Over recent years, there has been a shift in consumer attitudes about wellness, beauty, and healthy ageing awareness and acceptance, generating high demands in the market from new patient segments. The beauty industry was documented to grow by more than 10% a year, also by widening its accessibility primarily via aesthetic clinic chains (Leclerc et al., 2021). Innovative aesthetic equipment developed by manufacturers as well as the broadening variety of options in this industry raised the demand for aesthetic treatments, contributing to profitable expansion (Leclerc et al., 2021).

The skincare industry is expected to rise at an 8.9% compound annual growth rate (CAGR) from $1.7 billion in 2021 to $2.5 billion by 2026 (Globaldata, 2022). Consequently, customer perception should be prioritised. The term embraces the overall assessment of the client’s impression of service to the client’s attitude towards the service (Chahal & Kumari, 2010; Parasuraman et al., 1985; Shie et al., 2022).

This leads to the emphasis of the importance of patient-centric marketing (Bellio & Buccoliero, 2021; Giraldo et al., 2014). Patient centricity means incorporating and prioritising patients’ views and needs to ensure optimum patient’s experience journey (Wolf et al., 2014). Thus, the patient experience becomes pivotal to assessing quality of service delivered by the aesthetic clinic.

Efforts to increase customer experience are inseparable from factors influencing it. An earlier study by Schmitt (1999) pointed out that customer experience is a multidimensional construct that consists of sensory, affective, cognitive and active. However, previous studies (Molinillo et al., 2020; Palací et al., 2019) argued that cognitive and affective dimensions are the most important and distinctive of experience. Another study (Ahn & Picard, 2014) demonstrated that affective experience (AE) and cognitive experience (CE) could predict marketing success. In that sense, this study uses these two kinds of dimensions.

One of the most crucial outcomes of marketing objectives is to keep loyal customers and evaluate through their revisit intention (RI) (Park et al., 2021). It is also worth noting that revisiting intention has become one of the most critical aspects of the healthcare industry (Bellio & Buccoliero, 2021; Woo & Choi, 2021). The dependent variable in this study is the patient’s desire to visit a specific cosmetic clinic as frequently as possible within a given time frame, as specified in aesthetic clinics.

A study on an aesthetic clinic in Thailand identified these preceding factors, such as image, trust, brand loyalty and social media marketing activities; however, the results showed limited predictive capability influencing this outcome (Nurittamont, 2022). A study done in Taiwan (Wu, 2011) showed that satisfaction in aesthetic clinics did not always result in customer loyalty, and dissatisfaction did not necessarily result in a different outcome. Another study by Giraldo et al. (2014) on aesthetic medicine patients found a strong influence from relationship quality to loyalty but less from satisfaction. Aside from the undesired results, more previous studies on customer experience are too general and might not apply to the aesthetic industry. Consequently, further research in aesthetic clinics is required, particularly using patient experience.

According to Endeshaw (2021), many studies on healthcare service quality have been done by a service quality evaluation (SERVQUAL) that mainly referred to Parasuraman et al. (1985). However, that approach puts more emphasis on functional quality and neglects technical and other quality dimensions, such as outcome quality (OQ) that explain the result of the healthcare (Chahal & Kumari, 2010; Swain & Kar, 2018). Therefore, both are equally important and are associated with patient satisfaction.

Currently, the aesthetic industry utilises visual media as the primary tool for the target audience, and because patients these days are more influenced by digital platforms, social media engagement that is rich with visual content becomes an effective channel to engage with the patient (Trunfio & Rossi, 2021). Social media activity in the healthcare industry could affect patients’ motivation to return by luring in new clients while keeping hold of old ones. The most potent element of social media activity is provided through electronic word of mouth (EWOM, or EW). Electronic word of mouth is one method of using online customer communities to effectively promote one’s brand or cause, and also includes basic advice for website and forum moderators on how to make such presentations helpful to the users of their respective online communities. Studies have found that EWOM and RI have a positive association (Abubakar et al., 2017). Unfortunately, few studies are done to explore this phenomenon in the aesthetic clinic.

Based on the aforementioned consideration, three research questions were raised as follows:

To answer those three research questions, this study attempts to integrate functional and technical quality measurement in healthcare services and how patient experience, in turn, affects RI with the moderation of social media engagement. In this study, the influence of product quality (PQ), service encounters (SE), servicescape (SS), OQ and EWOM towards patient experience was explored through the affective and cognitive dimensions. This study has two benefits. Firstly, it helps to analyse and identify the factors that might lead to a profound patient experience, particularly at the aesthetic clinic. Secondly, it helps to determine how these experiences are perceived through the differential subjective process. This empirical study deployed a new model on what actionable factors might be effective in enhancing patient return intention in the aesthetic clinic, which, hence, will result in managerial implications for supporting a more effective business growth strategy.

