local inFluEncEr EndorSEmEntS impact on conSumEr purchaSE intEntionS: a rEplication StudY For thE indonESian SmEs

: Several studies have acknowledged that better access to global financial support is one of the SMEs’ major challenges. However, only a few of them are suggesting alternative feasible strategies to assist SMEs to gain market share against major firms. A considerable approach by SMEs is to promote their products and brands match with SMEs capacities through local influencer endorsement to stimulate consumer purchase intentions. Indonesian culture suggests Islamic religious moral leader figure as a type of influencer which is believed to have a significant influence on consumers’ purchase intention in particular on local social communities. The purpose of this study is to analyze the impact of local influencer endorsements on consumer purchase intentions and the moderating effect of negative publicity on the SME market perspective in Indonesia. A structured questionnaire was created to acquire primary data from 273 respondents using a quantitative research design. The findings from this study indicate that attractiveness, trustworthiness, and familiarity are factors supporting the image of local influencer endorsements, which positively influence consumers’ perceptions of quality, purchase intentions, and brand loyalty. This study also shows that negative publicity on local influencer endorsements does not have a moderating effect on consumer purchase intentions.


introduction
According to a 2018 report from the Ministry of Cooperatives and SMEs of the Republic of Indonesia, the Indonesian SMEs growth has increased to 63,933,617 units with a labor absorption of 116,673,416 workers and 298,208.7 billion IDR export income, excluding the income from oil and gas industrial sectors. Although these contributing trends to the country's economy are increasing, the fact that most SMEs' constraints are still related to their access to finance and poor marketing knowledge is rather troublesome (Tahi Hamonangan Tambunan and McKague, 2011). Although SMEs' poor marketing knowledge has been identified as one of SMEs' constraints, it has never been raised as one of the major constraints contributors. Even though, marketing knowledge may lay a foundation that offers important solutions for most SMEs' marketing issues. This argument is supported by the fact that most Indonesian SMEs are still relying only on traditional personal selling methods to promote sales namely door-to-door canvassing techniques and WoM (Word of Mouth) to their existing customers. Another fact also indicates that there are still not enough studies available to suggest alternative marketing strategies for Indonesian SMEs business to compete with major firms.
It is known from previous studies that celebrity endorsements positively affect customers' purchase intentions. Many established corporations commonly use these endorsements as marketing campaign strategies to promote their brands (Osei-Frimpong et al. 2019). McCracken (1989) defines a celebrity endorser as "any individual who enjoys public recognition and who uses this recognition on behalf of a consumer good by appearing with it in an advertisement", where a certain level of recognition and its usage for promotion are emphasized.
In general, SMEs would not be able to compete directly with major established corporations in the context of a celebrity endorsement marketing strategy, whether to gain attention or in funding local celebrities moreover international celebrities. To survive and grow, firms have to strive, earning their market share by knowing how to use marketing mix communication strategies (Keller, 2013), which also true for SMEs business. Thus theoretically, the same concept of celebrity endorsement as a marketing communication strategy should be able to be modified for smaller groups of people, in particular communities that are within the reach of SMEs' business scopes, mainly in regards to promoting their brands' emotional appeal (Solomon, 2017) by using influencer endorsement approach. This study attempts to explore a gap in the previous study of celebrity endorsement effect on purchase intentions by synthesizing the existing research model into a smaller scale to find alternative attainable strategy solutions whereby Indonesian SMEs can gain their larger market share by analyzing the impact of local influencer endorsements on consumer purchase intentions and the moderating effect of negative publicity on the SME market perspective in Indonesia.

