Effects of Social Media Marketing Towards Probiotic Chicken Customers’ Purchase Intention: Customer Engagement as a Mediator

_____________________________________________________________________________________________________ Social media has become an inseparable part of marketing, particularly for small and medium enterprises. However, there is a limited knowledge on how it helps small and medium enterprises in selling probiotic products. This study aims to explore how social media marketing influences the purchase intentions of probiotic chicken customers. This research uses a quantitative approach to collect data through an online survey from 215 participants. The results show that the factors of entertainment, trendiness, and e – word of mouth of social media marketing, alongside customer engagement, significantly affect customers’ purchase intentions. Data analysis reveals that the factors of social media marketing have greater influence towards purchase intention when mediated by customer engagement, compared to their direct influence towards purchase intention. Therefore, improving the quality of the social media marketing factors enhances customer engagement. Hence, customer engagement also plays an important role in social media marketing by directly influencing the customers’ purchase intentions.

Brands from different industries are swarming the platforms to find new customers, including the probiotic chicken industry. The probiotic chicken industry is part of the broiler chicken industry, where chickens are fed with Lactobacillus bacteria and herbs as food support to grow fast without the use of antibiotics and hormones (Trifiana, 2019). Currently, probiotic chicken is rising in demand as 90% of Indonesian households are consuming healthier menus during the covid-19 pandemic (Kantar Worldpanel Indonesia, 2020). Moreover, nutritionists predict that probiotics will rise in trend in 2021 (Afrilian, 2020). Probiotic chicken brands in Indonesia are using social media as one of their primary marketing channels, but it cannot be determined whether social media creates a significant impact to their sales.
At the same time, issues of customer engagement appear to have received substantial consideration in research concerning social media. Dessart et al. (2015) had found evidence of online brand communities experiencing engagement among community groups and direct customer-brand relationships in the online realm. The subjects of their research reported elements of customer engagement helped in driving purchase. Thus, with the growing technology, brands need to think ways to make use of social media not only to obtain revenue but also to create a long-lasting engagement with their customers. Cummins et al. (2014) also mentioned that the extent of customer engagement in social media include possible increase in customer satisfaction, retention, loyalty, customer lifetime value, and profitability. Kaplan & Haenlein (2010) mentioned how it is customary within social media for customers to be able to interact with brands. Through customer-brand engagement in social media, bonds can be easily formed to create value, and thus could lead to favourable results to the brand.
Nonetheless, the subject of social media along with customer engagement is still under-researched (Barger et al., 2016;Dessart et al., 2015;Husnain et al., 2017). Knowing the potential that social media offers, such as Instagram offering an engagement rate of 1.18% per post for food and beverage industry, and Facebook offering 0.10% respectively, they become an opportunity that are hard to miss by marketers (Chen, 2021). But it is also known that social media marketing does not generate equally profitable results for all industries. It has been revealed that social media is an effective platform for brands that include fashion, food, and technology, but is less effective to financial services (Guo, 2016). Therefore, there is a lot of potential that can be explored from social media as a marketing platform towards the purchase intention of a probiotic chicken business. This research also intends to analyse the role of customer engagement within the relationship between social media marketing and customers' purchase intention.

Social Media Marketing
Social media is a collection of applications that are based on the web that allow the creation and the exchange of user-generated content (Kaplan, 2015). User-generated content comes from users of social media who voluntarily give information, data or media in a useful or entertaining way (Krumm et al., 2008). The extent of social media has challenged businesses to understand how to best utilize the various social medias available on the internet to attract users to their brands and convert sales. Each social media application is unique in their nature. Hazlett (2010) mentioned that social media is used as a reference for many services on the internet that drove the users to participate in creating and exchanging content, joining into online communities (e.g., Tumblr, Reddit), social media sites (e.g., Instagram, Facebook), and media sharing sites (e.g., YouTube, Tik Tok). The components of social media marketing have been part of discussions in several subjects, such as consumer goods. Kim & Ko (2012) characterized social media marketing activity entertainment, interaction, trendiness, customization, and word-of mouth (WOM) which were implemented to luxury products. It is also known that in the fast beverage industry, social media marketing activities are categorized into entertainment, interaction, trendiness, advertisement and customization (Jo, 2013). Sano (2009) also identified the components of social media advertising including interaction, trendiness, customization, and perceived risk which was applied on an insurance service. www.msocialsciences.com

