com). e questionnaires were filled out by grassroots/middle-level/senior managers of the enterprise information department. A total of 123 valid questionnaires were returned, with an effective recovery rate of 51.25%. e descriptive statistics of the sample are shown in Table 2.

Reliability and Validity Measures.
We used Smart PLS software (version 3.0) to conduct a partial least squares analysis on the data. Reliability describes the internal consistency of the measurement results. Reliability testing was measured using Cronbach's α (CA) and composite reliability (CR). rough a calculation, the CA and CR values of each variable are greater than 0.7, and 0.8, respectively, indicating that each variable has good internal consistency and that the measurement model has good reliability. Validity includes content validity and construct validity, and construct validity is measured by convergent and discriminant validity. Convergent validity is used to judge whether different observed variables can be used to measure the same latent variable. e criterion is the average variance extracted (AVE) of the latent variable and the factor loading of the construct. rough the calculation, all AVE values and factor loading values are greater than 0.5, and 0.7, respectively, indicating that the measurement model has a good convergent validity. Discriminant validity refers to whether there are significant differences between latent variables. If the square root of the average variance extracted from a variable is greater than the correlation coefficient between the variable and the other variables, the discriminant validity of this variable is good. e calculation shows that the measurement model has high discriminant validity.

QCA and Data
Analysis. Traditional statistical methods (e.g., regression analysis) mainly study the net effect of independent variables on dependent variables, and it is difficult to explain the interdependence between independent variables. With a qualitative comparative analysis (QCA), which is a case-study-oriented method, it is believed that the result occurs based on the interaction of related elements.
e combination of each element is called the configuration [81][82][83][84]. A QCA is divided into two methods: a clear-set QCA and a fuzzy-set QCA. e value of a variable in a clear-set QCA is zero or 1, and in a fuzzy-set QCA, it is a real number between these same values. Because fsQCA is more distinguishable than a clear-set QCA, fsQCA was used in this study [83]. Similar to the significance level and coefficient of determination in a regression analysis, fsQCA uses the consistency and coverage proposed by Ragin to test the relationship between independent and dependent variables. Consistency can be used to determine whether an independent variable is a sufficient or necessary condition for a dependent variable [82]. When the consistency is greater than 0.8, the independent variable X is considered a sufficient condition for the dependent variable Y, and when the consistency is greater than 0.9, the independent variable X can be considered as a necessary condition for the dependent variable Y. Coverage reflects the explanatory degree of the model. Based on this, we select the relative advantage, perceived security risk, top management support, IT capability, competitive pressure, government support, and provider capability to study the interaction mechanism between variables. A complete QCA analysis process includes a theoretical analysis and the extraction of antecedents, data collection, variable recalibration, selection of case frequency and consistency thresholds, construction of a truth table, a standardized QCA, and the reporting of results using QCA symbols.
We first need to calibrate the measured variables and convert them into sets [81]. e calibration of variables requires three thresholds to be set based on theoretical knowledge and actual conditions: a full-membership threshold, an intermediate membership threshold, and a full nonmembership threshold. e eight variables selected in this study were measured using 2-4 items. e score of each variable for each item was obtained according to the 5-point Likert scale scoring standard, and the average score of each variable was calculated as the final score. We set the maximum and minimum scores of each variable as the full-membership threshold and full nonmembership threshold, respectively. e mean values of the scores were set as the intermediate membership threshold for variables with little difference between the mean and median scores. e mean and median scores of some variables were significantly different, whereas the median scores were 4.33, the samples with a score of 4.33 are more than 20% of the total sample. erefore, the intermediate membership thresholds of these variables need to be set to 4.32-4.34. If the intermediate membership threshold of these variables is set to 4.32, it will lead to a calibrated data distribution. erefore, the intermediate membership threshold of these variables was set to 4.34. By setting the three thresholds, fsQCA converts the scores of these variables into fuzzy scores between zero and 1.

