The Influence of Instafamous and Content Marketing on General Attitude Toward Local Brands and Interest in Buying Cosmetic Products

. The growth of local cosmetics in Indonesia has increased rapidly due to the increasing public interest in local cosmetic products that provide products according to the skin colour of Indonesians. Local cosmetic companies must compete with fellow local brands and brands from other countries to do marketing effectively. One is using Instagram social media to collaborate with Instafamous or create Instagram content. This study aims to determine the influence of Instafamous and Content Marketing on Interest in buying local cosmetic products mediated by General Attitude Toward Local Brands (GATLB). This type of quantitative research uses non-probability sampling methods and purposive sampling techniques on 100 respondents. Data was collected with online surveys and analyzed with SEM PLS through the SmartPLS 3.9.9 application. The results of this study show that Instafamous and Content Marketing hurt buying interest. While GATLB as a mediating variable plays a full role or complete mediation, Instafamous variables and Content marketing do not directly influence buying stake but have an indirect influence through GATLB as a mediation variable.


INTRODUCTION
The era of globalization requires companies to be more creative in marketing their products.This era combines automation technology, namely the Internet of Things (IoT), cloud computing and cognitive computing [1].Social media is one of the most effective forms of marketing because it can quickly provide awareness of a brand and be passed on from one user to another [2].One social media platform often used is Instagram, an application-based mobile phone platform that can share photos and videos with artistic visuals and provide social networking services [3].
Cosmetic companies also use social media to promote and compete in the industrial world [4].Given the rapid growth of the cosmetic industry, data from the Ministry of Industry in 2020 showed that the local cosmetic industry grew 9.32% from the previous year.Many new local brands have emerged in Indonesia that have only been established less than ten years ago.
Although local brands show positive growth, people tend to have inferior perceptions of local brands [5].In addition, according to research [6], young consumers in Indonesia crave foreign brands due to their globalization orientation, so they have favourable emotions towards foreign brands.This makes local cosmetic companies have to be more active in marketing their products, one of which is by collaborating using endorsement methods, namely including messages from brands and/or creations in their uploads by giving rewards or payments to Instafamous [7,8].
Instafamous can influence consumer attitudes towards a brand.An Instafamous with good credibility can become an opinion leader who can influence the public's mindset [9].In addition, Instafamous can also affect consumers' buying interest in cosmetic products with their ability to connect with their followers [10].
In addition to the influence of Instafamous, content marketing also has a vital role in digital marketing.Content marketing is creating a lot of content and good content and, most importantly, knowing what the audience wants [11].Content marketing can affect brand attitudes, where consumers form mindsets because of the encouragement of product information provided by brands.Positive attitudes from consumers can affect their interest in buying these products [12].
Research conducted by [13] shows a significant relationship between Instafamous and brand attitude, where expertise and trustworthiness influence the formation of brand attitude and brand image.Other studies conducted [14,15,16] show that there is a significant direct relationship between Instafamous and buying interest, but research conducted [15] shows that Instafamous credibility does not affect buying interest directly but affects consumers' positive attitudes (Brand attitude).

METHODS
This study uses a quantitative approach using non-probability sampling and purposive sampling techniques, where it is limited to certain types of people who have the following criteria: 1. Women who have never used local cosmetic products (BLP, SASC, ESQA, Goban, Rollover Reaction).
2. Know and have seen the promotion of local cosmetic products carried out by Instafamous, which has more than 1 million followers (Tasya Farasya, Rachel Goddard, Abel Cantika, Nanda Arsyinta, Indira Kalista).In research using SEM (Structural Equation Modelling) one method of determining the sample size for SEM is multiplying 5-10 times the num-ber of indicators of the entire variable.This study has 17 question items, so the sample size required is at least 16x5 = 85 to 17x10 = 170 respondents.So, the researcher, in this case, used 100 female respondents from throughout Indonesia.The sample of Instafamous is limited to 5 Instafamous with more than 1 million followers, namely (Tasya Farasya, Rachel Goddard, Abel Cantika, Nanda Arsyinta, and Indira Kalista) and local cosmetic brands that have been established for less than the last 10 years (Rollover Reaction, Goban, Sasc, BLP, Esqa).The selection of local brands looks at the age of local brands that have not reached 10 years and are actively creating content on Instagram.
The data collection technique in this study was questionnaires with a Likert scale of 1-7, each of which had the respondents' opinions.Data was collected through an online survey with questionnaires in Google Forms and disseminated through social media networks WhatsApp and Instagram Groups.It was cut off after reaching 100 respondents.
The questionnaire consists of 2 parts.The first part contains the respondent's identity consisting of full name, area of domicile, gender, occupation, age, and education.At the same time, the second part contains statement items, namely 7 statement items for Instafamous variables, 6 for Content Marketing variables, 6 for GATLB variables and 4 for purchase interest variables.The validity test was conducted with SPSS program version 28.0.1.1 with a sample of 50 outside the research respondents.Test the hypothesis using SEM-PLS Analysis using the Smart PLS application version 3.2.9, which consists of two submodels, namely the measurement model or the so-called outer model and the structural model, often called the inner model.
Based on the hypothesis formulation above, a research conceptual framework can be assembled as follows.

