THE IMPACT OF USING INFLUENCER ON CONSUMER PURCHASE INTENTION WITH ATTITUDE TOWARDS INFLUENCER AND BRAND ATTITUDE AS MEDIATOR

This study is conducted to determine the effect of perceived credibility, trust, experience, and attractiveness as an evaluator of influencers’ effectiveness on consumer purchase intention in fashion products mediated by attitude towards influencers and brand attitude. Influencers are those who regularly share content related to certain fields or skills on social media platforms. This study is based on the Theory of Planned Behavior (TPB) and other previous studies related to influencers’ effectiveness on purchase intention of fashion products. The population in this study is 385 women in Surabaya with a nonprobability sampling method using a purposive sampling technique. The data analysis method used in this study is Structural Equation Modeling-Partial Least Square (SEM PLS) with a smart PLS 3.0 program. The analysis results in this study show that perceived credibility, trust, and attractiveness have a positive and significant effect on attitude towards influencers. In contrast, perceived expertise has a negative and insignificant impact on attitude towards influencers. Attitude towards influencers fully mediates the influence of perceived credibility, trust, and attractiveness on brand attitude and purchase intention. Brand attitude fully mediates the influence of attitude towards influencers on purchase intention. In this study, variable perceived expertise used did not affect attitude towards influencer variables. This is different from previous studies that have been done before. For further researchers, it is suggested that they can again use perceived expertise and other variables that can support more deep assessments that give rise to consumer attitudes towards an advertisement made by fashion product influencers in subsequent studies using different populations. The analysis model that has been used in this study can then be reused in conducting further research with varying numbers of populations and types.


Figure 1. Growth of The Internet and Social Media Users in Indonesia
nesses in the fashion industry from year to year makes the existing competition increasingly tighter, requiring companies to have high value and continue to improve management performance in terms of planning, organizing, and controlling resources owned (Ismanto, 2016). To convey product value and increase its existence, the company must adjust its marketing strategy to existing market conditions. In this digital era, businesses are required to be able to adapt to digitalization.
According to Kumar and Patra (2017), promotion is an action that can be done to disseminate information to increase sales capacity. With digitalization, one of the efforts that can be done in conducting promotional activities is using social media for promotional means. Through social media, businesses can take advantage of larger platforms to conduct business activities and reach consumers by the opinions expressed by Stelzner (2012); Purwana et al. (2017). Figure 1 shows that active social media users in 2019 increased by 44 million users from active users in 2017. That is in line with the results of bps processed data in 2019, where fashion products ranked second most sold through social media as a means of e-commerce transactions (Badan Pusat Statistik, 2019). Going back to the years before the digital age emerged, many businesses used celebrity fame to help with the marketing of their products. Influencers are individuals who regularly share content related to a particular field of expertise on social media platforms. Influencers can also be one of the marketing strategies used where familiar individuals on social media use their connections in social media to promote certain products or services (Childers et al., 2019;Nijs, 2019). According to the research conducted by Chetioui et al. (2020), an influencer marketing strategy can affect attitude towards influencers, brand attitude, and purchase intention owned by consumers. Based on research that has been done by Chetioui et al. (2020) previously, there are several variables used to evaluate influencers of fashion products, namely perceived credibility, trust, perceived behavioral control, subjective norms, perceived expertise, and perceived congruence. The variables mentioned are those that affect the attitude or view of consumers towards the influencer. In addition to the variables used as evaluations in the study, there is a term attractiveness used in research of Taillon et al. (2020); Osei-Frimpong et al. (2019); Wiedmann and von Mettenheim (2020), Weismueller et al. (2020), which can also be used as a gauge of influencer effectiveness. Attractiveness, according to these studies, refers to the physical condition of an influencer. According to, consumer buying interest in this generation can be strongly influenced by the presence of influencers because they spend most of their time accessing social media. DECEMBER 2021 Dewi Mustikasari Immanuel, Alexandria Bianda H. S. Source: Nezakati et al. (2015)

