Social media are deemed as platforms typically in the marketing areas where individuals connect with one another and exchange knowledge and/or opinions (Kaplan and Haenlein, 2010). Social media have caused three significant changes in business activities due to their unique characteristics such as dynamic, interconnected, egalitarian, and interactive. For example, social media make it feasible for businesses and customers to connect in ways that weren't before conceivable (Lee and Kim, 2018). Different channels, such as social networking sites like Facebook and Instagram, microblogging services like Twitter, and content communities like YouTube and Tiktok, establish and improve this connectivity by enabling social relationships to develop from shared values and interests (Kaplan and Haenlein, 2010; Kwon and Cho, 2019). In addition, social media have changed how businesses and customers interact with and affect each other. Social interaction entails active (i.e., through communication) and passive (i.e., observations) actions that directly or indirectly affect the decisions and consumption patterns of others (Chen et al., 2011). Due to those reasons, social media are actively being utilized by firms and organizations for a variety of purposes such as leveraging marketing activities, increasing brand awareness, promoting products or services, and managing customer relationships (Parveen, Jaafar et al., 2015). Existing studies also report that social media promotes social interactions that provide firms with financial value (Muller and Peres, 2019). While several scholars have sought to conceptualize social media, their findings are mixed (Tarsakoo and Charoensukmongkol, 2019) because different contexts are used in conceptualizing the construct of social media: Thus, it is vital to identify the dimensions of social media, in part because the dearth of a common understanding of this term makes it challenging for scholars to understand our findings as intended. In this study, we limit our current view to social media marketing capability from the service organization’s point of view. Academic literature abounds with numerous different definitions and oftentimes decentralizes the definition of social media due to the fact that the nature of technologies is dynamically changing. For example, Kaplan and Haenlein (Reddy, York et al., 2010) view social media as a group of Internet-facilitated applications that are based on the ideological and technological foundations of Web 2.0. They argue that social media enable the creation and exchange of different contents produced by users. In addition, Van De Velde, Meijer et al. (Han and Hyun, 2015) accentuate the nature of message creation in defining the term social media. In terms of the dimensions of social media marketing capability, Fuchs (2021) has revealed several sub-factors such as information, communication, communities, and collaboration. In addition, Whiting and Williams (2013) identify seven factors including social interaction, information seeking, pass time, entertainment, relaxation, communicatory utility, and convenience utility. This study focuses on two distinct characteristics of social media marketing capability: namely, information delivery capability and communication capability for the following reasons. Our literature review found that previous studies have examined various factors in social media marketing, such as social media engagement, trust, and reputation. However, these factors may not necessarily apply to our specific research context. Additionally, there is a scarcity of studies focusing on the specific context of cosmetic services in South Korea during the COVID-19 pandemic. To address this gap, we have chosen to focus on two capabilities: information delivery and communication capabilities. Information delivery capability is crucial for providing timely and accurate information about cosmetic services offered by Korean hospitals, helping to reduce perceived risk. Communication capabilities are essential for building trust and establishing personal connections with patients, especially given the travel restrictions and social distancing measures imposed during the pandemic. During these uncertain times, providing timely and accurate information about hospital safety measures on social media can increase patient confidence in the safety of cosmetic services. Social media platforms also enable hospitals to showcase their services through visual and interactive content, including virtual consultations, which helps international patients understand and choose the services available in Korea. Furthermore, social media provides convenient communication channels for remote consultations, appointment scheduling, and updates, improving accessibility and convenience for overseas patients. In the subsequent sections, we will provide further details on each of these capabilities.

On the one hand, social media as information delivery capability refers to all activities by which cosmetic service hospitals present and deliver information on their services. it is assumed that presenting and delivering information through social media is an effective strategy to raise brand recognition and awareness (Gunelius, 2011). When it comes to our research setting, because overseas patients are not familiar with Korean cosmetic medical services, they are more likely to seek more information about Korean cosmetic medical services. Since information presented via social media is effective in increasing service awareness, visualizing medical treatment processes, and reaching millions of people beyond geographical constraints at a lower cost. Thus, whether hospitals offer genuine and timely information is likely to increase their service recognition and attract the customer’s attention. On the other hand, social media as a communication capability indicates that the utilization of social media enables businesses to control how customers perceive their value. According to Santos-Vijande et al. (2012), communication via social media aids companies in managing customer expectations by enhancing their perceived value and image. This study views social marketing communication capability in the context of medical tourism as the hospitals’ ability to shape customer value perceptions by employing social media channels for communication activities (i.e., sales promotions, public relations, micro-marketing, and personal selling). For example, hospitals can use various social media platforms for advertising by offering exclusive discounts to certain clients to make them feel special (Gunelius, 2011): thus, customers are highly engaged in their marketing activities. Communication with customers can be more effective when there is an interaction between the service provider and the consumer (Sharma and Patterson, 1999). For example, after receiving information about a particular medical procedure, customers have a tendency to have questions and explore more about that certain procedure or might be interested in the interaction with the medical staff or coordinators to discuss the details of the procedures (Gardner, Dowden et al., 2010). At this stage, it is imperative for medical organizations to be able to respond quickly and engage in interaction with potential customers.

