THE EFFECT OF INSTAGRAM CONTENT TOWARDS INTENTION TO VISIT UC_IBMRC WITH ONLINE ENGAGEMENT AS MEDIATING VARIABLE

: Social media is an online media that is now very developed in the world of communication and business. Instagram is one of the social media that is widely used by people to get information or run a business. In this study, researchers wanted to see the effect of Instagram content on interest in visiting with online engagement as a mediating variable. This research was conducted quantitatively with structural equation modeling where the researcher had an initial research model. After testing the validity and reliability, the researcher has a final research model. The respondents in this study were 96 UC_IBMRC Instagram followers. The results of this study indicate that Instagram content has an effect on visiting interest. Instagram content affects online engagement. Online engagement has an effect on interest in visiting and at the same time becomes a variable that partially mediates the relationship between Instagram content and interest in visiting.


Introduction
The development of technology today is growing very rapid. In the current era of globalization, it is marked by the presence of various technologies, one of them is communication technology and information technology. In the era of globalization, information is colored by the development of the internet and various social media. According to (Panjaitan & Prasetya, Interest to Visit According to (Adinda and Pangestuti 2019) consumers in deciding to visit must have several considerations first. In the process of choosing, there is one aspect where prospective consumers can determine what the purpose of the choice in the minds of consumers is. This strong and motivating urge to choose an action is called interest. In addition, according to (Munawwaro, 2018) interest is defined as a source of motivation that encourages someone to do what they want to do if they are free to choose. When a person sees an advantage for them then they are interested and bring satisfaction. When satisfaction is reduced, interest also decreases. And according to (Edithania, 2018 in Trirahayu and Putri 2019) that visiting interest is assumed to be the same as consumer buying interest. So buying interest is defined as an individual's tendency to take action before the purchase decision occurs.

Type of Research, Population and Sample
The research method used in this study is a quantitative method. Quantitative research is research that is based on a systematic, planned, structured and uses numbers (Sugiyono, 2012). Bungin (2015) says that quantitative research has a focus on recording as much data as possible from a wide population through statistical formulas. According to Sarwono and Martadiredja (2017), quantitative research can use either probability sampling or nonprobability sampling. This research uses non-probability purposive sampling method. Abdillah and Jogiyanto (2015) state that the population is the entire object of research or referred to as the universe. While Bungin (2015) mentions the population as a group of objects that will be targeted in research and also the entirety of the research that will be used as a source of research data. The population of this study is 2315 followers from Instagram UC_IBMRC. According to Bungin (2015) the sample is a representative of the population that can be represented as a subject in the study. The researcher uses the slovin formula to determine the number of samples n= (N/N*e2+1) n = (2315/2315*0.12+1) n = 95.66 n = 96 Based on the existing slovin formula, the researcher must find a minimum of 96 respondents with an error level of 10%.

Mediation Effect Test
According to Ghozali and Latan (2015), the mediation effect test is used to see the relationship from the independent variable to the dependent variable through an intermediary variable. There are three possibilities in the mediating effect test, namely: 1. Partial mediation effect, is when the hypothesis test from variable X to Y is accepted and the hypothesis test from M to Y is also accepted 2. Full mediation effect, is when the hypothesis test from variable X to Y is rejected and the hypothesis test from M to Y is accepted 3. There is no mediating effect, that is, when the hypothesis test from variable X to Y is accepted and the hypothesis test from M to Y is rejected

Result
In this study, researchers used three variables, namely Instagram content (X), online engagement (M) and visiting interest (Y). For the Instagram content variable, the researcher uses six indicators. For the online engagement variable, the researcher uses six indicators. For the variable of interest in visiting, the researchers used three indicators. Based on the data analysis that has been done, the following is the mean and standard deviation for each indicator. Based on the results of data analysis, it can be seen that for the Instagram content variable, the highest mean is in the X2 indicator, for the online engagement variable, the highest mean is in the M3 indicator and for the visiting interest variable, the highest indicator is Y1. For the standard deviation of the Instagram content variable, the X2 indicator has the most homogeneous answer. For the standard deviation of the online engagement variable, the M3 indicator has the most homogeneous answer. For the variable of interest in visiting, the Y1 indicator has the most homogeneous answer.

Model Evaluation
In this study, the first step the researcher did was to evaluate the change in the model. According to Ghozali and Latan (2015), the evaluation of model changes is used to see whether the research model, along with the indicators used as measuring tools for each variable, is appropriate or not. Figure 1 is the initial model of this research Picture 1. First Model Source: Processed Data (2021) Researchers used the model according to Figure 1 for this study. There are six indicators for the Instagram content variable, six indicators for the online engagement variable, and three indicators for the visiting interest variable. For the first stage, the researcher tested the validity of the loading factor to see the value of each indicator.  (2021) Based on Figure 2, there is a slight change where the online engagement variable which was originally six indicators became four indicators. The variable for Instagram content is fixed with six indicators and the variable for visiting interest is fixed with three indicators.

