Can Instagram Attributes Affect Store Loyalty Mediated by Application Engagement and Interaction Intentions?

The amplifying importance of Instagram and its flourishing embracing proposes a massive prospect for firms to promote their products that deliver an endeavor to identify customers’ behaviors well. The research aims to recognize application attributes of perceived ease of use (PEOU), information quality, perceived usefulness, and system quality toward store loyalty from the Technology Acceptance Model (TAM) lens. Moreover, it elaborates on the theory of Planned Behavior (TPB) to construct interaction intention and application engagement as mediator variables that apply the conceptual model among application attributes and store loyalty. This study uses a quantitative approach where questionnaires are distributed to 272 consumers who know or follow the Insta-gram of the Coffee Shop in Salatiga, Indonesia. The outcomes expose which interaction intention and application engagement are essential in the linkages among perceived usefulness, system quality, and information quality toward the store. Conversely, both mediator variables could not mediate the linkages between PEOU and store loyalty. Theoretically, there is a synergy amid theory TAM and TPB, revealing that perceived usefulness is most influential as antecedents affect store loyalty through interaction intention and application engagement.


INTRODUCTION
Digital marketing has made businesses overgrow and made progressive and massive business transformations (Hachimi et al., 2021), including for Micro, Small, and Medium Enterprises (MSMEs).According to the Office of Cooperatives for Small and Medium Enterprises of Central Java Province (2020), 44.3% of MSMEs for the processing industry are located in Central Java, of which 9929 of them are MSME food and beverage processing which is one of the leading sectors in Indonesia and compared other types of businesses in category C, that industry has the highest GDP contribution (Susianto et al., 2019).The processing industry category has a minimum wage of 99.04% (BPS, 2017); especially for coffee shops, the number reached more than 2,950 outlets in 2019 in Indonesia, an increase three times compared to 2016, when only 1,000 outlets (Humas Jateng, 2020).The results of an increase in MSMEs in the coffee sectors are directly proportional to the need for consumption, reaching 4,800 bags in 2018-2019 based on International Coffee Organization (ICO).Coffee shops have consumers who use social media which they use it to obtain information and interact with fellow consumers (Databoks. metadata.co.id, 2020).We Are Social & Hootsuite (2022) stated that 191.4 million Indonesians have used social media.It indicates that digital marketing penetration is relatively good, around 68.9% of the total population.Among the few social media available today, Instagram Is seeing a steady increase in active users.Therefore, its adoption is also growing among wellknown businesses, which motivates community development on Instagram.
This study attempts to identify several factors that influence store loyalty through the lens of the Theory of Planned Behavior (TPB) with the official Instagram account of a store.TPB by Ajzen (1991) states that behavioral intentions drive a person's behavior.McLean (2018) adds that consumer attitudes can be considered a collection of individual beliefs about a shop, so a consumer's active engagement with the application can influence loyalty.The research focuses on the attributes of the Instagram application, which refers to the specific characteristics or qualities of a software application that define its behavior, functionality, and performance.Numerous studies have examined consumer behavior in accepting and using technology based on the Technology Acceptance Model (TAM) approach (Jiang et al., 2013).TAM positions two primary constructs perceived ease of use and usefulness (Chau & Lai, 2003).Besides that, the DeLone and McLean Information Systems Success Model (D&M IS Model) is a framework used to evaluate and understand the influence of information systems on organizational performance, which brings information quality and system quality variables for attributes of the Instagram application.
These attributes determine how well the application meets the needs of its users and how effectively it performs its intended functions.It consists of behavior beyond buying and results from inner motivation, which ultimately directs users to take specific actions focused on the coffee shop.The willingness of consumers to take action in the form of interactions with stores on an ongoing basis is loyalty, and maintaining relationships with consumers will impact consumer loyalty to the store (LS).The ease attribute of Instagram consists of perceived ease of use (PEOU) and perceived usefulness (PU) that can affect store loyalty (SL).It is supported by research that found a significant effect between the PEOU and LS (Keni, 2020;Wilson et al., 2021), while studies by Harjanti (2017) show that PEOU perspectives do not affect loyalty, where consumers who agree to feel the PEOU of a website do not guarantee that they will become loyal consumers.Another thing, PU was found to have a significant effect on loyalty (Irfansyah, 2021;Maryanto & Kaihatu, 2021), inversely proportional to research by Bahari et al. (2018), where PU has no significant effect on loyalty because it does not differentiate results by examining the responses given by highly experienced and less experienced website users.Based on the arguments, there is an empirical gap in the relationship between the effect of Instagram application attributes on store loyalty.
Moreover, the quality content and Instagram application like information quality (IQ) and system quality (SQ) increase customers' loyalty to the coffee store.It is aligned with other empirical studies, namely the IQ, has a significant effect on loyalty (Handayani et al., 2021;Putri & Pujani, 2019), while (Vikramaditya, 2021)'s research found that the quality of information has an insignificant negative effect on loyalty, which means the quality of application information has not been able to influence the level of consumer loyalty.In addition, system quality significantly affects loyalty (Assegaff & Pranoto, 2020;Fitri & Raflah, 2021), unlike Kurniawan et al. (2021).who found that the quality of information systems does not indirectly affect consumer loyalty.Thus, there are research gaps between IQ and SQ toward store loyalty.
To answer the research gap above, engagement with the application and interaction intentions is needed to mediate the effect of Instagram attributes on store loyalty.Consumer loyalty to a store can be activated in several different forms through interaction with the store (Dessart et al., 2015), so in 2011, Kim & Adler conceptualized the potential of mobile applications for further engagement with stores.By looking at these things, even though there is much research on store loyalty, it cannot be found looking at loyalty through several Instagram attributes using the TAM and TPB theories.Thus, this study will analyze how the influence of Instagram application attributes on loyalty to a coffee shop in Salatiga.
Despite this, limited academic research has explored the attribute factors influencing customer engagement with an Instagram application and the interaction intention to increase loyalty stores.Thus, while customers are adopting Instagram applications as a channel for getting content information, coffee stores must be aware of the factors influencing customer engagement with an Instagram application and the subsequent outcomes for the interaction intention.Consequently, this research proposes to tackle these gaps in comprehension.This research aims to recognize Instagram attributes that are PEOU, PU, IQ, and SQ toward store loyalty from the lens of the TAM.Moreover, it elaborates on the TPB to construct interaction intention and application engagement as mediator variables that apply the conceptual model among application attributes and store loyalty.

