Online Marketing Enterprise of Jombang Culinary from TAM and VEM Perspective on Social Media

The purpose of this research was to develop, evaluate, and test antecedent model with trust as an intermediary variable of Technology Acceptances Model (TAM) and Virtual Experiential Marketing (VEM) for buying intention. It used purposive sampling technique with 300 respondents. The questionnaire results were evaluated and analyzed using Structural Equation Model (SEM) method with the analysis tools (the third version of SmartPLS). The research result shows that there is no effect between TAM and trust. In addition, there is a direct and positive influence of VEM on trust. The trust and TAM have a positive effect on buying intention. However, VEM does not affect buying intention. The results indicate that the trust cannot mediate TAM and buying intention. Meanwhile, trust can mediate between VEM and buying intention.


INTRODUCTION
The emergence of information technology has a great impact on the global market structure (Sin et al., 2012). In the business of information technology, it has brought fresh air for business people from large industrial sector and home industry sector. One of the most important information technology products is the Internet. With the Internet, it can eliminate time and distance barriers and change the human pattern in interacting and communicating. By sitting in front of the computer connected to the Internet, the potential customers can see, compare, and know all the information about a product that will be purchased without coming to the store. Business owners can use online stores and social media to conduct promotions and sales by uploading information and photos of the products through the Internet. Then, the online store or social media will be accessible automatically by prospective customers for 24 hours. Consumers can also make the transaction process without meeting with the buyer.
The Internet users in Indonesia reached 82 million people or 30% of the Indonesia population with Rp130 trillion as the value of e-commerce transactions in 2013 (Sidharta & Sidh, 2014). From those data, it has proven that Internet has become a living part of society in Indonesia. Furthermore, Kominfo (2013) stated that 95% of Indonesian used the Internet to access social media with Facebook as the most frequently accessed social network.
With the increasing popularity of smartphones, it also increases the size and importance of Social Networking Sites (SNS), such as Facebook, Myspace, Plurk, Twitter, and Youtube (Zhou et al., 2010). Many companies use Facebook to engage in marketing activities such as strengthening brand image and product promotion (Trusov et al., 2010). Jombang is one of the districts in East Java that has many potentials. In addition to religious tourism, Jombang also has a wealthy culinary to be proud of. Micro, Small and Medium Enterprise (MSME) engages in food and beverage in Jombang. It is in the form of a community in the virtual world by the name Jombang Culinary (JOKUL). The total members are about 53.338 people. This is done to survive and compete especially against the big traders.
To survive in the increasingly fierce competition, it needs an innovation in product and strategy. One form of innovation in marketing done by the JOKUL is to make a sale through Facebook. The reason for choosing Facebook is because of its wide coverage, feedback facility, and quick comment. Moreover, it can send images, video, and chat. Free registration is also a privilege of Facebook. With the innovations in marketing, online marketing is expected to be a competitive advantage for JOKUL. However, choosing the right media is not enough. It needs a proper and reliable strategy so that the business can be sustainable and increasing. One of strategy that can be used is by applying the concept of Technology Acceptance Model (TAM) and Experiential Marketing (VEM). TAM offers a strategy to predict consumer behavior in accepting a technology for potential customers (Ha & Stoel, 2009). TAM is derived from the theory of Theory of Reasoned Action (TRA) stated by Ajzen and Fishbein (1985). The concept of TRA is developed to predict the intention of behavior or behavioral intention. Then, Davis (1986) adopted the concept with the assumption that the intention to behave was determined by two major variables, perceived ease of use and perceived usefulness. In the concept of TAM, it explains the cause and effect of the relationship between attitude, belief, and intention to behave (Ha & Stoel, 2009).
Perceived ease of use is the extent to which people believe that the use of technology is very easy (Ha & Stoel, 2009). This variable is one of the most important variables in influencing the desire to buy (Tsu Wei et al., 2009). The indicator of this variable is derived from the research of Alalwan et al. (2016). The second variable in TAM is perceived usefulness. It is the degree to which people believe that in using a particular technology or system, it will develop and accelerate the performance of their work (Davis, 1986). From the results of previous research, it indicates that perceived usefulness positively affects the online purchase intention (Wu & Ke, 2015;Cho & Fiorito, 2009). The indicators used in this variable, are taken from the research of Alalwan et al. (2016).