Figure 2. Model Analysis Theory of Planned Behavior
The theory that underlies this research is the theory of planned behavior owned by Ajzen, which was initiated in 1991 in Ajzen (2011). This theory is the theory of the development of the Theory of Reasoned Action (TRA). TRA is a theory that examines the dimensions of attitudes and subjec-tive norms and their influence on intention towards behavior. In contrast, TPB is a theory of TRA development that focuses on the dimension of attitude, subjective norms, with the addition of one variable that is perceived behavioral control and its influence on intention towards behavior. Figure 2 shows an additional dimension of perceived behavioral control, which is considered necessary in TRA development. The selection of TPB as the basis of this study is through the dimension of attitude towards behavior. Researchers can assess the variables used in this study where the dimension of attitude towards influencers refers to the individual's assessment of certain behaviors. However, in some previous studies, it has been found that credibility has no positive effect on consumer purchase intentions (Lim, Mohd Radzol, Cheah, and Wong, 2017;Wibowo and Heryjanto, 2020). Similarly, related trusts were found to have no positive effect on purchase intentions (Lina and Permatasari, 2020). It is contrary to the research of Chetioui et al. (2020), which says that perceived credibility and trust have a positive effect on consumer attitudes towards influencers that ultimately positively affect purchase intention. The existence of this gap makes this interesting to be re-examined. Therefore, this research aims to analyze the influence of influencer effectiveness on purchase intentions owned by consumers by mediated by attitude towards influencer and brand attitude where credibility and trust are included. And it is expected to be helpful to know the impact of the implementation of marketing strategies by using influencers for businesses to increase consumer buying interest.