Hypothesis Development
The Effect of Perceived Ease of Use PEOU is about the degree to which the users believe the technology will be easy to learn and use (Davis, 1989).PEOU is essential in users' intention to adopt new technologies or systems.The easier a technology or system is perceived, the more likely users intend to use it.
Besides that, when a product or service is easy to use, consumers are likelier to engage with it and use it more frequently.PEOU affects engagement with the Instagram application (McLean, 2018), while Yunita et al. (2020) revealed that it does not significantly affect consumer engagement.
If consumers have the intention and involvement in ease of use, it can lead to store loyalty.PEOU significantly affects loyalty (Wilson et al., 2021), in contrast with research by Marso (2023) found that loyalty has no positive effect on loyalty.Thus, the proposed hypothesis is as follows: H1a-b: Perceived ease of use significantly affects (a) interaction intention; and (b) Instagram engagement.H1c-d: Perceived ease of use has a significant indirect effect on store loyalty through (c) interaction intention; and (d) Instagram engagement mediation.
The Effect of Perceived Usefulness PU refers to the extent to which a user believes technology will help them perform a task more effectively or efficiently (Park, 2009) which included e learning selfefficacy, subjective norm, system accessibility, perceived usefulness, perceived ease of use, attitude, and behavioral intention to use e-learning, was developed based on the technology acceptance model (TAM).The perceived usefulness of technology is referred to as one of the most important constructs in influencing the adoption of the latest technology.Previous studies have empirically investigated whether perceived usefulness positively affects purchase and repurchase intention (Keni, 2020;Isma et al., 2021;Wilson et al., 2021) found that perceived usefulness has no significant effect on purchase intention and behavioral intention to use.
McLean ( 2018) revealed that perceived usefulness influences engagement with mobile applications.Conversely, other research shows that perceived usefulness does not significantly affect consumer engagement (Yunita et al., 2020).PU significantly affects loyalty (Maryanto & Kaihatu, 2021;Wilson et al., 2021) business competition is ubiquitous.The research analyzed the impact of perceived usefulness and customer satisfaction on customer loyalty in Grab users.It also studied the impact of perceived usefulness on customer satisfaction and the customer satisfaction (moderated by perceived ease of use, but another research by Harianto & Ellyawati (2023) found no positive effect between PU and loyalty.The Effect of Information Quality Information quality focuses on how much information can consistently meet the requirements and expectations of all who need the information provided in the application (Urbach et al., 2010).The quality of information is measured by accuracy, timeliness, completeness, relevance, understanding, accessibility, and consistency according to the desired characteristics, whereas the quality of information measures semantic success (Delone & Mclean, 2003;Petter et al., 2013).Jiang et al. (2021) explained that the quality of information (quality of content, expression, and utility) significantly increases the intention to adopt information.
Research by Masri et al. (2020) explained that the quality of information has a significant relationship with sustainable intentions.However, other studies have found that the quality of information has a negative but insignificant effect on the repurchase intention of system users (Fachri et al., 2021).
Information quality has a significant positive relationship with customer engagement (Gosain et al., 2019).Information quality positively influences consumer engagement (Islam & Rahman, 2017).
Information quality significantly affects loyalty (Kurniawan et al., 2021;Mahendra et al., 2021;Amin & Chandra, 2022;Patma et al., 2023) The Effect of System Quality System quality focuses on measuring the system application itself, taking into account performance characteristics, functionality, and usability so that the sytem's quality can be considered as the extent to which the system is easy to use in completing work (Urbach et al., 2010).System quality is measured in terms of perceived ease of use, response time, functionality, reliability, flexibility, data quality, portability, stability, integration, and importance of an information system (Delone & Mclean, 2003;Wu & Wang, 2006).
System quality is significantly related to sustainable intentions and intention to use (Masri et al., 2020;Alkhawaja et al., 2022) consumers can plan their trip without time and space limitations.This study proposes a model regarding the for-mation of the relationship quality (customer satisfaction and trust.System quality does not significantly affect user repurchase intentions (Fachri et al., 2021).
System quality positively affects consumer engagement (Islam & Rahman, 2017).System quality affects the influence of consumer engagement, where a sound system quality is predicted to make users feel that the application is a helpful platform for social interaction and can encourage them to keep using the application (Busalim et al., 2019).