This quantitative analysis employed cross-sectional data. Targeted population are patients who visited Youthology Aesthetic Clinic^® minimum twice. It is a new, well-established aesthetic clinic with potential of expansion. This study used purposive sampling to collect data from preset objectives or groups (Bougie & Sekaran, 2020). Structured questionnaire indicators examined the conceptual framework’s aspects. This study used a questionnaire modified from previous research, such as PQ from Swain & Kar (2018), SE from Chandon et al. (1997), SS from Bitner (1992), OQ from Kim et al. (2017), EWOM from Hennig-Thurau et al. (2004), CE and AE from Brakus et al. (2009), SME from Choi and Kim (2013) and RI from Khan (2022). Respondents used the Likert scale to rate their agreement with the statements from ‘strongly disagree (1)’ to ‘strongly agree (5)’ (Bougie & Sekaran, 2020).

The conceptual framework, a complex research paradigm, has 14 parts. PLS-SEM was recommended for exploratory research because it could evaluate complex models (Sarstedt et al., 2022). The PLS-SEM analysis was carried out to assess significance using SmartPLS^® Version 3.3, which was selected as it provides a bootstrapping menu to test significance (Memon et al., 2021).

The PLS-SEM analysis began with outer loading from the reflective model to assess indicator reliability. Twenty-one research model indicators satisfied outer loading requirements. Outer and inner models serve as the foundation for the primary method. Outer model is used to assess the validity and reliability of relationships between indicators and corresponding model components. The inner model is used to investigate if there is a meaningful connection between each component in the study model (Sarstedt et al., 2022).

The study had 280 eligible participants. Table 2 shows the study participants’ profiles. Female reactions predominate. Most responders lived in West and South Jakarta. Most respondents are upper-middle class, independent contractors with high purchasing power. They also had skin and face contouring consultations.

Analysing the outer model is the first step in the PLS-SEM model analysis process (Sarstedt et al., 2022). We excluded a few indicators that could not match the threshold value of 0.5, as required in the outer loading results in Table 3 (Hair et al., 2019).

All remaining indicators are reliable in measuring their respective constructs according to the model. The construct validity test was applied by analysing the average variance extracted (AVE). It can be deduced from the result that all indicators in this study model are considered valid to measure their respective constructs collectively (Hair et al., 2019). Heterotrait-monotrait (HT-MT) ratio is then applied to evaluate the discriminant validity. This method was chosen because it is reputed to provide a more accurate value (Hair et al., 2019). As all of the HT-MT values in Table 4 fall considerably below the 0.9 cutoff, it can be said that all of the indicators in this study model have achieved sufficient discrimination in measuring their respective constructs (Sarstedt et al., 2022). As a result, it can be concluded that every indicator in this study model is valid and dependable for measuring each construct explicitly.

Variable inflation factor (VIF) is then used to determine multicollinearity issues. The findings showed all the constructs had inner VIF values below 5 as suggested (Hair et al., 2019); thus, it can be said that there is no multicollinearity issue found in this model (Table 5). The coefficient of determination or R² of CE is 0.421, categorised as weak to moderate predictive accuracy (Table 5), as well as AE (0.462) and RI (0.428). Cognitive experience has a small effect size on RI with an f² value of 0.108, as well as AE on RI (0.116), as seen in Table 7. The value of the construct Q² predict of CE (0.398) and AE (0.454) shows medium predictive relevance, as well as RI (0.429), as shown in Table 6. Based on these results, it can be said that this model has consistency in its predictive ability.

The significance and coefficients of the variables in the structural model were evaluated to see whether or not the hypothesis could be supported. Based on the results shown in Table 8, all of the supported hypotheses have T-statistics greater than 1.645 (one-tailed with alpha 0.05) with p-value equal to or less than 0.05 and confidence interval (CI) 5% and CI 95% follow the direction of the hypotheses and there is no zero value between the ranges of CI 5% and CI 95%. As a result, we conclude that the analysis of the empirical data supports the 11 hypotheses that were created in this study.

Figure 2a and b shows that the moderation of SME to the link between AE and RI is consistent with the moderation of SME to the association between CE and RI. This means that when emotional intention rises, so will RI. Furthermore, as subjective experiences rise, so does the inclination to return.

It can be observed that the difference between +1 SD and −1 SD grows larger below 0 and that the meeting point between the +1 SD line and the −1 SD line is where the value of CE and RI grows. This also applies to AE and RI. The wider the difference between +1 SD and −1 SD, the stronger the influence of the moderating variable. The graph depicts a straight connection, indicating that the more the CE and AE, the greater the patient’s inclination to return.