HYPOTHESIS DEVELOPMENT
The Perceived Credibility and Attitude Towards Influencer. Credibility is the integrity and prowess that advertisers have in determining the effectiveness of advertising (Martins, Costa, Oliveira, Gonçalves, and Branco, 2019;Yang, Kim, and Yoo, 2013) and widely used to analyze the effectiveness of marketing activity, especially endorsement (Lim et al., 2017). So perceived credibility is essential before consumers choose or follow an influencer (Nam and Dân, 2018). Perceived credibility positively affects attitude towards influencers because consumers trust the message conveyed by an influencer if the consumer considers the influencer has credibility (Chetioui et al., 2020;Kalu, 2019). However, the results of the study conducted by (Lim et al., 2017;Wibowo and Heryjanto, 2020) said that credibility does not affect attitude towards influencers. This is because consumers feel that influencers do not always have adequate knowledge about the products offered. An indicator of per-The Impact of Using Influencer on Consumer Purchase Intention with ... ceived credibility is an influencer must meet the criteria of convincing, believable, credible, and a good reference (Martins, Costa, Oliveira, Gonçalves, and Branco, 2019;Yang, Kim, and Yoo, 2013). H1: Perceived Credibility has a positive and significant effect on Attitude Towards Influencers.
Trust and Attitude Towards Influencer. Trust refers to the honesty, integrity, and trustworthiness of advertisers who depend on the perception of the target audience or consumers (Chetioui et al., 2020;Ohanian, 1990). Trust can help create a business's long-term relationship with consumers (Chopra, Avhad, and Jaju, 2020). Based on several research results, trust positively affects attitude towards influencers (Chetioui et al., 2020;Chopra, Avhad, and Jaju, 2020;Macintosh and Lockshin, 1997;Ohanian, 1990;Suh and Han, 2002). Trust is an aspect used in considering advertising assessments conducted by influencers in answering whether they assess products objectively or due to financial rewards (Wiedmann and von Mettenheim, 2020). Highly trusted influencers have more opportunities to influence attitudes and choices to consumer buying interests (Chetioui et al., 2020). However, the results of research conducted by that trust do not positively impact attitude towards influencers. There are five indicators used to measure trust: dependable, honest, reliable, sincere, and trustworthy (Najib, Daud, and Rosa, 2019;Ohanian, 1990). H2: Trust has a positive and significant effect on Attitude Towards Influencers.
The Perceived Expertise and Attitude Towards Influencer. Expertise is the level of knowledge, experience, or competence the endorser possesses (Chetioui et al., 2020;Ohanian, 1990) and can provide more per suasion to consumers (Martensen, Brockenhuus-Schack and Zahid, 2018). Chetioui et al. (2020); Wiedmann and von Mettenheim (2020), in their research, revealed the results that perceived expertise from influencers positively affect attitude towards influencers. And this is due to consumer assessment on influencers who experts are more highly qualified and make the assessment of a product more accurate and valid. The indicators used to measure the expertise owned by an influencer are expert, experienced, knowledgeable, qualified, and skilled. (Chetioui et al., 2020;Ohanian, 1990). H3: Perceived Expertise has a positive and significant effect on Attitude Towards Influencers.
Attractiveness and Attitude Towards Influencer. According to Ohanian (1990;Rizqiah et al., 2020), attractiveness refers to the characteristics possessed by an endorser as measured in attractive physical appearance, traits, and lifestyle. Some previous research has stated that attractiveness positively affects attitude towards influencers (Kalu, 2019;Lim et al., 2017;Osei-Frimpong et al., 2019;Taillon et al., 2020;Weismueller et al., 2020;Wiedmann and von Mettenheim, 2020). Attractiveness owned by an influencer can encourage consumers to receive advertising products done by influencers more easily. Influencers who are considered to have a high level of attractiveness for consumers will further influence the attitude towards consumer influencers (Taillon et al., 2020). The indicator used to measure attractiveness are attractive, classy, beautiful, elegant, and sexy (Najib et al., 2019;Ohanian, 1990). H4: Attractiveness has a positive and significant effect on Attitude Towards Influencers.
Attitude towards Influencer, Brand Attitude, and Purchase Intention. Attitude can be interpreted as a consumer attitude that reflects the power and intention to interact with an account in the future (Casaló, Flavián, and Ibáñez-Sánchez, 2020;Chetioui et al., 2020). Chetioui et al. (2020); Huang, Wu, Wang, and Boulanger (2011) said that attitude towards influencers could form a positive brand attitude where the brand can be a priority of consumers who ultimately have a positive effect on purchase intention. It is supported by research conducted by Amos, Holmes, and Strutton (2008); Taillon et al. (2020) that attitude towards influencers positively affects brand attitude. In the world of online marketing, the message conveyed by an influencer will affect the attitude of consumers who also affects the brand attitude of a brand (Eshghi, Sarkar, and Sarkar, 2017). According to Casaló et al. (2020); Chetioui et al. (2020), indicators used to measure attitude towards influencers are: an influencer can be a role model for the community, an influencer can provide information about the latest fashion products, an influencer can provide information about unique fashion products, an influencer can display exciting content about the fashion industry, and influencer can be a reliable source of information. H5: Attitude Towards Influencer has a positive and significant effect on Brand Attitude. H7: Attitude Towards Influencer has a positive and significant effect on Purchase Intention.

Brand Attitude and Purchase Intention.
Brand attitude is an overall assessment and reflection of consumer response regarding a brand (Susanti, 2017). Liébana-Cabanillas, Japutra, Molinillo, Singh, and Sinha (2020), in his research, stated that marketing strategies through an influencer are considered to be able to build a brand attitude of a brand to influence the views of prepurchase consumers. Chetioui et al. (2020), in their research, revealed the results that brand attitude positively affects purchase intention. More about Chetioui et al. (2020); Huang et al. (2011) said that a positive brand attitude affects the consumer's priorities towards the brand and positively affects purchase intention. This is because increasing consumer assessment of a brand will improve brand attitude and encourage consumer purchase intention.
Indicators used to measure brand attitude that is a brand should be easy to remember, like, and be the priority of consumer choice (Susanti, 2017;Till and Baack, 2005). H6: Brand Attitude has a positive and significant effect on Purchase Intention.

