Identifying the Factors Affecting MSMEs’ Intention to Buy Flavor B2 Replacer as Halal Seasoning Solution

Consumption of halal food is currently not only a concern of the Muslim community, but non-Muslim communities have also paid attention to halal products due to their high health orientation. Lard oil is one of the ingredients to improve the taste of food; however, besides being haram, lard contains quite a complex fat content, so all people with health problems cannot consume it. Various breakthroughs have been made by innovating new products, such as the development of artificial flavors that can produce taste characteristics similar to lard but are healthy and halal to consume. This article examines the factors that influence the interest in using food and beverage providers to replace lard with B2 replacer products based on halal awareness, halal food knowledge and perceived value with attitude as a mediating variable. By taking a sample of 372 MSMEs food and beverage providers in the DKI Jakarta area using a purposive sampling technique, this study found that perceived value is the most powerful factor influencing purchase intention, followed by the halal awareness factor. However, halal food knowledge was found to have no direct effect on purchase intention.


INTRODUCTION
Indonesia has an up-and-coming market for MSMEs industry players where they can take advantage of the potential of existing demographic bonuses to develop their businesses. Indonesia has the most micro, small, and medium enterprises (MSMEs) in the ASEAN region, as the ASEAN Investment Report reported. These MSMEs can absorb labor and contribute to the Gross Domestic Product (GDP) and also in national exports. The contribution reports said that it reaches 60.5%, and the employment of 96.9% as of the national workforce total absorption (Limanseto, 2022). In Indonesia alone, most MSMEs-scale businesses or industries are engaged in the food sector, with the number of MSMEs food sector reaching 1.51 million business units in 2020. The proportion of MSMEs in the food sector reaches 36% of all national MSMEs, totalling 4.21 million business units. The high number of MSMEs in the food industry has opened up opportunities for developing companies in the flavor provider industry. The flavor is one of the raw materials for minor food and beverage production. From year to year, the flavor industry has developed both in terms of sales and the number of product variations offered.
Several factors are triggering the current development of the flavor industry, especially in Indonesia, namely the increasing need for food consumption, the population of Indonesia, which is increasing every year, and the increase women in the workforce, so that the need for ready-to-eat food is also increasing. Rapidly pork oil is a flavor often used by non-halal restaurants as a food flavoring (Syarief, 2010). Pork oil is produced from lard, which has quite a complex fat content, so it produces a taste that seems fatty and has a thick flavor. However, lard can only be consumed by some market segments because of religious limitations. One of the breakthroughs that can be taken is by innovating new products by making artificial flavors whose characteristics are similar to lard oil. So that it can be expected to produce a delicious and halal food taste. The Muslim community and people with non-Muslim beliefs have also started to more aware in halal food and beverage products. Based on Wibowo et al. (2020), health benefits are one of the reasons for consuming halal food by the wider community.
This indicates a broad market potential for a product to innovate by focusing on flavor products as a substitute for lard. This new product innovation has become a business strategy for one of the flavor companies to enter the food service market segment. There is a need for research that provides a broad picture of the purchasing behavior of consumers, especially non-Muslims in Indonesia, towards marketing these new products. Several previous studies (Astuti & Asih, 2021;Bashir, 2019;Bashir et al., 2019;Setiawati & Syahrivar, 2019;Vizano et al., 2021) found the awareness of halal factor influences consumer intentions to buy. It is a conscious of each Muslim state in which they suffice the knowledge of sharia in Islam about what is halal, recognizes the proper process of executing, and also putting first the consumption of halal food. If they bear more with the halal concept, they will tend to be much more concerning in what to buy and consume in every product that commonly spread and sold in society. However, other research (Nurcahyo & Hudrasyah, 2017;Yu & Lee, 2019) found no effect in terms of purchase intention and halal awareness. So the involvement in those are still experiencing inconsistent findings.