By using TAM, the company can know the attitudes and behavior of their potential customers, and the technology used to conduct the marketing process. If the consumers accept and believe that the technology like Facebook is useful for them and easy in use, and able to bring them closer to the company, it will trigger consumers to increase their desire to buy products offered by the company. TAM provides a useful basis for research related to consumer behavior in the acceptance of online shopping technology (Ha & Stoel, 2009). Facebook is used as marketing tool by JOKUL, although in Indonesia the trust toward online shop is still low. The media used is the same, and the similar products are offered by competitors. Facebook has the same template and the same facilities. Thus, it becomes a challenge for the company to design and create contents that will be displayed on Facebook. It should be interesting and feasible to be published. It is important to apply experiential marketing as the second solution after the TAM.
The company must be aware that developing an experiential environment is very important to do because it will become a competitive advantage and it is difficult to imitate or eliminate by the competitors (Tsaur et al., 2007). The application of experiential marketing is not only done offline or in the real world, but it can also be done in the virtual world (Internet) through Facebook. The concept is called as VEM. VEM can embrace Internet and its network including blogs, chat rooms, interactive images, virtual communities, multiplayer online games, and technology to create an immersive experience through visual and audio cues (Luo et al., 2011).
The origin of VEM comes from the theory by Gilmore and Pine (2002). Consumers are ready with the technology that has been used by the company. Then, the social media has been designed in a way that leads to a different experience. However, it is still considered not enough to move the consumers to buy interest through Facebook. Trust is an important variable to be considered to optimize the applied strategy. Trust is defined as a subjective belief that online service providers will fulfill the transactional obligations such as the obligations they have promised and are understood by consumers (Kim et al., 2008). Trust plays more important role for an online business than offline business. It is because consumers cannot see the stores directly and check the items that they want to buy (Hong & Cho, 2011). According to Wen (2010), consumers' confidence in the online shop has a positive relationship with the intention of buying online products offered on the site. Moreover, Wu and Ke (2016) agreed that trust variables could mediate the relationship between TAM and buying intention. Understanding consumers' buying intention is paramount because the final decision to make a purchase can be predicted (Bai et al., 2008). Consumers decide what they buy based on the information they get (Kim et al., 2008;Pappas, 2016).
It is known that there are many frauds occurred in cyberspace because it only displays the product images and brief information in the online shop. Consumers cannot meet directly with the buyer and hold the goods that we will buy. With this circumstances, many irresponsible people use the opportunity to deceive the consumers. Therefore, it is not infrequently what consumers' orders do not match with the photos on the webpage or consumers have transferred the money, but the goods never come. One of the reasons why many consumers use the Internet but they do not want to buy is because of the security factor in doing business via the Internet (Gefen et al., 2003). Uncertainty and risk can be reduced if consumers have trust (Ha & Stoel, 2009;Tugiso et al., 2016).
The selection of TAM variables is based on inconsistent research results of Wahid and Iswari (2007) and Widiana, Supit, and Hartini (2012). The research object of both researches is similar which is MSME. Wahid and Iswari (2007) stated that the adoption of information technology for MSME in Indonesia is very weak and only touches the level of functional integration. However, Widiana, Supit, and Hartini (2012) suggested that TAM through ease and confidence variable made consumer repeat buying because consumers felt satisfied.
The use of VEM is based on the research by Chen et al. (2008). They stated that VEM could influence consumer behavior towards buying intention and loyalty. One of the requirements that had to be met by company in creating VEM was web design that created a positive effect on the user and improved the favorable consumer response (Dailey, 2004). However, templates on Facebook cannot be changed.
This research is to develop, evaluate, and test antecedent model with trust as intermediary variable of TAM and VEM for buying intention. It is expected that by knowing those variable, it can be used as basis for designing content and appearance which will be uploaded to Facebook that has unique and different value and can influence consumers to buy the product. Thus, it can be one of the applied strategies to compete in online business.
Based on the background that has been presented, seven questions are formulated as the problem in this research. First, it is to see whether there is a positive influence of VEM consisting of sense, interaction, pleasure, flow and community relationship on trust. Secondly, it is to find out whether there is influence of TAM consisting of perceived ease of use and perceived usefulness on trust. Third, it is to see whether there is positive influence of trust on buying intention. Fourth, it is to find whether there is effect of TAM on buying intention. Fifth, it is to see if there is influence of variable of TAM on buying intention through trust. Last, it is to analyze whether there is indirect influence of experiential marketing variable on buying intention through trust. In this research, there are several hypotheses proposed. First, there is positive influence from TAM consisting of perceived ease of use and perceived usefulness on trust (H1). Second, there is positive influence of VEM consisting of sense, interaction, pleasure, flow and community relationship on trust (H2). Third, there is positive influence of trust on buying intention (H3). Fourth, there is positive influence of TAM consisting of perceived ease of use, and perceived usefulness on buying intention (H4). Fifth, there is positive influence of VEM consisting of sense, interaction, pleasure, flow and community relationship on buying intention (H5). Sixth, there is indirect influence of TAM on buying intention through trust (H6). Seventh, there is indirect influence of VEM on buying intention through trust (H7). The direction of the relationship is described in Figure 1.