Complexity
After the data calibration, fsQCA 3.0 was used to analyze the truth table. After deleting the cases with frequencies of zero and 1, and a consistency of less than 0.8, 76% of the samples were retained, and a standardized analysis was used to identify the antecedent configuration of enterprises adopting the cloud services. e QCA analysis results report three solutions: complex, intermediate, and parsimonious solutions.
e conditions that appear in both the parsimonious and intermediate solutions are the core conditions, whereas the conditions that only exist in the intermediate solution are peripheral. In addition, variables with a consistency of greater than 0.9 are necessary conditions, and must be the core conditions. e core condition has a strong causal relationship with the outcome, and the peripheral condition has a weaker relationship.

Analysis of Necessity.
We tested the necessary conditions for the presence of cloud service adoption. For cloud service adoption, the consistency values range between 0.41 and 0.87 for both the presence and negation of the casual conditions.

Complexity
No single condition on its own leads to cloud service adoption because none of the casual conditions exceeds a value of 0.9. us, we proceed with a fuzzy-set analysis to identify sufficient combinations of casual conditions that explain the cloud service adoption. e results of the analysis of the necessary conditions are presented in Table 3.

Analysis of Sufficiency.
e findings from fsQCA on the configurations for cloud service adoption by SMEs are presented in Table 4. e solutions were grouped according to the core conditions. Every combination in the solution can explain the same outcome at a specific amount. Consistency values are presented in Table 4 for each solution as well, with all values being higher than the recommended threshold (>0.75). Consistency shows the degree to which the relationship has been approximated, and coverage evaluates the empirical relevance of a consistent subset. e overall consistency was similar to the correlation showing the robustness of the solution. e overall solution coverage indicates the extent to which a high satisfaction may be determined from the existing configurations and is comparable to the R-square value reported in traditional regression analyses. e overall solution coverage of 0.74 indicates that the six solutions explain 74% of the outcome.
Furthermore, fsQCA computes the empirical relevance for each solution by calculating the raw and unique coverage. e raw coverage describes the amount of the outcome explained by a certain alternative solution, whereas the unique coverage describes the amount of the outcome that is exclusively explained by an alternative solution. e coverage of each configuration was greater than zero. us, all configurations are empirically relevant.
For cloud service adoption by SMEs, solutions 1-6 present combinations for which the examined factors may be present or absent, depending on how they combine with each other.

Solution 1.
In configurations 1a-1c, government support is the core condition and is present, which means that enterprises with political, economic, and social resources provided by the government are more inclined to adopt cloud services. Moreover, in configurations 1a-1c, top management support, IT capability, and competitive pressure are peripheral conditions and are absent, indicating that government support is particularly important for enterprises with unclear support of the top management, incomplete IT technology, and no significant pressure from competitors in the same industry to adopt cloud services. In addition, the relative advantage and provider capability in configuration 1a are peripheral conditions and absent, whereas perceived security risk is an irrelevant condition. In configuration 1b, perceived security risk is a core condition and is present, and the provider capability is a peripheral condition and absent, and a relative advantage is an irrelevant condition. For configuration 1c, perceived security risk is a peripheral condition and is present, and a relative advantage is a peripheral condition and is absent, whereas the provider capability is an irrelevant condition. It can be seen from configurations 1a-1c that government support has a significant influence on the decision of an enterprise to adopt cloud services. For enterprises supported by the government, even if other conditions affecting the adoption of cloud services are incomplete, it is easier for them to incorporate cloud services into the decision of the enterprise operation and management. ese solutions explain the cloud service adoption of SMEs at rates of 23.5%, 23.7%, and 27.2%.