Engagement and Interaction Intention with the Instagram Application
Engagement with the application increases consumer loyalty to stores, and then after continuous retention, the effect of engagement with the application on loyalty becomes stronger (McLean, 2018).Jayasingh (2019) found that consumer engagement significantly increases store loyalty, where store engagement leads to positive store loyalty, usually in the form of consumers advocating for the store on social media.Consumer engagement positively affects consumer loyalty (Agyei et al., 2020).H5a: Instagram Engagement has a significant effect on store loyalty.
The main factor of TPB is the intention of the person to perform a specific action.The intention is how much effort is made or an indication of how hard some-one is trying to behave.Intentions can turn into behavior when the individual concerned controls the behavior.Interaction intention refers to the individuals' goals or objectives when they engage in communication or social interaction with others (Casaló et al., 2017).Interaction with consumers affects brand loyalty (Kholis & Ratnawati, 2021).H5b: Interaction intention has a significant effect on store loyalty.

METHOD
This study uses a quantitative method by collecting primary data by distributing online questionnaires using google forms.The sampling method used is nonprobability sampling, namely purposive sampling, by selecting the criteria for respondents who are consumers of the coffee shop and know or follow the Instagram of the coffee shop in Salatiga, the name is Coffee Tanem.The questionnaire is divided into three parts; in the first part, respondents' criteria were agreed to become participants.The following section contains the demographic statistics of respondents, including age, gender, education, and coffee shopping expenses in a month.The final section contains measuring variables and indicators with 39 question items using a 5-point Likert Scale, "1 = totally disagree" to "5 = strongly agree".
Based on the Slovin method, with a population (N) of 599 populations, with an estimated error rate of 5%, the minimum number of samples (n) is = 239,83 rounded up to 240 coffee shop Instagram users.According to Kelley et al. (2003) state that the acceptance rate of a good sample response rate must be above 45%.Thus, the minimum sample data target, previously 240 respondents, was increased to 272 to provide more accurate and precise results.
Data were analyzed using Partial Least Square-Structural Equation Modeling (PLS-SEM) with SmartPLS software version 3.0.The use of SmartPLS to apply SEM to predict or explain the construct or latent variable that is the target.PLS is causal modeling approach that aims to maximize the variance of the criterion latent variable, which predictor latent variables can explain.SEM analysis consists of confirmatory factor analysis (CFA), and regression analysis/path analysis, which can also be used to examine the validity and reliability of research instruments (Hair et al., 2009).

RESULT AND DISCUSSION
The results of descriptive research there is a research sample of 272 respondents.The following are the five characteristics of respondents used in this study: active account, gender, age, education le-Figure 1. Research Framework vel, and amount of spent on coffee in a month.The characteristics of the respondents in more detail can be seen in Table 1.

Confirmatory Factor Analysis (CFA)
The CFA test aims to test whether the indicators grouped based on their latent variables remain consistent in their constructs so that the model developed is by theoretical framework.According to Gholami et al. (2013); Islam & Rahman (2017) that the measurement of latent variables through convergent validity and reliability is used to test the value of loading factor, average variance extracted (AVE), composite reliability is used to test the value of loading factor, average variance extracted (AVE), composite reliability, and Djikstra-Henseler rho (ρA) higher than 0.60 (Ghozali & Latan, 2015).Testing the reliability and validity of each latent variable is explained in Table 2 and Table 3