These findings have management implications for the aesthetic clinic, as the clinic should improve the patient’s CE and AE by either expanding the content of social media to be more instructive and informative or by incorporating a sense of emotion into the material.

From the path linking the independent variables to the dependent variables (Table 9), four pathways are showing significant mediation (T-statistics > 1.645) of CE to RI and four have significant mediation of AE to RI. Thus, the eight pathways are shown to be interconnected and show a link between the antecedents and RI.

In the context of cognitive and experience constructs, the variables must be maintained and continuously improved to achieve desirable OQ. It is worth noting down that the variables that must be a concern for management to be improved and cultivated are EWOM and SS. In cognitive construct indicators, the indicators that must be maintained are EW1 and OQ1. Meanwhile, PQ2, PQ4, SE6 and SE7 have run well but are not considered necessary by the clinic. EW1 and OQ1 are running very well and need to be maintained. Moreover, OQ3, EW4, SS3, EW3 and SS1 are considered essential but are still not performing well. Figure 3 shows importance-performance map analysis (IPMA) analysis results in indicators in estimating AE, which has the most prominent correlation towards RI. In the AE construct, SE is a variable that must be kept constant and enhanced. The factors of EWOM must also be considered for the management to be improved and nurtured. The affective construct indicator, EW1, must be kept up because it serves the clinic effectively, while SS1, SS3, EW3 and EW4 require much attention because they are underperforming but are important for the clinic.

This study examined how patient experience antecedents impact behaviour intention. This research defined patient experience as several encounters across time and emphasised cognitive and emotional consumer reactions (Molinillo et al., 2020; Palací et al., 2019). This study’s conclusion was drawn to illustrate the significance of the structural relationship depicted in Figure 4.

Product quality, SS, OQ and EWOM positively influenced patient experience. Service encounter influenced AE but not CE. Patients judge SE more on the clinic’s medical staff’s empathy, warmth and friendliness. Thus, service emotions matter more. Service interactions may not always result in a favourable experience because consumers may need more than one visit to the facility to build a solid opinion (Voorhees et al., 2017). In contrast to other studies which focus on the promotion (Nurittamont, 2022), this study shows that patient experience of services at the clinic can adequately predict RI directly. It is true that patients remember the clinical services through brands, but the fundamental thing is the deep impression obtained from their experience at the clinic. Thus, it is recommended for clinic managers to routinely conduct patient experience surveys.

Electronic word of mouth, SS, PQ and OQ influenced CE the most. This finding supports the latest research from Hsu (2021) on a beauty enterprise, which shows that EWOM positively influences customer attitudes. The finding regarding EWOM is also in line with the study by Koay et al. (2020), where EWOM from social media marketing activity can influence brand experience and loyalty. Thus, clinic management must prioritise EWOM with strategic digital marketing. Positive online conversations about the clinic should be monitored and encouraged to become positive EWOM material that promotes clinic services.

This finding elucidates the notion from Bitner (1992) about the importance of SS in superior service. Patients who come to the clinic will evaluate the service based on its utility, such as complete facilities in the waiting room. This study is in line with Jeloudarlou et al. (2022), which showed a favourable correlation between SS and customer experience. Patient cognition also affected clinic PQ and treatment outcomes. Patient knowledge determines PQ. This study verifies patient evaluation-based OQ (Swain & Kar, 2018) in the healthcare service. Thus, the patient or consumer may evaluate the treatment’s success based on pre- and post-treatment findings. This finding is also in line with a study from Kim et al. (2017), which shows that the treatment results significantly affect RI. This OQ variable must account for the duration between treatment and ideal outcomes, as cosmetic clinic treatments frequently take time to show effects.

Parallel to PQ, EWOM, OQ, SS and SE were major AE antecedents. This result is in line with a study by Paulssen and Guru (2020). This reveals that customers’ evaluations of PQ and overall excellence or superiority over alternatives affect how enthusiastically they would use the same thing in the future. Thus, a clinic manager must work to maintain beauty PQ. This also becomes a concern because many unauthorised or counterfeit products are used in beauty clinics, which might harm the patient’s overall health and welfare. It is recommended for clinic managers to always inform consumers the quality certification of the product and the authenticity of the drugs or devices used before starting the treatment. This kind of communication should be the standard procedure before treatment.