Validity and Reliability Test
After evaluating the model changes using the loading factor, according to Abdillah and Jogiyanto (2015), the next step is to test the validity of the AVE, cross loading and test the reliability of Cronbach alpha and composite reliability.  (2021) Based on Table 3, it can be seen that all tests passed the minimum recommended value, where the AVE was at least 0.50. For the reliability test of Cronbach alpha and composite reliability, the recommended minimum value is 0.70. Based on table 4, for the cross loading validity test, the test carried out also meets the minimum requirements, namely the indicator on the corresponding variable has a value above 0.70 and is the largest compared to indicators on other variables.

Coefficient of Determination Test and Hypothesis Testing
According to Ghozali and Latan (2015), the coefficient of determination test is used to see how much influence the variables used in the study have.  (2021) Based on Table 5, it can be seen that the coefficient of determination for the online engagement variable is 0.116. This means that the Instagram content variable has an effect of 11.6% on online engagement and the rest is influenced by other variables. While the variable of interest in visiting has a coefficient of determination of 0.328. This means that the variables of Instagram content and online engagement have an effect of 32.8% and the rest is influenced by other variables. Based on Table 6, it can be seen that Instagram content has an effect on visiting interest. Instagram content affects online engagement. Online engagement has an effect on interest in visiting as well as a variable that partially mediates the relationship between Instagram content and interest in visiting accepted visits. The t-statistic value is 2.641 and the p-value is 0.009. This is in accordance with research written by Sagala and Rachmawati (2016) which states that Instagram content is the key to someone's interest in being able to visit someone's Instagram profile page. Rahayu and Baridwan (2020) also mention that Instagram content that is made attractive both in terms of photos, feeds design and also attractive invitations can greatly affect the affection for visiting someone's Instagram profile page. So far, UC_IBMRC's Instagram has always prioritized design so that followers are interested in visiting the profile. In addition, researchers and colleagues always change the design of feeds regularly so that followers don't get bored. Researchers and colleagues also use good and interesting captions and adjust the language of young people so that the majority of followers aged 17-25 years are more interested. This is in accordance with research conducted by Rosdiana (2019) which states that captions on Instagram play an important role in interest in visiting profile pages. Researchers and colleagues also try to always look for the latest materials in order to adapt to trends.

Influence of Instagram Content on Online Engagement
The second hypothesis in this study is that Instagram content has an effect on online engagement. The t-statistic value is 4.307 and the p-value is 0.000. Based on the loading factor test, it turns out that the likes and follow indicators are not the right measuring tools to measure online engagement. This is in accordance with research conducted by Rachmah and Mayangsari (2020) which stated in one of their findings that likes and follows are not appropriate for online engagement because there is no direct interaction between the two parties. In addition, Rachmah and Mayangsari (2020) also mention that even though Instagram account users do not like or follow activities, because even though they do not carry out these two activities, Instagram users can still have intense interactions with other users. Instagram UC_IBMRC quite often provides content that can increase online engagement from followers. The types of content used by researchers and colleagues are posteducation, sharing of outstanding students, lecturers or alumni and light content that can be shared with friends for discussion or jokes in their spare time. Researchers also recap daily Instagram insights so that they know in detail the online engagement made by followers on the content presented. In addition, from this Instagram insight, researchers can see which types of content are liked and disliked by followers.

The Effect of Online Engagement on Visiting Interest
The third hypothesis contained in this study is that online engagement has an effect on interest in visiting and is accepted. The t-statistic value is 5.824 and the p-value is 0.000. In addition, online engagement is a variable that partially mediates the relationship between Instagram content and interest in visiting. This is in line with research conducted by Al Khasawneh et al (2021) which states that online engagement has a positive effect on visiting interest. In a study conducted by Khasawneh et al (2021) it was also stated that the higher Instagram users interact such as giving likes, comments, shares, saves, the higher the level of compatibility of the Instagram users with the content shared and the interest in visiting will also be higher. Researchers conduct periodic evaluations using Instagram analytics in each uploaded post to see what types of posts are liked by followers and also what kind of posts are liked by followers. Through Instagram analytics, researchers and colleagues also see the development of the number of likes, comments, shares, comments, saves for each post every day. This is in accordance with research conducted by MacDowall and De Souza (2018) which states that in Instagram posts it is important to maintain interest and periodic evaluation through tools that have been provided in the form of Instagram analytics.