METHODS
The research method used is survey. According to Singarimbun and Effendi (2006), it takes sample from a population and uses questionnaire as data collecting tool which is individual. This survey research uses explanatory research type (explanatory research) which explains the causal relationship between the variables through hypothesis testing. The unit of analysis is the consumer of JOKUL as the individual. The population in this research is all consumers who have made purchases of JOKUL products through JOKUL Facebook. To determine the sample, the researchers use purposive sampling technique with the criteria of consumers who have ordered products on JOKUL Facebook at least once. Moreover, questionnaire is a data collection technique which is done by giving a set of questions or written statement to the respondents to be answered. According to Ferdinand (2014), the number of good samples for the Structural Equation Modeling (SEM) ranges from 100 to 200. Based on that theory, this research uses sample of 300 respondents disseminated through the Internet using messaging function on JOKUL Facebook.
The data processing in this research uses SEM method by SmartPLS third version. Meanwhile, the criteria of PLS output to do interpretation use two models. Outer model (indicator test) uses convergent validity, discriminate validity, average variance extracted and composite reliability. Then, inner model (hypothesis test) uses R2 for endogenous latent variable and T-statistical parameter coefficient.

RESULTS AND DISCUSSIONS
The validity evaluation is done by calculating the convergence validity and discriminate validity. Convergent validity is known through loading factor. An instrument is said to fulfill the convergence validity test if it has a loading factor above 0,6. The convergent validity test results are presented in Figure 2 and Table 1. Based on that, it can be seen that all indicators that measure TAM, VEM, trust, and buying intention are greater than 0,6. Thus, the indicator is said to be valid to measure the variable.
The calculations that can be used to test the construct reliability are Cronbach alpha and composite reliability. The test criteria state that if the composite reliability is greater than 0,7 and the Cronbach alpha is greater than 0,6, the construct is reliable. The Cronbach alpha and composite reliability calculation results can be seen through the summary presented in Table 2. It can be seen that the value of composite reliability on TAM, VEM, trust, and buying intention is greater than 0,7. Thus, all indicators that measure TAM variables, VEM, trust, and buying intention are reliable.   The influence of TAM on trust results in 0,841 of t-statistics. The test results show that the value of t-statistics is <1,96. This means there is no significant influence of TAM on trust. It shows that H1 is rejected. The results of this research are in accordance with Wahid and Iswari (2007). They stated that the adoption of information technology for MSME in Indonesia was very weak and only touched the level of functional integration only. From the consumers' side in Jombang society, they usually use social media, but they will make a purchase immediately regardless of the store that sells offline.
The influence of VEM on trust results in 8,056 of t-statistics value. The test results show that the value of t-statistics is >1,96. This means there is a significant influence of VEM on trust, so H2 is accepted. The results are in accordance with the research of Chesney et al. (2017). They suggested that VEM could influence consumers' confidence to buy especially with the product figure that supported communication and communicate in chat.
The influence of trust on buying intention is 5,491 of t-statistics. The test results show that the value of t-statistics is >1,96. It means there is a significant influence of trust on buying intention. Hence, H3 is accepted. The results are in accordance with Bansal, Zahedi, and Gefen (2016). Trust is a very significant variable in e-marketplace and plays a very important role in influencing buying intention. The event of JOKUL is to strengthen the hospitality and consumers can meet with sellers in there. From the meeting the sellers directly, the bonds of trust is formed.
The influence of TAM on buying intention is 2,107 of t-statistics. The test results show that the value of t-statistics is >1,96. This means there is a significant influence of TAM on buying intention, so H4 is accepted. The results are contrast with Putra (2017). Putra (2017) stated that TAM could not affect the variable of behavioral intention without going through mediator variables. However, in this research, TAM can affect buying intention without mediator variable. It is because they have bought product offline previously so when they are familiar with the new offline store, the next order is through social media.
The influence of VEM on buying intention shows 1,437 of t-statistics. The test results show that the value of t-statistics is <1,96. This means there is no significant influence of VEM on buying intention. Thus, H5 is rejected. The results are in accordance with the results of Bauer et al. (2015). They suggested that the characteristics of online shopping were always problematic with the level of consumer confidence. Consumers do not necessarily believe. If consumers do not believe, the interest of buying does not occur.
