The Use of AISAS Method in Telemedicine Advertisement (Sentence Case)

Background: The COVID-19 pandemic has accelerated digitalization in the health industry, indicating a transition in services toward telemedicine. In Indonesia, telemedicine is quickly expanding in tandem with technological advancements and legal regulations, with users increasing by 700 percent in the first year of 2020. This is because telemedicine is seen as a cutting-edge approach of dealing with the COVID-19 problem. Looking at the rise of digital marketing in the company and in the current market, it is not impossible for every hospital to begin using digital marketing in every activity, using various methods such as setting up websites and setting up social media. Thus, internet technology leads an advertising company named Dentsu to introduce a new buying process, namely AISAS (Attention, Interest, Search, Action, Share). Objective: This study is intended to analyze the use of AISAS method in telemedicine advertisement. Methods: This research involved 45 respondents. Respondents consisted of social media followers from hospital social media account selected by the purposive sampling. The SEM-PLS analysis in this research was used to analyze the direct effect of social media on purchase intention based on AISAS modeling. Results and Discussion: The results of this study indicate that social media has a significant and positive effect on the attention, interest and search factors of consumers. Meanwhile attention, interest, and search don’t have a significant effect toward purchase intention. Purchase intention has a significant and positive effect on action. Action doesn’t have significant effect to share. Conclusions: This study indicated telemedicine advertisement using social media as media promotion with AISAS modelling does not increase customer’s purchase intention.


Introduction
Therefore, this study aim to analyze the effect of social media on the AISAS model's components, analyze the effect of AISAS model's components toward consumer purchase intention, and to analyze consumer purchase intention toward action in AISAS model's component.

Ethical approval
This research followed the accepted ethical guidelines for doing research with human respondents for ethical approval. The researchers received Ethical Approval No. 2187/KEP-UNISA/VII/2022 from Health Research Ethics Committee, Universitas 'Aisyiyah Yogyakarta. Indonesia. Address: Kampus Terpadu, Jl. Siliwangi (Ringroad Barat) No. 63 Nogotirto, Gamping, Sleman, Yogyakarta 55292, Indonesia. Respondents were informed about the research's objectives, risks, and advantages of participation, and they were encouraged to ask any questions they had about the survey. Respondents were assured of the secrecy and privacy of their responses, which helped eliminate the potential bias introduced by self-reported data.

Participant characteristics and research design
There was a total of 45 respondents in this survey. The criteria used in determining the sample were: (1) respondents were social media followers of Instagram belonging to @rumahsakitjih, @rspantirapihyogyakarta, @rssardjito_official, @rsbethesdajogja, @pkugamping, @pkujogjamedia, and @siloamyogyakarta; (2) followers who already use the telemedicine services; (3) aged 16 years old. This study is a analytic observational study using cross sectional approach

Sampling procedures
This study the population cover hospitals with telemedicine service and all the patient who follow the social media of @rumahsakitjih, @rspantirapihyogyakarta, @rssardjito_official, @rsbethesdajogja, @pkugamping, @pkujogjamedia, and @siloamyogyakarta, and have experience of using telemedicine service in the period of January 2021 -December 2021.

Sample size, power, and precision
Chin (1998) suggested that the sample size requirement of PLS should be collected at 10 times the dimension of most question items. The dimension of most question items is share and action. There are 4 questions items. Therefore, the minimum sample size for research must be at least 40.

Measures and data collection
Data used in this research were primary and secondary data. Primary data used in this study were obtained based on an online questionnaire distributed to social media followers of @rumahsakitjih, @rspantirapihyogyakarta, @rssardjito_official, Anky Angga Alhudha, Winny Setyonugroho, Firman Pribadi/ KESANS The Use of AISAS Method in Telemedicine Advertisement (Sentence Case) @rsbethesdajogja, @pkugamping, @pkujogjamedia, and @siloamyogyakarta. Secondary data used in this research were obtained from books, journals, and the internet.