Aside from halal awareness, halal food knowledge is another factor influencing buying interest (Astuti & Asih, 2021;Öztürk, 2022;Rektiansyah & Auwalin, 2022;Setiawati & Syahrivar, 2019). Consumer knowledge of halal food will provide the basic information they need as a decision-making factor (Wibowo, et al., 2020). Consumers need definitions and facts about subcategory products (Nurlatifah, 2021;Wibowo et al., 2019). It will cover various of products in some categories like attributes, terminology, evaluation of attributes, and the common information (Yu & Lee, 2019). However, several previous studies found no effect between halal food knowledge and consumer buying interest (Maulani et al., 2022;Wibowo et al., 2019). So for now, it accomplished that the involvement for halal food knowledge and purchase intention is still experiencing inconsistent findings.
The last factor that can potentially increase consumer purchase interest is perceived value (Aksoy & Basaran, 2017;Yoshihiro et al., 2019;Yu & Lee, 2019). Perceived value is frequently conceptualized as a consumer's usual evaluation of the application of a service or product primarily based totally on perceptions of what's obtained and what's given (Zeithaml et al., 1996). In this concept, what is received and given express benefits and sacrifices? However, Lomboan (2017) found no influence between perceived value and consumer buying interest. So the involvement of perceived value and purchase intention is still experiencing inconsistent findings. In addition to the three potential factors mentioned above, several previous studies tested attitude variables in mediating the involvement of halal awareness, halal food knowledge, perceived value, and purchase intention of halal consumers (Maulani et al., 2022;Rektiansyah & Auwalin, 2022;Setiawati & Syahrivar, 2019;Wibowo et al., 2019;Yu & Lee, 2019). For this reason, this research objectives to find out the factors influencing the interest in using food and beverage providers to replace lard with B2 replacer products based on halal awareness, halal food knowledge and perceived value with attitude as a mediating variable, as well as how much influence attitudes have on purchase intention and examine the role of mediation on attitudes towards new product innovation flavors "B2" (the best for booster) pork oil replacer as a new solution for halal seasoning.

LITERATURE REVIEW Halal Awareness
Awareness relates to the perceptions and responses of intellectual to conditions when they use, eat or drink (Nurlatifah, 2021). Awareness is an approach carried out to buy service or product which represents the first process of purchasing that a consumer inexperienced with it then becomes attach to it. Namely, awareness measures as informastion of selected situation or subject (Wibowo et al., 2020). Previous research (Astuti & Asih, 2021;Öztürk, 2022;Rektiansyah & Auwalin, 2022;Setiawati & Syahrivar, 2019) has confirmed that the awareness has a positive effect to consumer attitudes toward purchasing food products. As an additional, it positively affects consumer intentions (Astuti & Asih, 2021;Bashir, 2019;Bashir et al., 2019;Setiawati & Syahrivar, 2019;Vizano et al., 2021). Furthermore, attitudes mediate the involvement of awareness and consumer buying interest (Muzakir & Damrus, 2018;Setiawati & Syahrivar, 2019

Perceived Value
Perceived value is frequently conceptualized as a consumer's usual evaluation of the application of a service or product primarily based totally on perceptions of what's obtained and what's given (Zeithaml et al., 1996).
The beneficial perceived value elements contain the factors of extrinsic and intrinsic the product, perceived quality, and other relevant better levels of abstraction, like experience or prestige. In addition, the forfeit component of perceived value includes monetary and non-monetary costs like effort and time (Zeithaml et al., 1996). Earlier studies have determined a positive and significant involvement of perceived value and consumer buying interest in halal products (Aksoy & Basaran, 2017;Yoshihiro et al., 2019;Yu & Lee, 2019). In the other hand, the perceived value was also found to influence attitudes toward halal products (Yu & Lee, 2019). Furthermore, attitudes were found to mediating the perceived value and consumer buying interest (Yu & Lee, 2019). Thus, this study hypothesizes that: H3: perceived value positively affect purchase intention η2 = γ3ξ3 + ς2 ……………… (7) H6: perceived value positively affect attitude η1 = γ6ξ3 + ς1 ……………… (8) H10: attitude mediates the relationship perceived value and purchase intention η2 = γ3ξ3 + γ6ξ3 + ς2 ………. (9) Attitude is the level of consumer preference for a product (Akın & Okumuş, 2020). It is a psychological tendency that determines favorable or unfavorable evaluations for carrying out a behavior (Maulani et al., 2022). In this case, a strong and positive attitude shows that consumers' expectations are met and aligned with their motives (Bashir, 2019). In turn, the consumer's decision-making process will be simplified and increase the intention to bought a product. Thus, this research hypothesizes that: H7: Attitude positively affect purchase intention η2 = β1η1 + ς2 ……………. (10)

METHODS
This study was designed using a quantitative method. According to Saunders et al. (2019), the quantitative method uses research data in the form of numbers and analyzes it using statistics. Then, this is causality research, known as explanatory research, conducted to identify the level and nature of causal relationships. According to Saunders et al. (2019), causal research is study which describes the causal involvement of the independent (exogenous) and dependent (endogenous) variables, where this research consists of 3 independent variables, namely halal awareness (X1), halal food knowledge (X2) and perceived value (X3) with two dependent variables which are attitude (Y1) and purchase intention (Y2).