Customer Engagement
Customer engagement refers to the process of human psychology which shapes customers' loyalty towards a newly discovered brand, along with the mechanism that allows sustainability towards repeated purchase (Bowden, 2009). Customer engagement is one of the media component of a brand that a customer has encountered that leaves an impression that is favourable towards the brand (Cummings, 2007). It is also known that the use of social media by the masses support customer engagement (Pütter, 2017). Unlike more traditional marketing strategies in which customers were mere recipients of information, the newer mediums of marketing, including social media requires engagement in consideration as it plays a big role to let the consumer participate in, and influence the way of the interaction within the business. Vivek (2009) suggests in his research on the scale of customer engagement, that the dimensions of the theory of consumption values (Sheth et al., 1991) are used when observing the outcome of customer engagement; as the level of customer engagement is associated with the value received by the consumer. Vivek (2009) derived both intrinsic and extrinsic values of a business's customer engagement to be associated with higher customer goodwill, word of mouth (WOM), connections with the organization, and affective commitment.

Purchase Intention
Intention is used by brands to predict the sales of upcoming products or to predict the repeated purchase of an existing product (Diallo, 2012). Purchase intention occurs within the customer journey where the customer has had their readiness to purchase a product (Dodds et al., 1991). In another word, purchase intention becomes an imperative indicator to predict the likelihood for a customer to purchase a product. Many studies have identified factors that are affective towards purchase intention. In the online realm itself, more than 80 factors are identified as influential towards purchase intention (Chang et al., 2005;Rose et al., 2012;Thamizhvanan & Xavier, 2013). Regarding social media, the extent of it shows that the applications may contribute in improving the consumer decision making process more than traditional methods (Bruhn et al., 2012). It is also known that customer engagement may serve in increasing purchase intention among customers (Husnain et al., 2017). The role of customer engagement as a mediator of the study could facilitate marketers in interacting with social media marketing to increase customer purchasing intention. Since it is not possible to explore all the factors that are affective towards customer purchase intention, this research is restricted to investigating the effect of social media marketing and customer engagement towards purchase intention.

Hypothesis Development
As a mean to provide communication with customers, now brands in the broiler chicken industry have shifted to social media. Social media is a perfect fit for providing two-way communication among brands and users. The interaction between users and brands in social media has helped create positive attitude towards the brands present. Furthermore, they may support stronger user commitment and affect the decision-making towards a purchase (Hutter et al., 2013). It has also been found that social media marketing influences purchase intention (Z. Wang et al., 2016). Therefore: H1: There is a positive relationship between the social media marketing factors and purchase intention.
Nowadays, with an extent of marketing mediums and online shopping experiences, customer engagement is a referred term in a brand's efforts in creating emotional ties which can influence word of mouth and upcoming sales (Dessart et al., 2015). A previous study by Barhemmati & Ahmad (2015) found that social media marketing is significantly related to customer engagement. Social media may be used to engage with customers by creating value for them. Pütter (2017) also argued that social media, a platform that is used by the masses, support customer engagement. Following the arguments, a hypothesis is made accordingly as follows: H2: There is a positive relationship between social media marketing factors and customer engagement.
www.msocialsciences.com Barhemmati and Ahmad (2015) mentioned in their research that the emotional ties between customers and brands that are made through engaging with the customers, offer the brands an immense prospect to achieve their marketing objectives that are related to customer relationship. These prospects may help the company in persuading customers to purchase the products or services offered. Based on the arguments, a hypothesis was made as follows: H3: There is a positive relationship between customer engagement and purchase intention.

Mediation
The following section discusses the relationship among social media marketing, customer engagement and purchase intention. Husnain and Toor (2017) implied that if managed proficiently, the engagement created through social media marketing may lead to an extended relationship between the brand and the customer and will likely increase sales. Regarding the arguments, another hypothesis is made accordingly: H4: Consumer engagement is a mediator of the relationship between social media marketing factors and purchase intention.

Conceptual Framework
Conceptual framework is illustrated in Figure 1 in accordance with the current research limitations. The independent variable consists of social media marketing components consisting of Entertainment, Interaction, Trendiness, Advertisement, Customization, Electronic Word-of-Mouth. The relationship between the independent and dependent variable is mediated by customer engagement. Meanwhile, the dependent variable of this study is Purchase Intention. The dimensions of Social Media Marketing variables of this research are taken from Kim and Ko (2012), Jo (2013), and Sano (2009)

Population and Sampling
Referring to Malhotra's Book of Marketing Research (2010), the researcher uses nonprobability sampling, specifically quota sampling for this research. Quota sampling is done by determining the control elements, which are; (1) Living in the areas of Jakarta, Tangerang, Bogor, Depok, or Bandung, (2) Either had bought products of, or seen the social media of any raw chicken brands. The population of this research are taken from areas with a growing number of probiotic chicken buyers and prospects to increase the validity of the data gathered. The control element or quota ensures that the composition of the sample is the same as the population accordingly. This sampling requires the respondents to follow the control elements, and that for a test-marketing study, the minimum size of the sample is 200.