Data analysis
Data analysis techniques used quantitative analysis. Indicator measurements were made using a Likert scale with a value of 1 to 5. The provisions of the Likert scale are: 1) Strongly Disagree, 2) Disagree, 3) Fairly Agree, 4) Agree, and 5) Strongly Agree.
Data processing of the research used descriptive analysis and SEM PLS analysis. Descriptive analysis was used to analyze the description of the characteristics of followers of @rumahsakitjih, @rspantirapihyogyakarta, @rssardjito_official, @rsbethesdajogja, @pkugamping, @pkujogjamedia, and @siloamyogyakarta social media accounts.
PLS-SEM was used to examine the AISAS model. This decision was made based on the characteristics of the constructs used in the AISAS model. It is the preferred technique if the study delves into prediction. Hence, the use of PLS-SEM in the present study provides a better explanation of the underlying constructs, and their relationships as well as the prediction quality in the AISAS model. The software we use to analyze is SmartPLS 3.0.
The SEM-PLS analysis in this research was used to analyze the direct effect of social media on purchase intention based on AISAS modeling. The hypotheses in Table  1 are formulated in Figure 1 that shows the conceptual research framework. We use SmartPLS 3.0 as a software to analyze.

Study Population Characteristics
Forty five (45) followers of Instagram social media answered the questionnaire. The results of the research were concluded by majority of answers in the questionnaire. Most respondents were women 26(57.7% of respondents). Most respondents were 16 to 25 years old (57.7% of respondents). Most respondent's recent education is bachelor degree 29(64.5% of respondents).

Table 2
Respondents' Characteristics In Figure 2.0 it shows the followers of each hospital. The respondents may follow one or more social media of each hospital. Total social media account who has most followers on Instagram is 22. Anky Angga Alhudha, Winny Setyonugroho, Firman Pribadi/ KESANS The Use of AISAS Method in Telemedicine Advertisement (Sentence Case)

Analysis of Structural Equation Model -Partial Least Square (SEM-PLS)
The analysis consisted of two parts that were outer model evaluation and inner model evaluation. This section, it will explain the evaluation of each model 1) Evaluation of Measurement Model (Outer Model) To test the validity and reliability researcher will use SmartPLS 3.0 program. The process in validity testing is using convergent validity which correlate score of each component with construct score and create loading factor value. The construct will be valid if the loading factor values is >0.7. In reliability testing to determine the reliability of the variable researcher will use the composite reliability score and cronbach alpha score. If the score were larger than 0.7 high levels of internal consistency reliability have been demonstrated among all the variables (Wong, 2013).   3 shown that in the construct model before dropping found the results of loading factor in indicators IN1, AC3, AC4, SH2, SH3, and SH4 were <0.7 therefore its eliminated. After removing indicators, each variable's loading factor value met the specified rules of thumb.
Discriminant validity was measured by comparing the roots of AVE of a construct must be higher than the correlation between these latent variables. Discriminant validity is assessed using by looking at the value of cross-loading in  A construct was considered reliable if it had a Cronbach's alpha value of 0.6 and a composite reliability value ≥ of 0.7. All variables and indicators of structural model variables after dropping had construct reliability tests shown in table 5.

2) Evaluation of Structural Model (Inner Model)
The main purpose of structural model testing is to find correlation between each construct which tested using t test in partial least square. Structural or inner model can be measured by looking at the R-Square model value which show how big is the effect for each variable in the model. Next step is look up in the path coefficient value which acquired from bootsrapping method. If the t value is >1.96 (significane 5%) the hypothese is accepted. Furthermore, based on the p-value, if the p-value < α (0.05), then H0 is rejected and vice versa (Wong, 2013).  Table 7 Path coefficient Inner Model Table 7 shows the result of accepted hypotheses have a probability value of less than 0.05, and the T-statistics value is greater than 1.96. From 8 hypotheses, 4 hypotheses are accepted and have significant effects, such as H1, H2, H3, H7; whereas H4, H5, H6, and H8 are rejected and have no significant effect.