RESULTS AND DISCUSSION
Characteristics of Respondent.The characteristics of respondents in this study were as many as 100 samples.The majority of respondents have an S1/bachelor degree background of 73%.Most respondents are young, namely 20-30 years old, with 99 people (99%).
Measurement Model / Outer Model.Measurement model or outer model was carried out on all variables involved in this study, namely independent variables or exogenous variables, namely the quality of Instagram content against dependent or endogenous variables.The outer model is measured by convergent validity, discriminant validity, and internal consistency reliability.This model is called a reflective measurement model because it evaluates the values of loading factor, composite reliability, Cronbach's alpha, AVE, and discriminant validity.The rule of thumb for the loading factor commonly used to measure convergent validity is ≥ 0.7 with an average variance extracted (AVE) value of ≥ 0.5 [18].If an outer loading value below 0.7 is found, the researcher should carefully examine the effect of item deletion on composite reliability and construct content validity.
Generally, indicators with outer loading between 0.40 and 0.70 should be considered to be removed from the scale only when removing these indicators leads to increased composite reliability [18].The following are the results of the outer loading analysis phase II (re-estimate) after eliminating one by one the indicators whose external loading value is below 0.7 as follows.The next test is the estimation of path coefficients, which is carried out to see the significance of the influence of the independent variable on the dependent variable by looking at the value of the parameter coefficient and the significance value of t-statistics.This test is carried out by calculating using PLS Bootstrapping according to the model below (Figure 2).
It can be seen from the statistical t-value between the independent and dependent variables to assess the significance of predictive models in structural model testing.According to [18], the significance value of level 5% must meet the tvalue of 1.96.The t-statistic value can be seen in the Path Coefficient table in the Smart PLS output according to Table 2.  Evaluating The Effect of Mediation.Testing of mediating effects in analysis using PLS first, testing the influence of exogenous variables on endogenous variables and significance if the tstatistic value >1.96 (Table 3).In subsequent tests, it looked at the effect of mediating variables on endogenous variables significant at t-statistics >1.96 (Table 4).[13].This can affect the positive attitude of consumers because instafamous who are famous through Instagram for sharing information about cosmetic products are more trustworthy than celebrities who work with certain brands to advertise their products.The credibility of instafamous can strengthen brand credibility, which will lead to positive behaviour in the form of considerable buying interest [19].
In the instafamous variable, the Popularity variable has the highest value.Popularity is related to the number of fans Instafamous owns and how often Instafamous interacts with followers through Instagram.Instagram is not the same as celebrities who play a role in art and entertainment, so Instagram has a more intense time interacting with its followers.This makes instafamous have closer engagement with their followers and creates a better consumer attitude towards the instafamous they follow.
The Effect of Content Marketing on GATLB on Local Cosmetic Products.Content marketing positively and significantly influences GATLB's local cosmetic products directly.The results of descriptive research on GATLB variables have been excellent, and very high attitudes were perceived by respondents.This means that content marketing has a significant influence on the positive attitude of consumers.This is in line with research conducted by [20], which found that content marketing influences brand attitude, where good content marketing allows positive consumer attitudes towards brands to be more easily formed.
Content marketing affects the perception and image of the product through the information provided so that users can get an idea of the product even though they have never used it.If the information displayed is good, the better the consumer's attitude towards the product.
The Instagram social media platform itself has a feature to promote content from social media accounts to other users looking for similar accounts, making it easier for users to get information about products.Just as the user follows and likes one of the local cosmetic product accounts, the user will be recommended content similar to the account they want.Marketing content that gets a lot of likes and views on Instagram can become viral.The more viral the content, the more popular the product offered.
The Effect of GATLB on Buying Interest.The results of this study show that GATLB has a positive and significant influence on GATLB local cosmetic products directly.The results of descriptive research on GATLB variables have been excellent, and very high attitudes were perceived by respondents.This means that GATLB's attitude significantly influences consumers' buying interest in local cosmetic products.
Local brands of cosmetic products in Indonesia are more in demand because consumers feel the products are more suitable for their needs, and they have confidence that the products are made with halal ingredients indicated by halal labels.Consumers' trust in the brand gives them a good attitude and vice versa.The development of local cosmetic products in Indonesia and the improvement of the quality of local cosmetic products proves that local cosmetics get positive attention from the community, making them interested in buying local cosmetic products.
According to research [21], consumers in Indonesia show strong local brand consciousness and offer positive emotional value to create a positive consumer perception of local brands.So, the more positive the consumer's attitude towards the brand, the greater the influence on buying interest.This is in line with research [22], which shows that a positive attitude of consumers can encourage buying welfare.
The effect of GATLB as a mediating variable between Instafamous and Content Marketing on buying interest.This study shows an insignificant di-rect relationship between Instafamous and Content Marketing to purchase interest.Still, the indirect relationship between Instafamous and content marketing mediated by GATLB is significant, so GATLB fully mediates the relationship between Instafamous and content marketing to purchase interest.
Consumer attitudes towards brands can be used to predict consumer purchase behaviour to increase buying interest [23].According to consumer behaviour theory, stimuli can influence individuals through perception, i.e., how consumers perceive and attach meaning to all information input.In this case, Instafamous can act as an opinion leader who can provide information to influence consumer attitudes towards local brands.Instafamous can build a positive attitude of consumers compared to ordinary celebrities because they can interact better with their followers on social media.This makes the information shared by instafamous more informative and exciting to follow.Therefore, selecting appropriate instafamous is important for companies to increase consumers' positive attitudes towards local cosmetic products to encourage consumers' buying desires.
Content Marketing also plays a role in providing perceptions to consumers about a product.The better the marketing content used, the better consumer perception.Local cosmetic products also have value in their products, made by the nation's children, so that they can make better consumer attitudes towards these products and encourage consumer buying interest.This is in line with research conducted by [17], where the relationship between social media use (social media content, activist and brand content) and buying interest is mediated by consumer attitudes, where positive consumer attitudes can increase consumer purchase interest.