Purchase Intention and Attitude towards Influencer.
Purchase intention is a mental statement that exists in consumers that describes the plan consumers own in purchasing products with a certain number and brand (Rachmy, 2018) and can arise due to the stimulation given to consumers of the company (Mubarak and Sanawiri, 2018). Chetioui et al. (2020); Cooke and Sheeran (2004) said that the relationship between attitude towards influencers and purchase intention would increase involvement with advertising conducted by influencers. The research shows that attitude towards influencers positively influences purchase intention (Chetioui et al., 2020;Taillon et al., 2020). This is because if consumers like, assume, use information from influencers, and assume consumers influence product selection, it will increase purchase intention. MacKenzie, Lutz, and Belch (1986); Taillon et al. (2020) said that attitudes influence behavior formation. So it can be concluded that if the attitude towards influencers is positive, it will produce behavior in the form of purchase intentions carried out by consumers. According to (Rachmy et al., 2018), there are four indicators in purchase intention, namely (1) explorative interests that refer to consumer behavior to always look for information on the product of interest, (2) preferential interests that refer to consumer behavior in prioritizing a product, (3) referential interests that refer to consumer behavior to provide product references to others and (4) transactional interests that refer to consumer's implement to buy a particular product. H7 = Attitude Towards Influencer has a positive and significant effect on Purchase Intention. The following are conceptual models in this study ( Figure 3).

METHOD
Determination of the number of samples in this study is based on an unknown population or called infinite population so that the size of the sample is 385 respondents (Davis and Cosenza, 1993). And based on specific criteria (purposive side) of the women's society in Surabaya with the age range of 19 to 34 years, have income, and have seen and followed influencers of a product on social media.
The data analysis in this study uses the Structural Equation Model-Partial Least Square (SEM-PLS), which is supported by the use of SmartPLS software. The first stage is an outer model measurement that is tested from a validity test, which can be used to measure whether the study results can be accepted, of course, with certain criteria.
The Impact of Using Influencer on Consumer Purchase Intention with ... Abdillah and Jogiyanto (2015) said that the validity of the construct is used to show how well the results are obtained through measurements following the theory that refers to a construct and reliability test used to measure the internal consistency of the measuring instrument. The construct validity test consists of a convergent validity that refers to a principle that the gauge contained in a construct should be highly correlated based on loading factor, with the rule of thumb used as a reference for the validity of convergent, i.e., outer loading > 0.7 and discriminant validity refers to the principle that the gauge contained in a different construct should not be highly correlated can be said to be valid if the cross-loading value > 0.7 in one variable. While the reliability test is measured using two methods, namely Cronbach's alpha and Composite reliability (Abdillah and Jogiyanto, 2015). Abdillah and Jogiyanto (2015), rule of thumb value Cronbach's alpha or composite reliability is required greater than 0.7 to be received even though the value of 0.6 is still acceptable.

Figure 3. Conceptual Models
The second stage is the measurement of the Inner Model or hypothetical test results which are conducted with two tests, namely R-square and test t. According to Abdillah and Jogiyanto (2015), it is used to measure the degree of variation of changes obtained from independent variables to dependent variables. The benchmark number used is 0.67 is considered high. 0.33 is considered moderate and considered weak if it is at the value of 0.19. At the same time, the t-test or path coefficient is used to indicate the level of significance in hypothesis testing. The number of path coefficients that must be shown in the T-statistic value for the two-tailed hypothesis should be above 1.96.
The last stage is a mediation effect test conducted to determine the relationship between independent and dependent variables through connecting variables or so-called mediation variables. The measurement is based on the the-statistic value contained in the Total Effect  (2015), mediation effect testing is conducted when the independent variable relationship to dependent variables is significant.