In addition, this research type is exploratory which aims to find patterns in the data assuming a lack of theoretical evidence or prior learning on the tested variables (Putra, 2022). Considering this is exploratory research, then the analysis method will use the PLS-SEM. The PLS beneficial is to perform predictive analysis (Fahmi, et al., 2022a;Fahmi, et al., 2022b;Kasmo et al., 2018;Putra, 2022). PLS-SEM is purpose to forecast the relationship for each construct with these predictions. As you can see in Figure 1 reveals the research model, and the sign (Putra & Ardianto, 2022) is ξ (exogenous latent); η (endogenous latent), β (path coefficient of endogenous variables to endogenous variables); γ (path coefficient of exogenous variables to endogenous variables); x (manifest measurement variable from latent exogenous variables); y (manifest measurement variable of latent variable); ς (remaining latent endogenous variables). Then, the form we propose is separated to two structural equation models in this way: This research was distributed to respondents who are MSMES food and beverage providers in the DKI Jakarta area and have knowledge of the product flavor "B2" (the best for booster) lard oil replacer. Respondents in the study obtained 372 respondents of 5159 total population with purposive sampling data collection technique, which considers the selection of respondents based on specific criteria to become samples from populations that are considered able to provide information needs (Andriani & Putra, 2019;Saunders et al., 2019). The following is the result of eliminating respondents who did not meet the sample criteria presented in Table 1. Based on the distribution of 413 questionnaires, only 410 (99%) of respondents they have met the first criterion, namely businesses that provide food and beverages. Furthermore, from a total of 410 remaining respondents, only 398 (97%) of respondents met the second criterion: business people who opened their businesses in the DKI Jakarta Region. Then, out of a total of 398 remaining respondents, only 390 (98%) met the third criterion, namely business people who had heard of the "B2" (the best for booster) replacer flavored pork oil, and only 372 (95%) respondents who used it. For this reason, the model in this study will be analyzed with 372 data that have been collected and meet the purposive sampling criteria. After we validated the fulfillment of the sample, the following is an analysis of the demographic distribution of respondents in this study which is divided into several categories such as gender, age, business ownership, a form of place ownership, business capital, business turnover, and business span establishment. Based on Table 2, it can be seen that the data in this study were dominated by 284 (76%) male respondents, with the most ages being in the range of 25 to 30 years, 188 (51%). The following characteristic is seen from the type of business ownership, where the most significant number of respondents run a family business of 161 (43%), with the most venture capital in the range of 50 to 75 million of 181 (49%) with an average turnover of at most above 50 million per month by 247 (66%). The last characteristic can be identified when the business has been established, most in the range of 3 to 10 years, with 221 (59%).
This research consists of three independent variables, namely halal awareness (X1), halal food knowledge (X2), and perceived value (X3), with two dependent variables, namely attitude (Y1) and purchase intention (Y2). For this reason, the measuring items in this study were adapted from several previous studies, where halal awareness (X1) was measured using 13 items adapted from Nurlatifah (2021). While the variable halal food knowledge (X2) is measured using eight indicators adapted from Wibowo et al. (2020). Furthermore, perceived value (X3) is measured using 16 items adapted from Jamal & Sharifuddin (2015). Thus there are a total of 37 measuring items used to measure the manifest exogenous variables. Meanwhile, for endogenous variables, attitude variables (Y1) and buying interest behavior (Y2) in this study were measured using each of the four indicators adapted from Setiawati & Syahrivar (2019).