Operationalization of Constructs
The attributes and questions for the questionnaires are adapted respectively from previous literatures ( Table 1). The questions regarding social media marketing in the questionnaire were adapted from previous research done by Jo (2013), Kim and Ko (2012) and Sano (2009). The questions regarding customer engagement are adapted from Vivek (2009), Bove et al. (2008), Bansal et al. (2004), Anderson (1998), andMittal (1988). And the questions regarding purchase intention were adapted from Husnain et al. (2017) and . The questionnaire contains a series of questions that will assess the respondents' perception towards the topic. The questionnaire uses 5-Point Likert Scale, which consist of five points indicating; (1) Strongly disagree; (2) Disagree; (3) Neither agree nor disagree; (4) Agree; (5) Strongly agree. Vivek (2009) I am motivated to respond to communications from this brand This company goes out of its way for customers Bove et al. (2008) This company always tries to do the right things for customers This company has a genuine concern for customers I feel emotionally attached to this company Bansal et al. It is likely that I will buy the products recommended by this brand Husnain et al. (2017),  I will purchase this brand's products the next time I need probiotic chicken If a friend asked me for advice on which probiotic chicken products to buy, I would advise them to buy this brand's products I will definitely buy the products recommended by this brand I have strong intentions to purchase the products recommended by this brand Using social media of brands help me make better decisions before purchasing products Using social media of brands help in increasing my interest in buying products I am highly likely to buy products or services recommended by my friends on social media

Participant Demographics
Out of 219 responses collected, 215 is eligible to process, resulting in 98.17% eligibility rate. As seen in Table 2, most of the respondents are between the age of 18 and 25 (76.28 per cent). The distribution of males (47.4 per cent) and females (50.7 per cent) are almost equal. Of the 215 respondents, 20.93 www.msocialsciences.com per cent were employed, 66.5 per cent are students, 14 per cent own businesses and 6 per cent are stay at home wives or husbands.

Measurement Model
To ensure that the constructs used are valid and reliable, a convergent validity test, discriminant validity test, composite reliability test and indicator reliability test were conducted. The model uses constructs that are valid and reliable. Thus, all the constructs that do not pass the requirement are removed from the model. Table 3 shows the constructs that are valid and reliable, and further used in the model. All the constructs have Average Variance Extracted (AVE) values above 0.5 (Wong, 2013), making all of the constructs valid. The composite reliability also show values over 0.7 (Wong, 2013), ensuring that all of the constructs are reliable. The indicators are also deemed valid according to the discriminant validity result. Table 4 shows that AVE values surpassing the values of each of the constructs' correlations (Wong, 2013), the constructs in this model have passed the convergent and discriminant validity requirements. The outer loadings value of the indicators used in the model have all surpassed 0.7 (Wong, 2013). It can be concluded that all the indicators are reliable according to the indicator reliability results. The data used are also required to be free of collinearity. According to Table 6, the indicators used in the model have VIF values below 5 (Wong, 2013), making them free of collinearity problems.

Structural Model and Hypotheses Testing
The examination of the hypotheses is done by analysing the T-values based on 95% confidence interval (Setiaman, 2020). The relationships shown in Table 7 show T-values of over 1.96, meaning that they are significantly positively correlated. The model's coefficient of determination R² values surpassed the value 0.33 but are below 0.67, meaning that the constructs have moderate strength to explain the variances in the corresponding constructs. The remaining percentage of variances of the constructs are explainable by the constructs that are not used in this research. The results imply that entertainment can influence 13.6% of customer engagement and 12% of purchase intention significantly. Therefore, H1 and H2 are supported for the entertainment factor. Thus, entertainment's power to influence customer engagement is higher than its' power towards purchase intention. Similarly, eword of mouth can influence 41.6% of customer engagement and 16.5% of purchase intention significantly. Therefore, H1 and H2 are supported for the eword of mouth factor. Meanwhile, trendiness can influence 21.1% of customer engagement, and does not influence purchase intention directly. Therefore, H1 is supported for the trendiness factor. Additionally, customer engagement can significantly positively influence 40.4% of purchase intention. Hence, H3 is supported.