This study shows that EWOM is crucial to emotional experiences. Short video storytelling and testimony are emotive. Thus, seeing or hearing EWOM might evoke pleasant feelings like enthusiasm (Fredrickson & Branigan, 2005). The finding of OQ to AE is also in line with a study by Klaus and Maklan (2013), which found that OQ positively influences customer experience and loyalty. Another study by Alnawas and Hemsley-Brown (2018) further supported this result, where it was emphasised that any positive or negative statement made by potential, actual or former customers about a product or company will eventually shape any decisions by information-seeking, specifically in the online marketplace and any reputations involved, hence implying consequences on customer AE. From these findings, it can be suggested that clinic managers may develop their promotional content, with pictures, videos and listening stories that can trigger positive emotions in the patients. This content can amplify the patient experience that has been formed, thereby strengthening their intention to continue using the clinic’s services.

The R² value for CE in the research model assessment is found to be 0.421. This value represents that percentage of CE is 42.1% estimated by PQ, SS, OQ and EWOM. Particularly EWOM serves as the construct with the most prominent ability to estimate the CE. On another side, the R² value for AE is 0.462. This value represents that percentage of CE is 46.2% estimated by all the antecedents, with PQ contributing as the most prominent relationship. These two results are only slightly different where R² is found more in AEs. Furthermore, AE also has a slightly stronger influence on RI than CE (0.334 vs. 0.319). This result confirms previous research findings that AE has a more substantial influence on satisfaction than the CE (Ahn & Picard, 2014; Molinillo et al., 2020) and that the two dimensions can be independent at one time. Moreover, this overall experience reflected that customers have accomplished higher order of specific goals that represent a mix of utilitarian and emotional factors.

Overall, the two characteristics of experience have an equally beneficial impact on RI and align with earlier research that predicts customer loyalty (Choi & Kim, 2013; Giraldo et al., 2014; Jatiyananda et al., 2021; Woo & Choi, 2021). Thus, analysing cosmetic clinic customer experience must include cognitive and emotive elements, different from previous studies that focused on patient satisfaction to evaluate service by beauty clinics (Giraldo et al., 2014). This study shows that both experience dimensions may be used together to analyse antecedents in a new way. Thus, multimodal client experience should be evaluated (Schmitt, 1999). In an aesthetic clinic, the AE and CE can considerably predict client intentions to return.

Unexpectedly, SME moderation was not significant in the research findings. In another study, social media increased beauty clinic RI (Jatiyananda et al., 2021). Even if the coefficient value is low, an interesting phenomenon is discovered where AE strengthens RI while CE weakens it. Social networking weakened loyalty in a healthcare research (Heriyanto & Antonio, 2021). This is also mentioned in a prior study (Dolan et al., 2015) that social media consumer contact may have negative influence. The clinic’s social media material may not meet the client category. Another study indicated that brand experience regulated by social media marketing is stronger in the millennial subgroup than the non-millennial subgroup, suggesting that social media can influence certain target audiences more than others (Khan, 2022).

It is crucial to underline that social media is always public, searchable, shareable, volatile and dynamic. This may elicit various responses from target audiences, some of which may be negative. Patients may be influenced by third-party comments or overexpectations. Further studies are needed to investigate this finding.

This research model identified R² value during the analysis of empirical data, which is classified as medium predictive accuracy, despite when viewed from f² it categorised has a large effect size (f² > 0.35) to RI respectively. When using out-of-sample examination from Q² predict value, that is considered more advanced in assessing predictive power (Sarstedt et al., 2022). This research model classified has a medium predictive relevance. Therefore, this research model can be considered adequate to predict RI as a consequence of patient experience at the aesthetic clinic. Likewise, the R² values for both experience dimensions were classified as moderate. Thus, the antecedents in this study can be considered satisfactory in predicting patient experience.

Through this research model, new insight can be contributed, namely integrating technical quality such as treatment OQ, infrastructural quality such as SS and social support quality such as EWOM. This is following the relevant proposition in the science of healthcare management that service quality in healthcare is broader than just functional quality (Swain & Kar, 2018). Also, the conventional SERVQUAL model with five dimensions is no longer sufficient to assess patient experience (Chahal & Kumari, 2010; Swain & Kar, 2018). This study fulfilled the previous research suggestions that it is necessary to measure the quality of service specifically for different types of health services (Endeshaw, 2021), such as the services provided by the aesthetic clinic. Apart from some limitations, this study which is deliberately done in aesthetic clinic services has proposed a model that can be developed in further research.

The identified antecedents of this study – PQ, EWOM, SS, OQ, and SE – are likely to affect patient experience. Parallel cognitive and affective experiences processed the antecedents differently. For instance, EWOM mostly affects CE and PQ affects AE.

This study improves patient experience multidimensional assessment for academics and managers. The experiential marketing theory showed that the two elements of patient experience might predict cosmetic clinic service RI. The model’s predictive power makes it suitable for a similar healthcare industry. This empirical study can help managers design a patient-centric company plan to ensure that existing clients will return to the aesthetic clinic.