Discriminant Validity
Test using correlation values between variables from the Heterotrait-Monotrait ratio (HTMT) it was found that all correlations between variables did not exceed the cut-value of 0.90 (Henseler et al., 2015), so that the seven variables can be ascertained with discriminant validity.According to Henseler et al. (2015), the method using the HTMT ratio is more effective in testing discriminant validity when compared to the criteria from Formell-Larcker.Complete results can be seen in Table 4.   and comments, many users tend to interact; then user preference is also an essential factor; a user will be more interested in interacting with this type of content or accounts that suit their interests, even though the application is the ease of use.
The first-b hypothesis stated that the PEOU could not significantly effects Instagram engagement.That is because of competition and noise on Instagram, where this platform is viral with millions of users and content posted every day, which makes it difficult for content to stand out and get high engagement, especially when brands or coffee shop accounts are not yet top of mind when users use Instagram.In this regard, Instagram continues to change its algorithm for displaying content to users, which affects how content is displayed in user feeds and determines wherever that content will be actively viewed by users so that sometimes these changes will reduce the reach and engagement of content.This finding aligns with previous research by Yunita et al. (2020).
The first-c hypothesis states that PEOU does not significantly impact store loyalty mediated by interaction intention.The results align with previous research by Marso (2023) concerning PEOU on loyalty e-commerce customers and a study by Suryani & Ramdhani (2022) about PEOU not having significant effects on II.It is known that the interaction intention is a connecting variable between PEOU and store loyalty, which means that the higher the level of PEOU with an interaction intention, the higher the interest instore loyalty.
The first-d hypothesis states that PEOU does not significantly impact store loyalty mediated by engagement with the Instagram application.The research results align with previous research by Marso (2023) concerning the influence of PU on loyalty in the TikTok Shop.Yunita et al. (2020) about PEOU has no significant effects on EIA.It is known that engagement with the Instagram application is a connecting variable between PEOU and store loyalty, which means that the higher the level of PEOU with an interaction intention, the higher the interest in-store loyalty.
The second-a hypothesis stated that the perceived usefulness significantly increases interaction intentions.The Instagram platform provides various communication benefits, including sharing and expressing oneself and engaging visual content.Instagram provides various communication features, such as direct messages, comment fields, and live streaming features that allow users to interact with others.This finding is in line with previous research by Keni (2020); Octavia & Baridwan (2020); Isma et al. (2021).
The second-b hypothesis stated that perceived usefulness significantly affects Instagram engagement.Instagram's functionality is easy to use and intuitive, with an easy-to-navigate interface, so it motivates users to explore the various features available and continue to interact on the platform.Users tend to be more satisfied with the experience of using Instagram, thus encouraging users to be more actively involved in various activities on the platform, such as uploading photos or videos, uploading stories, commenting, and liking another user's content.The research findings are consistent with previous research by McLean (2018).
The second-c hypothesis stated that PU significantly affects store loyalty mediated by interaction intention.The result of the research carried out is in line with previous research by Wilson et al. (2021) concerning the role of PU in influencing loyalty, research by Keni (2020) about PU has significant effects on II, and a study by Kholis & Ratnawati (2021) that found II has significant effects on SL.It is known that the interaction intention is a connecting variable between PU and store loyalty, which means that the higher the level of PU with an interaction intention, the higher the interest in-store loyalty.
The second-d hypothesis stated that PU significantly affects store loyalty mediated by engagement with the Instagram application.The results of the research carried out are in line with previous research by Wilson et al. (2021) concerning the role of PU in influencing loyalty, research by McLean (2018) about PU has a significant effect on EIA, and a study by Agyei et al. (2020) that found EIA has significant effects on SL.It is known that Instagram engagement is a connecting variable between PU and store loyalty, which means that the higher the level of PU with an EIA, the higher the interest in-store loyalty.
The third-a hypothesis stated that information quality significantly affects interaction intention.The content presented is information that is relevant to the user's interests or needs, by the context of the platform, original, and has the credibility of the sender of the information where the account that sent the information has authority or expertise in specific topics, users tend to trust that information more and more allows to interact with the content.These findings support the research by Masri et al. (2020) and Jiang et al. (2021).