Solution 2.
In configurations 2a and 2b, relative advantage, IT capability, and provider capability are the core conditions. is means that the feeling regarding a cloud service can bring about a relative advantage, having a complete IT capability, and for cloud providers with high technology and management service capabilities, enterprises are more inclined to adopt their services. Cloud services, enterprises, and cloud providers, as the three central bodies, all play a vital role in the process of enterprise adoption of cloud services. Only when the three complement and cooperate with each other can they further encourage enterprises to adopt cloud services. Furthermore, in configurations 2a and 2b, government support is the peripheral condition and is absent, indicating that as long as the relative advantages of the cloud services are sufficient, the IT capability of the enterprise is sufficient, and the cloud provider's capability is sufficient to support the provisioning of high-quality cloud services. Even if there is no government financial support, enterprises will also adopt cloud services because of the high performance, versatility, and high efficiency of the cloud platform. In addition, in configuration 2a, top management support is a peripheral condition and is present, and the perceived security risk is a peripheral condition and is absent. In addition, the competitive pressure is an irrelevant condition; in configuration 2b, perceived security risk is a peripheral condition and is present, and top management support and competitive pressure are peripheral conditions and are absent. It can be seen from configurations 2a and 2b that the relative advantages of cloud services and cloud service provider capabilities significantly impact the adoption of cloud services erefore, cloud service providers should strive to improve their service capabilities and make enterprises deeply feel the relative advantages of cloud services, such that more enterprises will willing adopt cloud services. ese solutions explain the cloud service adoption of SMEs at rates of 41.8% and 26.9%, respectively.

Solution 3.
In configuration 3, the relative advantage, ITcapability, government support, and provider capability are core conditions and are present, top management support and competitive pressure are peripheral conditions and are present, and perceived security risk is an irrelevant condition.
is configuration shows that for enterprises with relatively complete IT capabilities, the government support, higher cloud service provider capabilities, and perceived advantages of cloud services will make enterprises more willing to adopt cloud services. In the discussion of configuration 1, it can be seen that government support can promote enterprises to adopt cloud services; in the discussion of configuration 2, it can be seen that the relative advantages of cloud services, the IT capabilities of enterprises, and the capabilities of cloud providers can jointly promote the adoption of cloud services by enterprises. erefore, if an enterprise with relatively complete IT capability chooses a cloud service provider with strong technical service capabilities and management service capabilities and feels the relative advantages of cloud services, it will undoubtedly make a decision to adopt cloud services with the strong support of the government. is configuration further illustrates the vital role that the government plays in promoting the adoption of cloud services by an enterprise. In addition, top management support and competitive pressure in configuration 3 are peripheral conditions and are present, which means that full understanding and trust of cloud services from top management, the resources allocated to cloud services, and the pressure brought by industry competitors will promote the adoption of cloud services by enterprises to a certain extent.
is solution explains cloud service adoption of SMEs at a rate of 55.8%.

Solution 4.
In configuration 4, provider capability is the core condition and is present, and the relative advantage, perceived security risk, top management support, IT capability, competitive pressure, and government support are peripheral conditions and are absent. is indicates that as long as the technical services and management services capability of the cloud service provider are mature and sufficiently trustworthy, even if there are no other conditions that promote the adoption of cloud services by an enterprise, the enterprise will also be willing to entrust the cloud service provider with the implementation of work related to cloud services, deploy websites and systems to the cloud service platforms, and allow cloud service providers to provide service deployment, service choreography, cloud service management, and security and privacy protection for the enterprise. In addition, perceived security risk is the peripheral condition and is absent, indicating that enterprises do not perceive that cloud services will bring risks to the organizational performance under the condition of strong cloud service providers, which will further promote enterprises to adopt cloud services. is solution explains cloud service adoption of SMEs at a rate of 25.4%.