mEthodS
A survey was conducted to collect the primary data needed. The survey comprised a structured questionnaire that was given to members of the Indonesian general public who were willing to participate as respondents. Several concerns regarding the exact number of total population existed, time and budget constraints to conduct the research are mainly the reasons behind the chosen method applied for this study. This research employed a single cross-sectional design approach, using Google forms electronic survey methods and a nonprobability convenience sampling technique. Before the main study, pre-test data from the first 30 respondents were collected to test the instrument used for this research (Malhotra et al. 2017).
The survey managed to collect 273 valid respondents with the demographic respondents' profile as presented in Table 1. The survey result was gathered in Jakarta, Indonesia. The survey was conducted within 33 days, from 22 February 2020 to 26 March 2020, using the electronic interviewing method through an internet network with the assistance of a google form application. The survey invitation was conducted via WhatsApp application blast which provides a direct link for the participants around the Indonesian region to a set of constructed questionnaires for them to enter their inputs.
The scales used to measure the constructs, as shown in Figure 1, were taken from an earlier study (Osei-Frimpong et al. 2019). All of the scales were measured on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), apart from attractiveness, trustworthiness, and familiarity, which were measured on a seven-point semantic differential  Figure. 1. Hypothesized model scale. Influencer Endorsement (IE) was constructed as second order, which was measured using attractiveness, trustworthiness, and familiarity. Attractiveness (ATTR) and trustworthiness (TRUS) were measured on a sevenpoint semantic differential scale with a five-item scale (Ohanian, 1990). Familiarity (FAMI) was measured on a seven-point semantic differential scale with a threeitem scale (Spry et al. 2011).
The consumer perception of quality (CPQ) was measured on a four-item scale (Spry et al. 2011). Consumer Purchase Intention (CPI) was measured on a six-item scale (Liu and Brock, 2011). Consumer Brand Loyalty (CBL) was measured on a three-item scale (Yoo and Donthu, 2001). Negative Publicity (NP) was measured on a three-item scale (Money et al. 2006), which was related to subjects considered taboo in Indonesian culture, namely drug abuse and sex scandals. All of these measures with each of their factor loadings are presented in Table 2. group relationships being based on morals and family values that resonate with their higher appreciation for leaders in social group interactions, which is defined by Hofstede's as significant power distance (Hofstede, 2005) in Indonesian culture. Gani (2004) in his study stated that Indonesian leadership characteristics can be considered as having a paternalistic leadership style where influential leaders play the role of a father with its attributes of wisdom and honesty. Marschall (1996) acknowledges two topological leaders in Indonesia; Social Identity Theory, Consumer Behavior, and Social Influencers Social Identity Theory (SIT) defines an individual's understanding that he/she belongs as a member of certain social groups that share emotional and value significance (Tajfel, 1972). Consumer behavior theory proves that social identity has a significant impact on consumer attitudes and beliefs which affect consumers' selections and their decisions to purchase particular products (Solomon, 2017). Social influence can be seen as how an individual who plays a role in a particular social group attempts to assimilate the group's characteristics, which is then often demonstrated in that individual's purchasing behavior towards certain products to attain his/her position in the social group (O'Shaughnessy and O'Shaughnessy, 2002). In any particular social group, one or several individuals are more able to influence the other members of the group; they contribute to shaping the characteristics of the social group and are known as group leaders (Stogdill, 1948). Social influencers are not necessarily those who claim to be leaders in a social group (Stanforth, 1995) but are individuals who are perceived by the other members to have leadership characteristics, to have a high degree of involvement with their social identity, and whose advice is sought for making decisions (Kang and Park-Poaps, 2010). These studies argue that influencer terminology is not exclusively attached to digital social media platforms only, but is also relevant to the more traditional context of social groups.

Model Development and Hypotheses
The characteristics of communicator agents in terms of source credibility are believed to have a significant effect on the persuasiveness of the messages in marketing communication strategies, including messages that significantly affect consumer purchasing intentions (Lafferty et al. 2015). Therefore, the selection of a spokesperson who has a combination of attractiveness, trustworthiness, and expertise in communicating products or brands to consumers is considered to be crucial (Ohanian, 1990). In a related study, perceived endorser familiarity is also considered to be an important factor in conveying successful messages to consumers, where the members of smaller social groups are more likely to be more familiar and close to each other (Spry et al. 2011

Celebrity and Influencer Endorsement
A study to identify how effective celebrity endorsement and influencer endorsement has indicated that compared to traditional celebrity endorsers where they gain public recognition through their professional talent, influencer endorsers gain their acknowledgment by branding themselves as experts on social media platforms (Schouten et al. 2019). This previous study was performed using the same dimensions and variables to compare the effectiveness of celebrity endorsement and influencer endorsement. Thus using this logic, it is safe to assume that the same dimensions and variables that are applied to performing a study on the celebrity endorsement effect on purchase intentions can also be used to performing a study on the influencer endorsement effect on purchase intentions.
Influencers terminology discussed in marketing communication strategies commonly refers to social media influencers, even though the term itself should not be exclusively used only to describe a particular form of media to be used for communication, a digital social media form in this matter. A broader definition of influencer can be found by understanding the behavior of individuals in certain communities through Social Identity Theory (SIT), consumer behavior, and social influencer approach.
satisfaction, and the perceived quality they experience when their expectations have been fulfilled which will lead to long-term customer brand loyalty (Juran and Godfrey, 1999).
h4: Consumers' perceptions of a brand's quality have a positive influence on their brand loyalty.
A previous study stated that brand attitude and brand image interrelate with each other and both act as variables on constructing brand equity and forming brand preferences, which positively affect consumers' purchasing intentions (Chang and Liu, 2009). A similar study by Garretson and Clow (1999) also showed that consumers' perceptions of quality have a positive effect on their purchase intentions.