The third hypothesis stated that information quality does not significantly effects Instagram engagement.That is because the coffee shop accounts do not use call-to-action (CTA) methods on their content and do not respond or engage with the comment section, only more res-ponses on repost stories.That keeps users from engaging even if the content is excellent; most followers only engage by liking the posts.This finding differs from previous studies by Islam and Rahman (2017) and Gosain et al. (2019).
The third-c hypothesis states that IQ significantly affects store loyalty mediated by interaction intention.The results of the research carried out are in line with previous research by Amin & Chandra (2022) concerning the effect of information quality on customer loyalty, research by Jiang et al. (2021) about IQ affects II, and studies by Kholis and Ratnawati (2021) that found II has significant effects on SL.It is known that the interaction intention is a connecting variable between IQ and SL, which means that the higher the level of IQ with and II, the higher the interest instore loyalty.
The third hypothesis states that IQ has no significant effects on store loyalty mediated by engagement with the Instagram application.The results of the research study differ from previous research by Kurniawan et al. (2021) about IQ affecting loyalty, and Gosain et al. (2019) found that IQ has a sign with EIA.It is known that Instagram engagement is a connecting variable between IQ and store loyalty, which means that the higher the level of IQ with an EIA, the higher the interest in-store loyalty.
The fourth-a hypothesis stated that system quality significantly affects interaction intention.It is due to the availability of features that function correctly and meet user needs, the speed, and smoothness of the platform's response, the practical and accurate quality of content filtering so that users see exciting and relevant content, protection of user privacy and security, and a pleasant user experience when having a quality Instagram system will make users more likely to have the in-tention to interact on Instagram actively.This finding aligns with previous studies by Masri et al. (2020) and Alkhawaja et al. (2022).
The fourth hypothesis stated that system quality significantly affects Instagram engagement.The performance of the Instagram system runs smoothly and responsively; the system is stable and free of technical glitches, is safe, and provides exciting features, so the system's quality plays an essential role in influencing the level of user Instagram engagement.This finding aligns with previous research by Busalim et al. (2019).
The fourth-c hypothesis states that SQ significantly affects store loyalty mediated by interaction intention.The results of the research carried out are in line with previous research by Mahendra et al. (2021) concerning the effect of SQ on customer loyalty, research by Alkhawaja et al. (2022) about SQ has significant effects on interaction intention, and study by Kholis & Ratnawati (2021) that found II has significant effects on SL.It is known that the interaction intention is a connecting variable between SQ and store loyalty, which means that the higher the level of SQ with an interaction intention, the higher the interest in-store loyalty.
The fourth hypothesis states that SQ significantly impacts store loyalty mediated by engagement with the Instagram application.The results of the research carried out are in line with previous research by Mahendra et al. (2021) concerning the effect of SQ on customer loyalty, research by Busalim et al. (2019) about SQ has a significant effect on EIA, and study by Agyei et al. (2020) that found EIA has significant effects on SL.It is known that the EIA is a connecting variable between SQ and store loyalty, which means that the higher the level of SQ with an EIA, the higher the interest in-store loyalty.
The fifth-a hypothesis stated that Instagram engagement has significant effects on store loyalty.Each user actively engages with content through liking or following on Instagram will make users feel emotionally connected to that content.Then users can interact with others via comments or admin via direct messages, where when users feel involved in conversations or get responses from account owners or other users, users feel valued and socially connected so that they can increase their loyalty to the coffee shop account.Another thing is that when account owners consistently update and upload relevant and exciting content, users stay engaged and loyal to the account.Consistency in delivering quality content can build positive expectations from users and make them want to return for new content.The results of this study support the findings by and McLean (2018), Jayasingh (2019), and Agyei et al. (2020).
The fifth-b hypothesis stated that interaction intention significantly affects store loyalty.Excellent and sincere interaction intentions play a crucial role in shaping follower loyalty to Instagram accounts; account owners show good intentions, provide quality content, and interact positively with followers to strengthen relationships and build loyal communities.The result of this study supports the findings of Kholis and Ratnawati (2021).