Solution 5.
In configuration 5, relative advantage and IT capability are core conditions and are present; perceived security risk and top management support are peripheral conditions and are present; and competitive pressure, government support, and provider capability are peripheral conditions and are absent, indicating that for enterprises with complete IT capabilities, if they feel that the cloud service will bring relative advantages for the enterprise, they will tend to adopt such services. Top management support is a peripheral condition and is present, indicating that enterprises with top management who support cloud service adoption decisions will be more willing to adopt cloud services. Compared with configuration 2b, it can be seen that even if the enterprise is not overly confident in the technical service ability and management serviceability of the cloud  Cloud service adoption by SMEs  1  2  3  4  5  6  1a 1b Black circles ("•") indicate the presence of a condition, and circles with a cross-out ("⊗") indicate the absence of a condition. Furthermore, large circles indicate core conditions, and small circles indicate peripheral conditions. Blank spaces indicate a "do not care" situation in which the causal condition may be either present or absent. service provider, it is possible to adopt cloud services to take advantage of its relative advantages to improve their business performance and operating efficiency. is is because the top management can feel that the cloud service will bring advantages to the enterprises and be willing to actively participate in the development of the cloud services strategy.
Other peripheral conditions indicate that, regardless of whether the enterprise feels competitive pressure, receives government support, perceives the risks that cloud services will bring, or has a capable cloud service provider, as long as the cloud service brings obvious relative advantages to the enterprise, which has relatively strong IT capability, it will adopt the services. is solution explains the cloud service adoption of SMEs at a rate of 21%.
5.2.6. Solution 6. In configuration 6, IT capability and government support are core conditions and are present, and perceived security risk and top management support are peripheral conditions and are also present. Relative advantage, competitive pressure, and provider capability are peripheral conditions and are absent, indicating that the IT capability of the enterprise is relatively complete. e enterprise is supported by the political, economic, and social resources of the government for adopting cloud services, and will adopt the cloud services. Top management support is a peripheral condition and is present, indicating that the top management is willing to allocate resources to implement cloud services and encourage employees to use cloud services, which will further promote the adoption of cloud services by an enterprise. Other peripheral conditions indicate whether the enterprise perceives the relative advantages or security risks that the cloud services will bring, whether it feels competitive pressure, and whether the capabilities of the cloud service providers are complete; in addition, as long as the enterprise has high IT capabilities and is supported by the government, it will be confident to adopt cloud services owing to the "cloud service coupons" (financial cloud service subsidies) issued by the government. is solution explains the cloud service adoption of SMEs at a rate of 20.8%.