This research begins by evaluating the normality of 372 data sets collected. The results of this study indicate that the critical value (skewness) is less than ±2.00, and the kurtosis value is not more than 7. It can be concluded that each measuring item in this study has good data normality. In addition, we have conducted a pilot test to conduct initial testing to evaluate the validity and reliability of 37 respondents. The pilot test results provide an initial overview for us to evaluate each research statement in order to avoid habits so that it is easy for potential respondents to understand.

RESULTS AND DISCUSSION Results
Before testing the research hypothesis to answer the formulation of the problems in this study, we first evaluate the measurement and structural models. The measurement model, commonly called the outer model, consists of an evaluation of convergent and discriminant validity followed by an evaluation of construct reliability. To start evaluating the measurement model, we evaluate convergent validity by looking at the significance of the outer loadings of each of the existing indicators. According to Hair et al. (2018), the value of outer loadings must be greater than 0.70. However, according to Sarstedt et al. (2019), if the external loadings value of the reflective indicator is above 0.50, it is recommended to save it if the construct reliability value meets the requirements. So the decision that can be taken for the minimum loading acceptance limit is 0.50. The results of the tests in Table 3 show that the outer loadings have values above 0.50 and a significant pvalue below the 0.05 threshold, so it can be said that all indicators in this study have been able to measure the intended variable or, in this case, are valid.
Another method to see convergent validity is to look at the square root of the average variance extracted (AVE) value which must be greater than 0.5; this limit indicates that the latent variable can take into account the variance of the indicator by more than 50%. This study found that each latent variable in this study was found to have been measured by indicators of more than 50%. Where halal awareness can be measured by 13 indicator items of 62.5% (0.625), then halal food knowledge can be measured by eight indicator items of 64.9% (0.649), perceived value can be measured by 16 indicator items of 74.2% (0.742), attitude is able measured by four indicator items of 81.6% (0.816) and purchase intention can be measured by four indicator items of 67.4% (0.674).
After the indicator items have been evaluated for convergent validity, the next step is to evaluate the reliability of each latent variable using Cronbach's alpha and the composite reliability of both rho_a and rho_c. This study obtained results showing the high reliability of each latent variable; this is shown in Table 3, where each variable has a Cronbach's alpha value above 0.60 and a composite reliability value for both rho_a and rho_c above 0.70 (Rinaldi & Putra, 2022).
The last step in the evaluation stage of the measurement model is the problem associated with discriminant validity for each construct with a correlation value between constructs in the model (Henseler et al., 2015) using the Fornell Larcker criterion. However, according to Henseler et al. (2015), the Fornell Larcker criterion approach failed to identify discriminant validity in the most significant cases. To that end, Henseler et al. (2015) suggested assessing convergent validity using the confidence intervals shown in Table 4.
The confidence interval value was obtained through a bootstrapping procedure with a re-sample of 5000 to obtain a confidence interval (CI) value of less than or equal to 1.00 to identify no problems with discriminant validity (Henseler et al., 2015). The results of this study found that the overall value of the existing correlation is below 1.00, both in the range of 2.5% and 97.5%. After the measurement model can be estimated and meets the testing criteria, the structural model is tested.
For this reason, the evaluation of the structural model begins by testing the inner VIF value. The test assesses multicollinearity and interprets path coefficients (i.e., collinearity between constructs) (Ramayah et al., 2018). The multicollinearity assumption is used to see no perfect or significant correlation between exogenous variables. The correlation value between the observed variables (VIF) cannot be more than 10 (Hair et al., 2018). The method used to test the occurrence of multicollinearity can be seen from the variable correlation matrix generated through the VIF value. The results of the tests in Table 5 show that the VIF value for each path correlation is below 10, namely in the range of 3.296 to 4.257, so it can be concluded that there is no multicollinearity between exogenous variables in this study.