Based on previous findings and reasonings, we develop a multi-dimensional construct of social media marketing capability and divide social media into two sub-dimensions: information delivery capability and communication capability. Thus, this study explores how each dimension works towards perceived risk.

Before conceptualizing perceived risk, it is important to define the term “risk”. According to the definition by Singh and Bhowal (2011), risk is a term that indicates a possible negative effect on an asset or value that can happen from some current situation or future process. Perceived risk is defined as the subjective assessment that individuals make about the characteristics and seriousness of a risk (Kim et al., 2020; Singh and Bhowal, 2011). Many previous studies on risk and perceived risk were conducted in numerous different fields including banking, tourism (Le and Arcodia, 2018; Barger and Grandey, 2006) food service (Hwang and Choe, 2020) education, e-commerce, agriculture, and healthcare. When predicting consumer behavior in healthcare services, previous studies have paid special attention to patient’s perceived risk because healthcare services are replete with life-or-death situations (Um and Lau, 2018). According to Paek and Hove (2017), it is very important to pay attention to perceived risk and risk communication in health care because it tells what types of risk most people care about and cope with Paek and Hove (2017). While many experts calculate risk as the likelihood of the occurrence of a risky event and its possible negative effect, customers evaluate the risk based on their various experiences and they tend to estimate higher risk levels than the experts do (Rahman, Khattak et al., 2013). When faced with risk, customers are likely to engage in information-seeking behaviors to mitigate their subjective risks (Nam, 2019) because more information reduces anxiety. Unlike other services, healthcare services inherently involve high risks and moving to another country to receive a healthcare service (i.e., medical tourism) involves more time and costs. As a result, individuals need more information and need to communicate with hospital managers to ensure that their physical and financial risks are minimized. In this sense, we assume that reducing perceived risk is a critical factor in making a decision on receiving a Korean cosmetic service. Thus, consistent with previous findings (Dowling and Staelin, 1994; Stone and Gronhaug, 1993), we refer to perceived risk in this study as an individual’s overall anticipation of negative consequences and feelings of uncertainty about the cosmetic service offered by the Korean hospitals and consider that individual perceived risk is formed before the actual service is consumed.

A growing amount of research has been conducted in the investigation and exploration of the concept of customer satisfaction. One of the primary reasons why customer satisfaction has drawn much attention is because it plays a critical role in determining firm performance (Um and Lau, 2018). Many researchers from various academic fields have contributed greatly to the conceptualization of the term customer satisfaction. For example, Torres and Kline (Canales, Kasiske et al., 2006) view customer satisfaction as a person’s evaluation or perception of how a particular product or service is provided. Oliver (1980) emphasizes that satisfaction is a result of and a judgment about a product or service after its consumption. In this sense, Um and Lau (2018) argue that satisfaction is generated when products or services meet an individual’s favorable level. From the provider’s point of view, customer satisfaction matters for three reasons. First, satisfied customers show a strong tendency to repurchase the product or service, the repurchase intention that may contribute to financial performance (Barger and Grandey, 2006). Second, satisfied customers are willing to attract new customers by sharing their experiences and giving positive word-of-mouth (Barger and Grandey, 2006). Finally, service providers keep their satisfied customers from switching to their counter partners and can convert them into loyal customers (Gounaris, Dimitriadis et al., 2010; Leverin and Liljander, 2006). Given that the healthcare industry is a service-oriented industry with an inherent motivation to concentrate on satisfying the needs of customers, we treat satisfaction as a key outcome variable in this research. Therefore, in accordance with previous research findings, we conceptualize satisfaction as the overall evaluation and judgment of international patients about a Korean cosmetic service after receiving Korean cosmetic services.