RESULTS
Outer Model Measurement. The Average Variance Extracted (AVE) value owned by each variable has a reference value where each variable has a construct validity that strongly correlates with the indicators or statements used in this study. The rule of thumb used to measure the validity of the construct is the AVE value > 0.5, so it can be said that the correlation indicator used in each variable used in this study is valid.
Based on the rule of thumb used as a reference for the validity of convergents, the value of outer loading (loading factor) must be >0.7. Thus, variables with loading factor values below 0.7 are declared invalid because they do not meet the reference value. Based on the loading factor value results, two indicators do not meet the reference value, namely one indicator on the variable Perceived Credibility (PC1) with a value of 0.651 and attractiveness variable (A1) with a value of 0.681 so that both indicators are declared invalid. The following is the result of the loading factor value after elimination of invalid indicators: In the discriminant validity test, each indicator must have a high cross-loading value on its latent variable compared to the cross-loading value of the indicator on other latent variables with the reference value of cross loading should be > 0.7. Therefore, elimination should be done on the PC1 indicator with a value of 0.651, and A1 with a value of 0.681 does not meet the reference value and is declared invalid.   There are two methods or values that can be used in reliability tests, namely Cronbach's Alpha and Composite Reliability, where the rule of thumb value of both must be > 0.7. All variables contained in this study have a value of Cronbach's Alpha and Composite Reliability > 0.7, so it can be said that all indicators or measuring instruments in this study are accurate and precise.

Inner Model Measurement. The Outer
Loading value is used to show the data processing results of each statement or indicator of each variable used in this study in the form of mean values and standard deviations. The greater the mean value indicates that the respondent approves of the indicator, and the smaller the standard deviation value means that the indicator is rated as a good indicator. While the lowest mean value means that the respondent disapproves of the indicator. The greater the standard value deviation indicates that the indicator is less good at reflecting the variable.
In the hypothesis test, the value of used to measure the rate of change of independent variables against dependent variables. Based on value , Attitude Towards Influencer and Purchase Intention variables have a value of moderate, while the Brand Attitude variable is below the moderate value point close to weak with a value of 0.210 where the weak when it stands at 0.19.
There are several values used to measure and demonstrate the degree of significance in conducting hypothesis testing. One of the values used to measure the importance is the T-Statistic value which for the two-tailed hypothesis value must be above 1.96 to be considered significant. Based on Table 4.10, most influences between variables have had a T-Statistic value above 1.96, so it is considered significant. However, the variable perceived expertise does not significantly influence the attitude variable towards influencers with a T-Statistic value of 1.91 below 1.96.

T-Statistics
This study has two mediation variables, namely attitude towards influencer (AT) and brand attitude (BA) variables. The majority of statistical t values have exceeded the value of 1.96. The influence of attitude towards influencer variable on brand attitude has a statistical t value of 10.94075; then, the variable is considered to mediate fully. The effect of the brand attitude variable on purchase intention has a statistical t value of 13.54631; the variable is considered to mediate fully. The influence of attitude towards influencer on purchase intention has a statistical t value of 13.57612; the variable is considered full mediating. So it can be concluded that the two mediation variables contained in this study are attitude towards influencer variables and brand attitude variables are deemed to be fully mediated.
The Impact of Using Influencer on Consumer Purchase Intention with ...

DISCUSSION
The Perceived Credibility dan Attitude Towards Influencer. This study found that perceived credibility (PC) positively and significantly affects attitude towards influencers (AT). This is supported by previous research conducted by Goldsmith, Lafferty, and Newell (2000); Lim et al. (2017) that influencers with high credibility will be more influential and positively impact consumer perception of the influencer. In addition, consumers will also trust an influencer to buy a product if the consumer considers an influencer to have good credibility (Kalu, 2019;Wathen and Burkell, 2002). An influencer's credibility is vital for prospective consumers because if the influencer has good credibility, then the influencer's assessment of the advertised product is not only done for the sake of content alone. In addition, influencers with good credibility will further influence the positive attitude of prospective consumers towards a brand, thus encouraging the emergence of purchase interest. However, this study was found to differ from previous research conducted by (Lim et al., 2017;Wibowo and Heryjanto, 2020), who explain that credibility has no significant effect on attitude towards influencers and purchase intention. So, this previous research suggested the opposite is the need for influencer services because it does not affect the interest in buying.