After testing the inner VIF value, the coefficient of determination is evaluated to see the magnitude of the variance of the exogenous variables in explaining the endogenous variables. Table 5 shows that attitude (Y1) as an endogenous variable can be explained by 69.7% by three exogenous variables, namely halal awareness (X1), halal food knowledge (X2), and perceived value (X3). Furthermore, that buying interest (Y2) as an endogenous variable can be explained by 65.2% by four exogenous variables, namely halal awareness (X1), halal food knowledge (X2), perceived value (X3), and attitude (Y1). The coefficient of determination R-square (R 2 ) shows how much the exogenous variables explain the endogenous variables. The value of R-square (R 2 ) is zero to one. If the R-square value (R 2 ) gets closer to one, then the independent variables provide all the information needed to predict variations in the endogenous variables. Conversely, the smaller the Rsquare (R 2 ) value, the more limited the ability of the independent variables to explain variations in the endogenous variables. The R-square value (R 2 ) has a weakness; namely, the R-square value (R 2 ) will increase every time there is an addition of one exogenous variable even though the exogenous variable has no significant effect on the endogenous variable (Putra & Ardianto, 2022). Evaluation of the structural model is continued to evaluate the fit model. This test was carried out to test the model's suitability, which was carried out using two test models, including standardized root mean square residual (SRMR) and normal fit index (NFI). As stated by Ramayah et al. (2017), the model will be considered to have a good fit if the standardized root means square residual (SRMR) value is below 0.10 (Hair et al., 2014). The following suitability index is the normed fit index (NFI) with the calculation of the Chi2 value (Rinaldi & Putra, 2023). The final stage in evaluating the structural model is to test the predictive-relevance value using the PLSpredict technique. The aim is to strengthen the relevance of the model's predictions. Shmueli et al. (2019) suggest using the PLSpredict algorithm to predict the performance of the PLS model. PLSpredict is based on k-fold cross-validation, validating disagreement samples (Shmueli et al., 2019). SEM-PLS suggests using ten folds (k=10) and ten repetitions (r=10) in the prediction test of the PLS model. This study uses the root means a squared error of predictions (RMSE) because the data supports a symmetric distribution of prediction errors (Shmueli et al., 2019).

Discussion
After evaluating the measurement and structural models, the last stage is testing the hypothesis. This stage is intended to test the acceptance of the research hypothesis that has been proposed previously. This stage consists of three values to be evaluated, starting from evaluating (1) path coefficients, (2) T-Statistic values through bootstrapping procedures, and (3) pvalues. Hair et al. (2017) state that the path coefficient value will be from -1 to +1. A path coefficient value of more than 0,000 and close to +1 represents a strong positive relationship, and vice versa indicates a strong negative relationship. At the same time, T-Statistics (bootstrapping) is used to see the significant value between constructs. Hair et al. (2017) suggested the bootstrapping procedure with a re-sample value of 5,000. The limit for rejecting and accepting the proposed hypothesis is ±1.9659, and if the t-statistic value is above 1.965, then the hypothesis is accepted. Furthermore, the p-value represents the correlation between research variables, where a relationship can be influential when the pvalue is below 0.05; otherwise, the relationship will have no effect when the p-value is above 0.05.
Meanwhile, in testing the indirect relationship, it can be seen that attitudes have two mediating roles, namely (1) partial and (2) full. Where partially, attitudes mediate the influence between halal awareness (β = 0.081, t = 3.337, p = 0.001) and halal food knowledge (β = 0.102, t = 3.844, p = 0.000) on purchase intention. This effect was found to have no difference when compared to the direct relationship, where halal awareness and halal food knowledge were found to have an effect directly and indirectly on purchase intention. For this reason, it can be concluded that with or without an attitude, both halal awareness and halal food knowledge still affect purchase intention. Furthermore, attitudes were found to have a fully mediating effect on the effect of halal food knowledge (β = 0.163, t = 4.804, p = 0.000) on purchase intention. This effect was found to have a difference when compared to the direct relationship, where halal food knowledge was found to have no good direct influence on purchase intention. For this reason, it can be concluded that through attitudes, halal food knowledge can influence purchase intention.