This study establishes several hypotheses by adopting the protection motivation theory. Protection motivation theory is one of the many theories that attempt to explain perceived risk and risk tolerance (Rogers 1975; Trujillo-Barrera, Pennings et al., 2016). The theory denotes that people are more likely to defend themselves when they expect undesirable outcomes and want to avoid them: thus, they desire to take preventive action (Rogers, 1975). According to protection motivation theory, people tend to take protective action when they are motivated and have the agency to do so, and when there is a relationship between perceived risk and injuries and events. For instance, Sheeran, Harris et al. (Han and Hyun, 2014) discovers that improving risk appraisal components (e.g., perceived risk and perceived severity) has a positive impact on changing intentions and behavior toward safety. Thus, protection motivation theorists argue that the motivation to protect oneself necessitates not only managing the risk but also possessing the knowledge and abilities to take preventative measures. The main tenet of this theory leads us to explore how the use of information delivery capability and communication capability reduces the perceived risk of cosmetic services and such reduced perceived risk could contribute to satisfaction in return. We will introduce each hypothesis with theoretical evidence.

This study assumes that delivery capability can lower perceived risk. Protection motivation theory postulates that individuals have a tendency to protect themselves when they notice or expect undesirable consequences, and try to avoid those types of events and think that doing so can help them protect themselves (Rogers, 1975). Applying the mechanism of protection motivation theory to our research context, this study assumes that international patients, who desire to receive a cosmetic service in South Korea, may have a certain level of risk due to a lack of knowledge about the procedures and outcomes of cosmetic service (Seow, Choong et al., 2021) that they are going to undergo in South Korea. Therefore, they are likely to rely on information provided by the Korean medical clinic from which they receive a cosmetic service in order to reduce the risk of negative outcomes of the cosmetic procedure in South Korea (Wang, Liu-Lastres et al., 2019; Seow, Choong et al., 2021). Consistent with protection motivation theory (Rogers 1975; Hwang and Choe, 2020; Seow, Choong et al., 2021), this study anticipates that the international patients who are seeking cosmetic services in South Korea during the COVID-19 pandemic are likely to gain the motivation to take preventive action by actively accessing social media channels provided by the hospital to minimize their perceived risk associated with the cosmetic services. The information delivered via social media by the clinics is (1) effective in visualizing procedures and outcomes (before- and after treatment), (2) easy to reach millions of people with little cost, and (3) more genuine than that via other traditional marketing channels (i.e., because it also presents the feedback of users who already receive the cosmetic service) (Taubenheim, Long et al. 2008, Kim, Lee et al. 2017). In addition, information via social media is free of linguistic, geographical, and temporal constraints (Okazaki and Taylor, 2013). Those characteristics make information delivered on social media more persuasive to international patients who hold a higher level of risk on a cosmetic service and have a lack of access to the details of its procedures and outcomes. Thus, we establish the following hypothesis that:

H1: Information delivery capability can reduce perceived risk

Protection motivation theory contends that people tend to act in risk-averse ways when they perceive the occurrence of an event to be serious and vulnerable (Rippetoe and Rogers, 1987). An individual forms protection motives through a thorough appraisal of threats and coping mechanisms and decides whether he or she intends to take protective actions (i.e., such as information seeking and communication with service providers) against potential threats (Seow, Choong et al., 2021). Applying this logic to our research contexts, this study argues that, in terms of communication, hospitals can not only shape and manage their customers’ expectations by offering detailed information regarding their cosmetic service via social media channels but also lower their customers’ perceived risk by providing a two-way communication through which interactions between the hospitals and their customers are stimulated. In addition, the extent to which hospitals utilize social media can determine their capabilities for responding quickly to their customer’s requests and offering personalized solutions in return (Huang, Baptista et al., 2015, Wang and Kim 2017). By communicating with potential customers and answering their questions about cosmetic procedures and other inquiries, hospitals could allow international patients to lower their perceived risk through two-way communication and interaction. Thus, this study establishes the following hypothesis that:

H2: Communication capability can reduce perceived risk

Focusing on the medical tourism context, we pay special attention to perceived risk when international patients make a decision on receiving a cosmetic service in Korea. Unlike general healthcare services, medical tourism involves more uncertain and risks because individuals have concerned about not only moving to another country where different cultures and languages appear but also receiving a cosmetic service there (e.g., medical procedures are typically complicated and treatment outcomes are not fully certain) (Um and Lau, 2018). Those circumstances allow overseas patients to rely on more information and communication with medical service providers to make sure that their financial and physical risks are under control (Belanche, Casaló et al., 2012). Thus, hospitals seek to present visual information regarding their services (i.e., price, procedures, before- and after- treatment) and interact with potential international patients in order to mitigate their individual risk levels (Kim, Lee et al., 2017; Bae and Chang, 2021; Li, Larimo et al., 2021). Thus, the social media information delivery and communication capabilities of a hospital may contribute to lowering perceived risk and those capabilities may lead to higher satisfaction through perceived risk mitigated by social media marketing capabilities. Thus, we test the following hypotheses:

H3: Both marketing capabilities can contribute to patient satisfaction via perceived risk

The online and offline (paper-based) survey was designed and carried out in South Korea. As the main target of this survey, we selected international patients who visited Korea to receive cosmetic surgical and non-surgical treatment at private Korean plastic surgery clinics located in Busan Metropolitan City, South Korea. We created a survey in two languages: English and Russian. The reason why we added the Russian version of the survey is because Russian-speaking customers accounted for the majority of the customer base in our survey samples. To make the process of data collection easier, we asked the hospital staff for cooperation. As a result, we could get the patients’ email addresses, and we sent a survey link to their emails. On the very first page of the survey, we introduced respondents to all the details and goals of this research. In addition, we also informed the respondents that all the information they provide will be utilized only during this study, and their responses will certainly be kept confidential. Before the analysis of the results, we examined if there were any missing answers multiple times. In the end, we could obtain 203 responses from the medical visitors. Table 1 presents our sample demographics.

Measurement items for each variable were created by adopting previously validated measures from the relevant literature. Two sub-dimensions of social media marketing were measured with three items each. They were modified from the study by Seo and Park (2018). The perceived risk construct was measured with six items, which were adopted from the work of Kim et al., 2009. Lastly, we measured satisfaction with four items, which were derived from the work conducted by Bae and Chang (2021). A 5-point Likert scale (1 is for strongly disagree, and 5 is for strongly agree) was used to measure each construct. In addition to main variables, such variables as COVID-19 perceived risk, kinds of cosmetic procedures, and the number of receiving cosmetic services abroad served as control variables for the following reasons. At first, COVID-19 perceived risk was selected to control because, regardless of the quality of cosmetic service, a higher level of COVID-19 perceived risk may discourage customers from giving up information-seeking and communicating with service providers. A single item was used to measure a personal level of COVID-19 a higher level of COVID-19 perceived risk in South Korea with a 5-point Likert scale (i.e., 1=low, 5=high). Second, types of cosmetic services (i.e., non-surgical and surgical ones) are another significant factor in determining the decision-making process because people tend to rely on more information and communication if a cosmetic service is deemed complicated and complex. To treat this variable as a dummy, we transformed it into 0= non-surgical and 1=surgical. Finally, the number of receiving cosmetic services abroad was selected to control because personal oversea cosmetic service experiences also determine decision-making processes and consequent behaviors: more experiences reduce a level of a higher level of COVID-19 perceived risk on cosmetic services in Korea. Thus, actual numbers were entered.

This study tests the presence of non-response bias by comparing the answers (i.e. ages, the number of visits, incomes, and marital status) of 20 early respondents with those of the 20 last respondents. The t-test results present no significant difference in their answers, thereby concluding that non-response bias is less likely. Moreover, common method bias is another issue to be tested because a single source was used for the data collection. A common latent factor test was performed (N. Podsakoff, 2003; P. M. Podsakoff, MacKenzie, Lee, & Podsakoff, 2003) and its results show that the produced model fit for the original measurement model and that for the same one with a common latent factor added were found to be similar, showing that common method bias is not a threat to this study (Satorra, 2001).

Several statistical procedures were undertaken before hypothesis testing. First, confirmatory factor analysis (CFA) was conducted to show the validity of the variables. (Refer to Table 2).

The results reveal fit indices (i.e., CMIN/DF = 3.112, comparative fit index (CFI)= .912, Incremental fit index (IFI) = .913, Tucker- Lewis's index (TLI) = .891, Root mean squared error of approximation (RMSEA) = .0102). This means that all the item loadings are statistically significant and the item loading values show a value greater than the benchmark of 0.5. Moreover, we calculated the composite reliability (CR) for each construct which ranged from 0.859 to 0.889: all values being greater than the benchmark of 0.7. Lastly, the average variance extracted for each variable was calculated. The values for AVE ranged from 0.580 to 0.690, all the values being greater than the benchmark of 0.5. Additional estimation of Cronbach’s α for consistency gave the result of a higher degree of internal consistency: the α value of each variable ranged from 0.855 to 0.894. We made a comparison of the square root of an AVE with the absolute value of the correlative coefficients of the latent variables. As illustrated by Table 3, the AVE value for each variable is higher than the squared correlations. Thus, it indicates that discriminant validity is established. Tables 2 and 3 present CFA results and correlation results, respectively.