Conclusion
To survive in a competitive environment, SMEs have had to adapt to the digital environment and adjust to customer needs globally, particularly in a post-COVID-19 world. e advantages of cloud computing (e.g., flexibility, scalability, and low entry cost) provide opportunities for SMEs with a restricted budget and limited resources. Small businesses can improve their operations and productivity enormously by moving to the cloud. Doing so will increase their revenue, thereby setting them ahead of the competition. To accelerate the digital transformation of SMEs and promote the development of the cloud computing industry, it is of great significance to explore ways to encourage SMEs to adopt cloud computing and make relevant and targeted measures.
SMEs need to consider multiple factors when incorporating cloud service adoption into their business management decisions. We argue that different combinations (or configurations) of multiple antecedents can explain SMEs' adoption of cloud services. To identify such combinations, we build on complexity and configuration theories, propose a conceptual model that includes seven antecedents (i.e., relative advantage, perceived security risk, top management support, IT competence, competitive pressure, government support, provider capability), and use fsQCA (asymmetric) to identify nine configurations for the adoption of cloud service.
e findings show that government support is present in three out of six solutions and always as a core factor, suggesting that it plays a critical role in driving SMEs to adopt cloud services. e Chinese government, particularly local governments, can significantly influence the business decisions of an enterprise. e government usually formulates policies conducive to a particular business in specific regions and industries, or issues policies for enterprises of a specific size and ownership type. If necessary, the government may even provide financial support to enterprises to encourage them to comply with government guidelines. Cai et al. [76] proposed that government support can promote enterprises to integrate the supply chain information, reflecting the influence of government support on enterprise decisionmaking.
Our findings also show that provider capability is present in three out of six solutions and always as a core factor, indicating that the strong technical service capabilities and management service capabilities of a cloud service provider can encourage an SME to adopt its cloud services. Indeed, Liang et al. [41] proposed that the capability of a cloud provider can affect the cloud provider support and thus the adoption of an e-government cloud, but it has not proven to have a significant promoting effect. Fu et al. [85] pointed out that the ability of a cloud service provider will affect the adoption of cloud computing by an enterprise, although research results have shown that this is not an extremely important factor. However, in this study, provider capability is present as the core condition in the three configurations, indicating that this factor has an essential effect on the adoption of cloud services of an SME.
In addition, our findings show that a relative advantage is present in three out of six solutions and is always a core factor. Many studies prove that the relative advantage can significantly promote the adoption of cloud services by an enterprise [23,28]. e results of this study confirm this conclusion.
erefore, cloud service providers should continuously strive to improve the quality of their services, and thus enterprises can further feel the relative advantages of cloud services, such as flexibility, scalability, low entry, and computing cost.
Finally, our findings show that IT competence is present in four out of six solutions and is always as a core factor, suggesting that it plays a critical role in driving SMEs to adopt cloud services. Oliveira et al. [23] and Martins et al. [36] proposed and proved that technology competence could significantly positively influence SaaS adoption by enterprises. e results of this study further indicate that the IT capabilities of an enterprise play an essential role in promoting the adoption of cloud services. Enterprises should strive to improve their own IT capabilities, increase investment in IT hardware and software, and improve the quality of the IT staff to prepare for the adoption of cloud services at any time.
6.1. eoretical Implications. In the context of the country using cloud computing as a development strategy, we need to increase our understanding of SME adoption and investigate internal factors and how they lead to more adoptions. Our findings identify multiple combinations that explain cloud service adoption in SMEs and offer two main contributions.
is study systematically sorted the latest research on the factors that affect cloud service adoption through a literature review. After such sorting, it was found that the existing literature mostly uses the TOE framework, DOI theory, and technology acceptance model as the theoretical basis when proposing research models, and mostly uses partial least squares (PLS), regression analysis, and structural equation modeling (SEM) as research methods. In recent years, after some theoretical frameworks have become more mature, and more research methods have gradually emerged, no research has specifically reviewed the latest studies on the factors affecting cloud service adoption. e combined effort in the study can help scholars find the research gap in this type of research, explore more factors that affect the adoption of cloud services, establish a more scientific and systematic research model based on existing research, and conduct innovative research in this field.
Another theoretical implication of this study lies in the innovative use of new research methods. We use fsQCA in the research field of factors that influence enterprises to adopt cloud services. Existing studies have mostly used PLS, regression analysis, and SEM to explore the factors affecting the adoption of cloud services of an enterprise. ese methods assume that the relations between variables are symmetric [86,87], but the relations can also be asymmetric in real life. ese methods are also dedicated to studying the independent effects of independent variables on dependent variables, although the relationship between independent variables in real life is often interdependent, and they will work together to lead to specific results. is situation occurs when SMEs make cloud service adoption decisions. ey often need to consider multiple factors to decide whether to adopt a cloud service, and these factors often exist simultaneously and affect the decision-making. In addition, the research of relevant scholars also shows that there is a causal asymmetry in an enterprise adoption of cloud services; for example, Tashkandi and Al-Jabri [28] pointed out that the relative advantage of cloud services has a significant positive impact on cloud service adoption, whereas Low et al. [5] verified a significant negative impact. Building on complexity and configuration theories, fsQCA can capture the causal asymmetry between many factors and the adoption of cloud services by an enterprise, resulting in new research models and theories. Based on the TOE framework, this study uses fsQCA to study the joint effects of the relative advantage, perceived security risk, top management support, IT capability, competitive pressure, government support, and provider capability on the adoption of cloud services, and identifies nine paths that lead an enterprise to adopt cloud services. is study innovatively uses the fsQCA to expand the conclusions of existing research, and at the same time, will provide researchers with a new way to think and operate tools to continuously enrich the research in this field.
In addition, this research enriches the application of QCA in the field of management. In the management area, existing studies have used the QCA method in human resource management, innovation management, international business, entrepreneurial management, knowledge management, corporate governance, social responsibility, stakeholder management, strategic management, organizational design, project management, and marketing management; however, no scholars have used the QCA method in the field of enterprise adoption of cloud services. is study uses fsQCA to explore the nine paths that will lead enterprises to adopt cloud services, which opens up new areas for applying the QCA method, further enhancing the understanding of this method for management scholars, and promoting the development of management theory.