Based on the results of the study, halal awareness has a direct influence on purchase intention; this shows that the higher the halal awareness of food and beverage providers for the new innovative product flavor B2 replacer (the best for booster) pork oil, it will increase their interest in using the product. To become one of their products. The results of this study are in line with research conducted by several previous industry studies (Astuti & Asih, 2021;Bashir, 2019;Bashir et al., 2019;Setiawati & Syahrivar, 2019;Vizano et al., 2021) and are in contrast to research which was carried out in cases of ready-to-eat instant noodles (Nurcahyo & Hudrasyah, 2017) and other halal industries (Savitri & Perwita, 2020). This shows that food and beverage providers are highly aware of halal, so that they will replace their food ingredients from lard with replacer products. Indonesia has one of the largest Muslim populations, so most of them have a solid sentiment to produce halal food products because they feel they have a big responsibility for the Muslim community as a whole (Tawil et al., 2015).
Subsequent findings found that perceived value has a direct influence on purchase intention; this indicates that the higher the product value perceived by food and beverage providers for the new innovative product Flavor B2 replacer (the best for booster) lard, it will increase their interest in using the product to become one of their products. This factor was also found to have the most significant influence compared to halal awareness and food knowledge. The results of this study are in line with research conducted by several previous industry studies (Aksoy & Basaran, 2017;Jayadi & Ariyanti, 2019;Yoshihiro et al., 2019) and are in contrast to several other studies (Maulani et al., 2022;Wibowo et al., 2019). This shows that food and beverage providers already have a high perceived value of halal food ingredients, so they will choose to replace their food ingredients from lard with replacer products. Food and beverage providers can also understand that lard is not only bad for Muslim communities but has an impact on health. This causes the importance of creating value for replacer products as a substitute for lard.
However, even though the halal awareness and perceived value factors are essential factors in increasing interest in using replacer products, the halal food knowledge factor was found to have no direct influence on purchase intention. This confirms the results of research conducted by Maulani et al. (2022) and Wibowo et al. (2019), who found no direct effect between halal food knowledge and direct purchase intention, but this result is in contrast to several other studies (Purwanto et al., 2021;Yoshihiro et al., 2019). This shows that the halal knowledge possessed by food and beverage providers has yet to be able to increase their interest in using replacer products. This finding is supported by the Islamic Finance Development Indicator Report (Refinitiv, 2022), in which the indicator of knowledge of Indonesian people occupies second place for three categories, namely halal education, halal research, and halal seminars compared to Malaysia. This is a concern for Indonesia's halal industry to increase the intensity of providing halal education, halal research, and halal seminars so that people have high knowledge of halal.
Subsequent findings in this study found that the three factors, halal awareness, halal food knowledge, and perceived value, could increase the positive attitude of food and beverage providers towards replacer products. Attitude is a person's way of looking at something from within themselves (Nugroho & Setiaji, 2018). Food and beverage providers are considered to have a positive perspective on pork oil replacer products. They are aware that all groups can consume replacer products without reducing the existing taste. In this study, the attitude of food and beverage providers has an important role, where this attitude helps them choose to replace pork oil with all the limited knowledge of their halal products.

Conclusion
This study found that the perceived value factor is the most significant in increasing purchase intention. Thus, producers of lard's new flavor, B2 replacer (the best booster), can focus on developing the value of their products to increase consumer purchase intentions effectively. Furthermore, this study found that the halal food knowledge factor was the most significant in increasing attitudes toward products. Thus, producers of the new innovative product Flavor B2 replacer (the best for booster) lard can provide detailed information about halal on the packaging and other product attributes to effectively increase positive consumer attitudes towards halal products. In addition, this study found that consumers' positive factors for halal products can change the direct influence of perceived value in increasing purchase intentions. Thus, producers of the new innovative product Flavor B2 replacer (the best booster) lard can focus on improving positive consumer attitudes to increase consumer purchase intentions effectively.
This study has several limitations that can be used as the main focus of further research. Psychological factors from food and beverage providers must be examined in further research, such as security, trust, self-trust, perceptions of selfcontrol, and subjective and social norms. Thus, the internal priority process for food and beverage providers to replace lard as a food ingredient can be explained more broadly. Furthermore, this research model can be used to investigate other consumers outside the DKI Jakarta area, considering that this research was only conducted in one area and cannot be generalized broadly. This model can also be used in other countries, such as Malaysia, which has a higher level of halal knowledge than Indonesia, even though its population is smaller than Indonesia's.