h5: Consumers' perceptions of the quality of a brand have a positive influence on their purchase intentions
Kamins and Marks (1991) argued that the more consumers are exposed to brands, the more they become familiar and the more consumers' loyalty can be expected towards particular brands. This is then translated further into brand preference, which finally leads to consumers' purchase intentions for specific brands.
h6: Consumers' loyalty towards a brand has a positive influence on their purchase intentions.

The Moderating Effect of Negative Publicity
Although previous studies have shown some evidence of success regarding the use of influencer endorsement in marketing strategy communication, negative publicity from influencers has also been proven to have a significant negative effect on influencers' attractiveness and credibility (Belch and Belch, 2018). The more effective influencers are, the more damage is caused by the impact of any negative publicity. This is because influencers' effectiveness towards brands is mostly achieved by their bringing their image and meaning to the products and brands they are endorsing (McCracken, 1989).
h7: Negative publicity from endorsers weakens the effects of the endorsement's influence on consumers' purchase intentions.
can be achieved by aligning conveyed messages with the endorsers' characteristics. This study also stated that expert endorser is only one of many endorser types that are used to convey persuasive messages to targeted audiences (Erdogan and Baker, 2015). Therefore, in the context of the Indonesian SME market, which is mostly related to FMCG commodity products, there is no urgent requirement for a particular expert type of endorsers needed. Therefore, it could be stated that the variables used to construct influencers' endorsement characteristics are mainly based on influencers' attractiveness, trustworthiness, and familiarity.
h1: Influencer endorsement as perceived in endorsers' attractiveness, trustworthiness, and familiarity has a positive influence on consumers' purchase intentions.

Several articles (Martensen et al. 2018) (Belch and
Belch, 2018) on marketing suggest that influencers with both constructed models of source credibility model (expertise and trustworthiness) (Ohanian, 1990) and source attractiveness model (likeability, similarity, and familiarity) (McGuire, 1985) are more likely to have a significant influence among their followers in establishing a perception of quality for brands or products.
h2: Influencer endorsement has a positive influence on consumers' perception of the quality of an endorsed brand.
A study on customer behavior claimed that influencers with the attributes of attractiveness, trustworthiness, and familiarity are likely to be more successful at promoting brand loyalty, as their influence is more conditionally accepted by their followers. These influencer's attributes lead their followers to a higher degree of attention and recall, more awareness, a certain positive belief, and attitude towards particular brands or products (Solomon, 2017).

h3: Influencer endorsement has a positive influence on consumers' brand loyalty
Consumers' perceptions of quality that come from their expectations of products or brands will turn into satisfaction when their expectations are met (Olsen, 2002). The higher their expectations, the greater their predictive accuracy, and f2 calculates each path model effect and measures the spectrum level of the predictor latent variable whether it has a weak, medium, or strong effect at a structural level.

rESultS data analysis
All of the scales recorded a Cronbach alpha of α > 0,7 with a correlation significance of ρ < 0.05. The PLS-SEM method with SmartPLS software (v.3.2.3) was used to compute the available data. All of the variables measuring the second-order construct (Influencer Endorsement -IE) were treated as formative measures, as shown in Figure 1. A bootstrapping procedure of 1000 subsamples was performed with a significance level using a two-tailed distribution. The indicator outer weights of the formative construct are presented in Table  3 with their levels of significance. All of the indicators show their significance with ρ < 0,05, which suggests their relative importance to the model construct (Hair et al. 2012). Table 2 shows all of the factor loadings and Composite Reliability (CR) show values above 0,6, which indicates that the items were aligned onto the constructs they were intended to measure.

plS Structural model result
A detailed list of the path coefficients with their respective t-values, R2 (coefficient of determinant), Q2 (cross-validated redundancy), and f2 (effect size) are presented in Table 4. R2 measures the accuracy of the model's prediction, Q2 measures the inner-models  The previous study performed in Ghana (Osei-Frimpong et al. 2019) was considered to be appropriate for the adaptation of a similar study in Indonesia, as both Ghana and Indonesia are in the same country category and level of emerging developing markets with several similarities in the consumer behavior aspects.