Trust dan Attitude Towards Influencer.
Trust (T) positively and significantly affects the variable attitude towards influencers (AT). This is supported by previous research that says that trust has a significant influence on attitude towards influencers caused if consumers can trust influencers in marketing products, it will help create long-term relationships and increase consumer buying interest in the brand of fashion products advertised by the influencer (Chetioui et al., 2020). In addition, an influencer must be trustworthy because the promotional content done by influencers impacts a wide audience, especially the influencer's followers. Therefore, trusted influencers have positive values to improve their attitude towards influencers. The results of this study are different from the results shown in the study; according to Lina and Permatasari (2020), trust does not significantly influence attitude towards influencers that can affect purchase intention.
The Perceived Expertise and Attitude Towards Influencer. Perceived expertise (PE) negatively and insignificantly affects variable attitude towards influencer (AT). This is not in line with previous research conducted by Chetioui et al. (2020) and Wiedmann and Mettenheim (2020), where perceived expertise has a significant and positive influence on attitude towards influencers because influencers who are considered experts in their fields have a more accurate and valid assessment of a product. This study found that the expertise or expertise of an influencer who suits his area of expertise does not affect consumer perception in assessing an influencer when advertising a product. This can be because each prospective customer has different product and brand feature preferences and preferences towards an influencer, so they do not have to be fixated on their expertise.
Attractiveness dan Attitude Towards Influencer. Attractiveness (A) has a positive and significant effect on variable attitudes towards influencers (AT). This is supported by previous research that says attractiveness has a positive and significant influence on attitude towards influencers (Lim et al.,2017;Kalu, 2019;Osei-Frimpong et al., 2019;Wiedmann and von Mettenheim, 2020;Weismueller et al., 2020;Taillon et al., 2020;Lina and Permatasari, 2020). Because if influencers have their attractiveness, it will cause the products advertised by influencers to be attractive in the eyes of prospective consumers. It influences the products advertised by influencers. Attractiveness refers to the characteristics of attracting an influencer who can influence consumers in the process of receiving ads that are done more positively about the ad (Rizqiah, Kusdibyo, and Rafdinal, 2020). To improve consumers' positive attitudes regarding an influencer, companies are advised to choose influencers with great persuasive abilities that will impact consumer buying interests (Osei-Frimpong et al., 2019). DECEMBER 2021 Dewi Mustikasari Immanuel, Alexandria Bianda H. S.

Attitude Towards Influencer dan Brand
Attitude. Attitude towards influencer (AT) positively and significantly affects the variable brand attitude (BA). This is supported by research that has been done previously by Chetioui et al. (2020) that attitude towards influencers positively affects brand attitude. This is because consumers tend to use a product to provide identity for themselves. The product advertised by influencers is considered to have a higher value than other products, so it also affects the attitude owned by the brand in consumer perception or brand attitude. Following the majority of respondents characteristics in this study who have an age in the range of 19 to 24 years and referred to as generation Z., This generation has the attributes of technologically fluent which tends to access the information needed easily and quickly compared to other generations (Ulfa, 2018). Therefore, these characteristics support a more in-depth consumer assessment of a product advertised by influencers because their ability to dig information has been better than other generations. So it affects the improvement of brand attitude in consumer perception through the quality and quantity of information they have.

Brand Attitude dan Purchase Intention.
Brand attitude (BA) positively and significantly affects the variable purchase intention (PI). Hasil is supported by research that has been done previously by Chetioui et al. (2020) that brand attitude has a positive and significant effect on purchase intention. According to Balathandayutham (2020), the research strategy of improving brand attitude can increase consumer involvement in purchasing a product. Generation Z's fluency in gathering information can influence their assessment of a product to define a product's brand attitude. With the increasing brand attitude of a product in a positive direction, it will affect consumers' purchase intention.