Mediation
In addition to analysing the direct effect, this research also tested the mediating role of customer engagement on the relationship between social media marketing and purchase intention using Baron and Kenny's (1986) approach. As shown in Table 7, the factors entertainment and eword of mouth are significantly related to the customers' purchase intention. This fulfils the first requirement that the independent and dependent variables' relationship must be significant. Additionally, both factors alongside trendiness also have a significant relationship with customer engagement. Thus, entertainment and eword of mouth fulfil the second requirement that the relationship between the independent variable and the mediator must be significant. It is known that the presence of customer engagement in the model reduces the effect of entertainment and eword of mouth towards purchase intention. Therefore, the last requirement is fulfilled and H4 is supported for the relationship between entertainment and eword of mouth towards purchase intention.

Discussion
The findings of the analysis show the relationship of social media marketing, customer engagement, and purchase intention. Previous literatures have identified that social media marketing and customer engagement are inseparable due to social media's interactive nature (Barger, Peltier, and Schultz, 2016), as the use of social media itself supports customer engagement (Pütter, 2017). Cummins et al. (2014) also mentioned that the role of customer engagement in social media could possibly increase profitability. The literatures are in line with the finding of this research that finds a significantly positive influence from social media marketing to customer engagement, and lower power to no significantly positive direct influence of social media marketing to purchase intention. Hence, the presence of customer engagement in this study accommodates the reason for companies to implement social media marketing to increase the customers' purchase intention.
Data analysis show that the Entertainment and E -Word of Mouth aspects appear to have a significant indirect influence towards the customers' purchase intention in the case of probiotic chicken industry in Indonesia. As predicted, customer engagement mediates the relationship between the aspects and purchase intention. Thus, probiotic chicken brands need to maximize these aspects of social media marketing, to enhance the brand-customer engagement and potentially leverage the customers' intention to purchase. While the relevance of the contents of the brands' social media appears to have a significant influence on the customer engagement. Hence, the brands can also consider maximizing the trendiness of its' social media contents to drive higher brand-customer engagement.

Conclusion
The purpose of this research is to explore the social media marketing aspects that influence the customers' intention to purchase probiotic chicken products. The researcher has analyzed the social media marketing aspects that includes entertainment, interaction, trendiness, advertisement, customization, and e -word of mouth. During the analysis, the researcher also found that customer engagement mediates the relationship between social media marketing and purchase intention. The data was collected using a Google Form questionnaire that was distributed to 215 respondents in Jakarta, Bogor, Depok, Tangerang, Bekasi, and Bandung city that have been a customer of or have seen the social media accounts that sell raw chicken. The data was analyzed using Partial Least Squares -Structural Equation Modelling.
The findings of this research may contribute to the world of business research in many ways. The study found that the aspect of trendiness is revealed to have a direct influence towards customer engagement. Thus, increasing the quality of the aspect potentially increases the customer engagement. While the aspects of entertainment and eword of mouth have both a direct and indirect influence on purchase intention and increasing the quality of the aspects may both potentially increase brand-customer engagement and purchase intention. Nonetheless, this study hghlights the ability of social media's outcome to impact customers' purchase intention. The study also highlighted that the relationship between the brand and the customer delivers the values of entertainment and eword of mouth of social media marketing to potentially increase the customers' intention to purchase. The emotional tie created with the customers may amplify the effect of the entertaining and shareable content to leverage the customers' intentions to buy probiotic chicken products.
The results founded in this study imply that probiotic chicken sellers that market their products through social media need to maximize their accounts' entertainment values. Because through giving the customers feelings of joy while observing their social media accounts, can both give the customers a feel of an emotional tie with the brand and could also directly increase the customers' intention to purchase. As of trendiness values, the information that is displayed on the brands' social media account need to be up-to-date and show information that are relevant to the current conditions of the society. Showing that the brand keeps up with the trend gives a passageway for the customers to feel closer or more related with the brand. This will increase the brand-customer relationship. Meanwhile, increasing electronic word of mouth values of the brands' social media accounts may lead to a stronger brandcustomer relationship and the purchase intention directly. The contents of the social media account has to be shareable, which means that customers can freely and willingly share it to their friends and families. With the high shareability of the social media contents, the brand-customer relationship will be built as the customers' intention to buy the probiotic chicken also rises.
samples from other cities. The research also only implies within the probiotic chicken sector, while other sectors in the poultry industry remain unexplored. Previously, it has been revealed that the coefficient of determination of each of the relationships is moderate, and less than half of the variances of the dependent variables remain unexplained. Hence, further research can be conducted to find out the corresponding variables that are not yet explained in this research.