The result of this replication research conducted in
Indonesia for local influencer endorsers shows a similar interconnections output pattern as it was with its initial research performed in Ghana applied for celebrity endorsers.
This research conceptualization suggests that consumers' perceptions of quality and consumers' brand loyalty are both important in their relationships supporting the influencer endorsement's positive impact on consumers' purchasing intentions. Although they significantly support the proposed research model, the variables show just a moderate interrelationship caused by dispersing respondent responses, which raises a possible indication of unequal appreciation of public trust towards types of Indonesian influencer leaders (i.e. official-design leaders and spiritual-moral leaders) that should be covered in further studies. The focused response from participants towards consumers' perception of a quality variable more likely would strengthen the interconnection value of consumers' brand loyalty variable which then leads to higher interconnection value to consumers' purchase intentions. Since the questioners structured as per initial research without identifying the local influencer endorser in particular as official-design leaders or spiritual-moral leaders, disperse participants' respond are expected between the two types of Indonesian leaders. Nevertheless, the result of this study remains supporting the concept that significant local figures indeed provide a positive influence on brands being endorsed which then likely to affect the consumers' perceptions of brand quality. Table 4 implies that hypothesis H7 (M = -0.059; ρ = 0.169; t-value = 1.294; f2 = 0.005), negative publicity as a moderating variable of influencer endorsement towards purchasing intentions was not supported and has a nonsignificant effect as the correlation significance value did not meet the requirements of ρ < 0,05 and a t-value of > 1.96. Figure 2 shows the path effect between Influencer Endorsement (IE) and Purchasing Intentions (PI) and how the moderating effect of Negative Publicity (NP at -1 SD and NP at + 1 SD) was not met while conducting this research.

the interaction Effect
Negative publicity in theory suggests the weakening effect of an influencer endorsement's impact on consumers' purchasing intentions (McCracken, 1989). However, this current study provides evidence that the negative publicity moderating variable by itself could not decisively determine the effect of influencer endorsement on consumer purchase intentions without taking into account the components constructing the influencer endorsement variable. The more significance that exists between the interconnection from the components constructing the influencer endorsement more likely would reduce the significance of the effect from negative publicity moderating variable and vice versa. The implication above correlates with the existing study regarding Social Identity Theory (SIT), consumer behavior, and social influencers which argue that the stronger bonds and influence from a significant figure is relatively stronger in smaller social communities compares to larger ones.
Contrary to a previous study by (McCracken, 1989), this study shows that although there is a trend indicating the potential for a weakening effect to occur, as shown in Figure 2, the result remains to show that there is not enough evidence to conclude the weakening effect to the negative publicity that may arise towards the influencer figures. A strong bond of connection among people relatively can be established in a smaller group of society; therefore the wise approach for managers is to study and target the society influencer figures which match with the SMEs products and brands to be endorsed. Furthermore, although the approach suggested is not limited to Islamic moral and spiritual influencer leaders only, it is now becoming clearer that this type of influencer figure is currently to be considered as the most suitable marketing influencer endorsement to SME products and brands for the Indonesian market.
Based on the arguments proposed above, SME managers should not hesitate to perform this alternative strategy approach on using local influencer endorsement to promote their products and brands moreover for this strategy to be copied by established major firms. Many major firms most likely already have their target consumers with more general characteristics where the firms' marketing strategy is based upon it. Furthermore, this approach is considered to be appropriate for SME practitioners, as the local influencer endorsement strategy would likely more efficient when it is built from a smaller local social community that is more immune to negative publicity and providing more bonds to its consumers before stepping into the next larger market.
exists in this study. This phenomenon can be explained with the assistance of the selective distortion concept by (Kotler and Keller, 2016), which argues that some people tend to interpret information in a way that supports concepts already structured in their mind in which they believe to be true.

managerial implications
This study has established the importance of influencer endorsement as an alternative marketing approach for SMEs in emerging developing countries such as Indonesia by using influencer figures in social groups. SME managers are expected to selectively match the characteristics of the influencers they use to endorse their brands. SMEs that choose to endorse Islamic brands are encouraged to approach a type of Indonesian moral leaders who are associated with Islamic religious leaders in their local societies. Furthermore, these Islamic moral and religious leaders may act as a bridge to other Islamic organization networks, namely madrasahs, Islamic boarding schools, Islamic political parties, and other Islamic religious networks.
It is also acknowledged that the deeper the influence of figures is planted on consumers' minds, the higher the effect of influencer endorsement on those consumers. At the same time, the more it becomes less sensitive