Attitude Towards Influencer dan Purchase
Intention. Attitude towards influencer (AT) positively and significantly affects variable purchase intention (PI). Hasil is supported by research that has been done previously by; Chetioui et al. (2020), which shows that attitude towards influencers has a positive and significant effect on purchase intentions. It is because attitudes owned by consumers based on their assessment of influencers will tend to create a positive action or behavior that will affect consumers' purchase intention. Generation Z, who were the majority of respondents in this study, had social characteristics where they were very intense in interacting on social media with various circles (Ulfa, 2018).
Through these interactions, they can express what they have in mind spontaneously. Due to solid interactions, consumers of this generation tend to be more active in assessing the advertising done by influencers in marketing a product to affect their purchase intention compared to other generations.

CONCLUSIONS
Purchase intention can be influenced by many things that consumers can consider to buy a fashion product advertised by influencers. Until now, no fixed factors can be expressed as an increase in attitude towards influencers and brand attitude and generate the push of consumer purchase intention on fashion products through advertising conducted by influencers. This research resulted in the acceptance of hypothesis 1, namely the influence of credibility perceived towards attitude towards influencer. Hypothesis 2 namely the influence of trust towards attitude towards influencer, hypothesis 4 the influence attractiveness towards attitude towards influencer, hypothesis 5 the influence of attitude towards influencer towards brand attitude, hypothesis 6 the influence of brand attitude towards purchase intention, and hypothesis 7 the influence of attitude towards influencer towards purchase intention. As for hypothesis 3, the influence of perceived expertise on attitude towards influencers is rejected because it has no effect. In addition, this study addressed the gaps found in previous literature with the rejection of the influence of variables and trusts that were accepted in this study.

IMPLICATIONS
From this study, it is known that in receiving or viewing product advertisements conducted by an influencer, consumers prioritize and assess influencers through product references provided at the time of the ad. Providing a good product reference will increase the credibility of an influencer in the eyes of consumers so that in the future, consumers can prioritize product references from an influencer who is considered to have good credit. Consumers also use advertisements conducted by influencers to add personal information related to the latest product offerings and other attributes in a product so that such information can be collected for the personal reference of consumers.
So in choosing the right influencer to do marketing, a product suggested by an influencer can give an accurate assessment of the product so that the delivery can be received well and by consumer expectations when receiving products purchased based on product references made by the influencer. In addition, for businesses, selecting influencers with good credibility will increase the brand attitude of the advertised product brand to increase the consumer's buying interest. It is best to choose influencers who can provide honest product reviews by their standards but can still be attractive to consumers. In addition, businesses can also choose influencers who can provide complete information and details about a product to be advertised. Because consumers also consider influencers as providers of information so that it can be used well by consumers to know a product and can increase purchase intention and directly affect attitude towards influencers. And also, in today's digital era, businesses should use influencer services as supporters of online marketing strategies. Because, through advertising done by influencers, the followers or fans of these influencers will be the first potential consumers for the advertised product. They will prioritize the brand of products advertised by their idol influencers so that this has a positive effect on purchase intention. In addition, influencer services are the right thing to do product marketing in this digital era due to the high number of social media usage in Indonesia.

LIMITATIONS
The limitations found in this study are due to the Covid-19 pandemic. The spread of questionnaires can only be done online so that researchers can not directly know the responses and understanding of respondents in interpreting and answering the available statements.

RECOMMENDATIONS
For the most important fashion industry, companies are advised to use the services of influencers with good credibility, trustworthiness, and appeal as one of the product marketing strategies in this digital era. With consideration, as many as 385 respondents from Surabaya in this study assessed that the effectiveness of an influencer could improve the brand attitude owned by fashion brands. So that consumers will give a positive assessment of the brand advertised by influencers and increase consumers' purchase intentions on fashion products sold by brands that conduct marketing strategies through influencer services. For other researchers, it can make variable perceived expertise a possible research gap to be re-examined.