CONCLUSION AND RECOMMENDATION
The results reveal interaction intention and application involvement that are important in the relationship between perceived usefulness, system quality, and information quality on store loyalty.Conversely, the two intermediary variables cannot mediate the relationship between perceived ease of use and store loyalty.Due to the frequent occurrence of Insta-gram updates, the interface can change at any time, which makes users who are not used to fast changes have to try to keep adapting; this will be very difficult for people who are not tech-savvy.Furthermore, Instagram engagement cannot mediate between information quality and store loyalty because some users feel the shared content needs unique value and the latest information about promotions or discounts.
Perceived usefulness is the most influential on store loyalty through interaction intention and Instagram engagement mediation.This is realized by admins consistently sharing content on Instagram 2-3 times a week and frequently re-stories user stories so that consumers can find helpful information quickly via Instagram.With easy and intuitive functionality, they can make it easier for users to explore its features to find and interact with shared information.
The practical implication involves using CTA for the engagement content, so users will have interaction intention and be more engaged with their Instagram account afterward.Limitations in this study, some hypotheses still need to be more significant.Recommendation for future research can retest that variable again, which is PEOU with II, EIA, and SL; IQ with engagement for direct or indirect effect and can use another place for the study case.Future researchers may focus on this point.
H2a-b: Perceived usefulness significantly affects (a) interaction intention; and (b) Instagram engagement.H2c-d: Perceived usefulness has a significant indirect effect on store loyalty through (c) interaction intention; and (d) Instagram engagement mediation.
. H3a-b: Information quality from Instagram significantly affects (a) interaction intention; (b) Instagram engagement.H3c-d: Information quality from Instagram has a significant indirect effect on store loyalty through (c) interaction intention; and (d) Instagram engagement mediation.
. H4a-b: System quality from Instagram significantly affects (a) interaction intention; and (b) Instagram engagement.H4c-d: System quality from Instagram has a significant indirect effect on store loyalty through (c) interaction intention; and (d) Instagram engagement mediation.

Figure 2 .
Figure 2. The Results of Full Model Analysis Source: Data processed through SmartPLS 3.0 (2023)

Table 2 .
. Outer Model Test Results Source: Data processed through SmartPLS 3.0 (2023) Note: *the indicator not use in path analysis test because the number of outer loading <0.60).

Table 3 .
Reliability Test