CONCLUSIONS
Positive consumer attitudes have an essential role in increasing consumer buying interest, so local cosmetic companies should choose the right instafamous to cause positive consumer attitudes towards local cosmetic brands to support purchase interest.Based on the results of this study, the variables with the lowest value are trustworthiness and expertise, so it needs to be the attention of local cosmetic companies to find instafamous experts in the cosmetics field who can pro-vide detailed and honest reviews of local cosmetic products.
Local cosmetic companies also need to consistently create content containing accurate and relevant information about local cosmetic products.
In addition, it is recommended that local cosmetic companies create marketing content about their products that can strengthen consumers' positive attitudes towards local products, such as using hashtags, invitations to use local cosmetic products, and emphasizing product quality that can compete with foreign brands and choices according to the needs of Indonesian consumers.
Theoretically, this study formulates a conceptual framework to predict consumer behaviour towards consumer buying interest in local cosmetic products.In particular, this study explores the causal relationship between the influence of instafamous on the interest in buying local cosmetic products, the effect of content marketing on the interest in buying local cosmetic products (response), GATLB as a mediating variable of this study can be known if Instafamous does not have a direct and not significant effect on buying interest as well as content marketing on purchase good.
Based on the results of this study, it was found that both Instafamous and Content Marketing on Instagram do not necessarily affect buying interest but must form a positive attitude of consumers.This is in line with the Theory of Planned Behavior, where consumer attitudes can predict consumer behaviour and internal and external factors that become stimuli in forming consumer attitudes.
This research also supports the Stimulus-Organism-Response theory, which is formed through stimuli in the form of the role of Instagram social media and Instagram content of local cosmetic products in providing product information to cause consumer perceptions and evaluations of regional cosmetic developments so that this research can be used to provide references and images to parties who need knowledge about the role of instafamous and content marketing in shaping responses to buying interest.

3 .
Know and have seen Content Marketing of local cosmetic products through local cosmetic brand Instagram accounts (BLP, SASC, ESQA, Goban, Rollover Reaction).

Figure 2 -
Figure 2 -Model of Path Coefficient with PLS Bootstrapping

Table 1 -
Correlation value (outer loading) and AVE variable has an R-Square of 0.363.This value shows that Instafamous and Content Marketing construct variables can explain the buying interest construct variable by 36.3%.In comparison, the other 63.7% is influenced by other factors not included in this study variable.

Table 4 -
Mediating effect on endogenous variables Based on the comparison between the path coefficients of exogenous and endogenous variables compared to the relationship between exogenous and endogenous variables through mediating variables, it can be concluded that GATLB fully mediates the influence of Instafamous and Con-tent Marketing on buying interest with a statistical t value of >1.96 and p-values of <0.05.The Effect of Instafamous on Buying Interest inLocal Cosmetic Products.The results of this study show that Instafamous does not significantly in-