The indirect effect of hypothesis testing is done to test whether there is the indirect influence of an independent variable on the dependent variable through intervening variable. The test criteria state that if the path coefficient is positive and t-statistics are ≥ t- Based on the test in Table 4, it can be seen that the influence of TAM on buying intention through trust is 0,831 of t-statistics value. The test results show that the value of t-statistics is <1,96. It shows that there is no significant influence of TAM on buying intention through trust. Thus, trust cannot mediate the effect of TAM on buying intention. H6 is rejected.
Then, the influence of VEM on buying intention through trust obtains by 4,537 of t-statistics value. The test results show that the value of t-statistics is >1,96. This shows that there is a significant influence of VEM on buying intention through trust. Hence, the trust can mediate the influence of VEM on buying intention. H7 is accepted. The results are similar with Gefen and Straub (2004) that trust is the most important element in trade. There is no transaction without trust based on it. It also confirms the result of Putra (2017) that TAM cannot be affected by behavioral intention directly without going through attitude.
The goodness of fit model is used to know the ability of endogenous variables, explain the diversity of exogenous variables, and know the contribution of exogenous variables to endogenous variables. The goodness of fit model in PLS analysis is done by using Q-square predictive relevance (Q2). The goodness of fit model that has been summarized in Table 5. Then, the R-square of trust variable is worth 0,454 or 45,4%. It can show that the diversity of trust variables can be explained by TAM and VEM of 45,4%. In other words, the contribution of TAM and VEM variable to the trust variable is 45,4%. Meanwhile, the rest of 54,6% is the contribution of another variable that is not discussed in this research.
The next is an R-square variable of buying intention has value 0,254 or 25,4%. It can show that diversity of buying intention variable can be explained by TAM variable, VEM, and trust of 25,4%. In other words, the contribution of TAM, VEM, and trust to variable buying intention is 25,4%. However, the rest is equal to 74,6% for the contribution of other variables that are not discussed in this result.
Then, the Q-square predictive relevance (Q2) is 0,593 or 59,3%. It can show that diversity of buying intention can be explained by the overall model of 59,3%. In other words, the contribution of TAM, VEM, and trust to variable buying intentions is 59,3%. Meanwhile, the rest of 40,7% is other variables that are not analyzed in this study.

CONCLUSIONS
Some hypotheses proposed in the research are accepted (H2, H3, H4, and H7). There are also some unacceptable hypotheses that are H1, H5, and H6. TAM cannot contribute to consumers' trust. It happens because consumers have a low level of involvement in Facebook of JOKUL. Consumers only want to order from the seller who has an online stall that is in JOKUL Facebook and has an offline stall. They never make transactions in the place. For consumers of JOKUL Facebook, it is just as a reminder that the store is there. If the seller who uploads or posts the goods on JOKUL Facebook and the seller does not have an offline stall, no one will buy it. Despite the times, the merchant re-post the product. Therefore, the result of H5 that VEM has a positive influence on buying intention is automatically refused. Although the product shown on Facebook is very interesting, if consumers do not know where the seller is offline and the location is not clear, the consumers will not want to buy the product.
Although the community of JOKUL has social media besides Facebook like Instagram and website, only Facebook become the target of consumers to order product in JOKUL. It is because they are lazy to learn about other social media. They just believe that the product display in JOKUL Facebook is very interesting, but they are still reluctant to buy because the sellers in JOKUL Facebook do not have an offline stall and clear location. Ordering facility is only used by the consumer to order product and interact with the seller. The consumers still prefer to visit the seller by offline to take the product and pay. By that reason, H6 which states TAM contributes to buying intention through trust is refused.
For further research, the researcher can use the other theory of TAM such Theory of Reasoned Action (TRA) which specialize in model user acceptance or Theory of Planned Behavior (TPB). The researchers can also use the other factors that determine the actual behavior of the individual like the factor which indicates how strong a person wants to try or how much effort is planned to implement the business. It is because TAM cannot predict user acceptance of technology on JOKUL Facebook.
The weaknesses of this research are the respondents who are inadequate considering the numbers of members who enroll in JOKUL are 123.000 members. For further research, it is advised to use qualitative analysis methods to determine the reasons why consumers buy online considering the first hypothesis stating that TAM does not affect trust and the fourth hypothesis that VEM does not affect buying intention (H4 and H5). Meanwhile, the strength of this research can explain the understanding level of Jombang society toward the use of social media especially JOKUL Facebook.