Practical Implications.
e results of this study have important guiding significance and enlightenment for enterprise management practitioners. With the in-depth development of cloud computing technology, cloud service providers have landed and officially commercialized both at home and abroad, and the term "cloud" has become the norm for corporate development. Particularly since the outbreak of COVID-19 in early 2020, the Ministry of Industry and Information Technology of China has required the use of cloud computing and other technologies to help enterprises resume work and production. Cloud services can drive process and business innovations and become a new profit growth point for enterprises. However, in the face of the state promotion of enterprises to adopt cloud services and the benefits that cloud services can bring to them, some enterprises have still refused to adopt cloud services owing to a weak awareness and the security issues of cloud services.
is study identified nine paths that lead enterprises to adopt cloud services. e results will provide relevant practical guidance for governments, cloud service providers, and enterprises to promote the adoption of cloud services, thereby enhancing the leading position and firm competitiveness of the cloud computing industry in China.
First, the research results show that government support is present as a core factor in three out of six solutions, indicating that it plays a vital positive role in the adoption of cloud services by enterprises. erefore, to promote the use of cloud computing to accelerate the digitalization, networking, and intelligent transformation of enterprises, and to promote the development of the cloud computing industry, governments at all levels should correctly and rationally understand the development path of cloud services and scientifically promote the adoption of cloud services by Complexity 13 enterprises. us, we propose two specific suggestions for governments. On the one hand, the government should strengthen the infrastructure of cloud services, formulate more preferential policies, and increase support for enterprises to adopt cloud services. At present, to promote the coordinated development of industries and maximize the creation of corporate value, China has successively formulated and promulgated the "Industrial Internet Plat-form Construction and Promotion Guidelines," "Promoting Enterprise Cloud Implementation Guidelines (2018-2020)," and other policy documents to help SMEs adopt cloud services. In the future, the government should strictly implement the content of the document and put forward more innovative opinions. Meanwhile, the government can set up a team of experts to provide policy interpretation services and legal support to enterprises.
On the other hand, the government should give priority support to enterprises that lack complete IT capabilities and top management support for cloud service adoption, providing them with relevant policy guidance and technical support. is is due to IT capability, and top management support is a peripheral condition and is absent in configuration 1. An expert team can also be organized to provide training to the IT employees of such enterprises, such as educating them on how to comprehend and use cloud services, or how to choose cloud service providers. Furthermore, cloud service "coupons" should be issued to these enterprises to fully promote the application of cloud services and give full-play to the positive role of such services in enterprise operation and management.
Second, provider capability is present in three out of six solutions and is always as a core factor, suggesting that it is an essential subject in the process of enterprises adopting cloud services. erefore, cloud service providers can not only help enterprises improve their IT architecture and achieve business growth but also reduce the pressure on enterprise operation and maintenance through high-quality services. To attract more SMEs to adopt cloud services, cloud service providers should continuously improve their technical service capabilities and management service capabilities and achieve higher technological advancement and service standardization, thereby improving the stability, security, and availability of the cloud platform infrastructure.
ese measures can further enhance their relative advantages. ey should enhance their own ability to protect user data security by building a data security capability scale, full life cycle security capability requirements, and data security capability dimensions; thus, enterprises can safely outsource their work related to cloud services to cloud service providers. At the same time, the providers should provide good after-sales services to enterprises that have already adopted cloud services to increase their satisfaction with cloud services and form a word-of-mouth effect.
ird, because IT capability is a core condition and is present in four out of six solutions, enterprises should focus on building their own IT capabilities, increase investment in IT hardware and software, deploy a good IT environment and essential software and hardware facilities, attach importance to the development of IT departments, and form excellent IT teams to support enterprises in applying cloud services better, thereby promoting the construction of enterprise informatization. At the same time, top management of the enterprises should promote and popularize cloud services, actively participate in the formulation of the cloud service vision and strategy, improve the ability of employees to accept new technologies, and focus on the promotion of IT department managers.
is will provide a solid foundation for enterprises to deploy cloud services, accelerate innovation, and reduce costs to ensure that enterprises remain competitive in the wave of industry adoption of cloud services.

Limitations and Directions for Future Work.
As a limitation of this study, the number of factors considered was relatively small. Only seven factors were studied on the joint effect of enterprise adoption of cloud services. In addition, the research object of this study was only SMEs in China. Because the sample selection is restricted by region and scale, the universality of the nine configurations that can enable enterprises to adopt cloud services obtained in this study needs to be further explored and analyzed. Based on this research, in the future, we will use the QCA method to study other factors that lead enterprises to adopt cloud services and expand the scope of research objects to enterprises in Asia, further increasing the breadth and depth of this research.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.