Discussion
This paper analyze the use of AISAS model in telemedicine advertisement in social media in enhancing customer purchase intention. The results presented in this study show that social media has significant effect toward attention. This result in accordance with Rini et al (Rini et al., 2018) which shows that social media effectively in gaining attention from followers in social media. Social media significantly affect attention because social media users will be interested in seeing more advertised products if messages are able to attract attention. The first stage for internet businesses to become known, recognized, and remembered by customers is to use a message that grabs attention (Fannani et al., 2020). In other hands social media shown has a significant effect on interest.
Gulseven (Gulseven, 2018) and Kim et al(Kim et al., 2017) says if followers of the social media like the messages and informations provided, it will improve interest in each individual toward the advertisement or the product that promoted in social media. Social media also has a significant effect on search. In previous study by Rini (Rini et al., 2018), shows that advertising activity in social media which is persuasive and informative to the customer, it will create question within the customers and ignite their curiosity about the product, so the customer will do search about the products. In line with the study by Fannani et al (Fannani et al., 2020), social media affect the customer to do search so customer can put trust on the information that is given to them.
On the other hand analysis on attention, interest, and search don't give a significant effect toward purchase intetion. Previous study by Raji et al(Raji et al., 2019), indicate content from the advertisement in social media doesn't have significant effect on purchase intentions. In line with the study by Abdurrahim et al (Abdurrahim et al., 2019), that interest has a negative effect toward customer's decision, because customer desires to make purchases are influenced by factors other than just their own interests. Advertisement in social media has a limit, it can be due to the advertisement's ambiguous and constrained information. Customers need to acquire more information to make purchasing selections because social media presence has restrictions on it, such as the number of words that can be displayed and the length of the show, and customers who do search up information about a product can become less interested in it (Cheah et al., 2019;Ruswandi et al., 2021). This search process is crucial, especially in the Net Generation, which makes using the internet to find information important (Xue et al., 2021). Based on this study, telemedicine advertisement does not fulfil the customer's expectations thus doesn't increase the purchase intention of the customer to buy or use the product (Sugiyama & Andree, 2011). It is somewhat surprising that attention, interest, and search has no significance toward purchase intention. It highly possible the customer give attention, interest, and do search to the advertisement yet they don't buy the product because they are in healthy condition, thus they feel they don't need to purchase the telemedicine services.
This research indicates that purchase intention has significant effect toward action. The consumer's decision-making process is influenced by how they think about the product (Fannani et al., 2020). If a social media follower has a compelling enough argument to make a purchase, they will do so. Thus customers who are interested in the product will likely to purchase it (Abdurrahim et al., 2019;Jane et al., 2013). The relationship between action and search are not significant. This result is in accordance with the study by Jane et al (Jane et al., 2013). Action doesn't always end up with share. It possible due to the influence of external factors studied, such as customer satisfaction is not fulfilled so that customer feel reluctant to provide a review of the product. Additionally, emotional differences associated to individual impulses to offer comments may be the reason for the Action variable's does not affect on Share (Ramadhani et al., 2020).
This study had several limitations. As conducted via an online survey, the researchers could not confirm the validity of the respondents' answers. Also, due to limited time, the researchers could only cover small portions of hospitals. However, this study could hopefully provide an information about the use of AISAS model in telemedicine advertisement.

Conclusions
This study aims to analyze the effect of the use of AISAS model in social media toward purchase intention. Social Media has a statistically significant effect on the components of the AISAS model, namely the Attention, Interest, and Search components. The Attention, Interest, and Search components of the AISAS model have no significant impact on consumer purchase intention. Purchase intention signifficantly affect Action in AISAS model. The AISAS model's Search component, is not significantly affect by the AISAS Action model component. In general, it may be said that employing social media as a promotional tool for telemedicine advertising does not improve consumer interest in purchasing. The issue of attention, interest, and search don't have a significane effect toward purhcase intention is an intriguing one which could be usefully explored in further research.