The Mediating Role of Brand Awareness in Social Media Marketing’s Impact on Brand Image

1. Introduction

The culinary business has become the pillar of the Indonesian creative industry. Statistical Central Agency records that this sector contributed 34% of the gross domestic product in 2023. Additionally, this culinary business has a high opportunity because of the prominent market and potential profits (Purwanto & Pancawati, 2024). As one of the business types, a coffee shop functions to sell foods and beverages. Besides, it provides a place for relaxing, working, and networking (Ferreira et al., 2021), preferred by the millennial generation in the United States (Halff, 2016) and Generation Y and Z in Indonesia (Poerwanto, 2019).

In Bandung City, these coffee shops are utilized by higher-education students to drink coffee while doing their tasks from lecturers (Hanifah, 2024). Based on One Indonesian Data (2021), they have existed in this city since 2016, totaling 14 units, and this number in 2017 was still the same. For 2018 and 2019, the number was 41. Then, this number became 50 and 99 units in 2020 and 2021. According to Baye and Price (2022), the more companies in the same business sector, the more competitive the market is.

One of the strategies to promote the coffee shop is through social media (Getaruelas, 2019; Grossman & Rachamim, 2024; Park & Namkung, 2022), such as Instagram (Getaruelas, 2019; Park & Namkung, 2022; Syehabuddin et al., 2022), Facebook (Getaruelas, 2019; Ibrahim et al., 2021; Syehabuddin et al., 2022), Twitter (Arhas et al., 2022; Getaruelas, 2019; Syehabuddin et al., 2022), TikTok (Sekarwangi et al., 2023), and YouTube (Getaruelas, 2019). Specifically, Kopitagram, one of the coffee shops in Bandung, utilizes Instagram (Purwanti, 2024) to promote Asian foods and beverages, including Indonesian and European ones (Permana & Hasbi, 2023). Instagram utilization can elevate its brand awareness and image among the competition with other coffee shops (Park & Namkung, 2022).

Some studies attempt to relate social media marketing to brand awareness (Ayuningtyas & Sijabat, 2022; Bilgin, 2018; Faisal & Ekawanto, 2022; Lim et al., 2024; Malarvizhi et al., 2022; Park & Namkung, 2022; Purwianti & Fionna, 2023; Salsabillah et al., 2024). Unfortunately, their outcome is contradictory. For instance, Bilgin (2018), Faisal and Ekawanto (2022), Malarvizhi et al. (2022), Park and Namkung (2022), Purwianti and Fionna (2023), Lim et al. (2024), Salsabillah et al. (2024) demonstrate a positive tendency. However, Ayuningtyas and Sijabat (2022), Malarvizhi et al. (2022), and Lim et al. (2024) document a pointless propensity.

Equally, some investigations try associating social media marketing with brand image (Bilgin, 2018; Faisal & Ekawanto, 2022; Hidayatullah et al., 2021; Malarvizhi et al., 2022; Park & Namkung, 2022; Salsabillah et al., 2024; Yanti & Muzammil, 2023). Regrettably, their outcome is opposing. For example, Bilgin (2018), Faisal and Ekawanto (2022), Park and Namkung (2022), Malarvizhi et al. (2022), Yanti and Muzammil (2023), and Salabillah et al. (2024) show a positive tendency. Unfortunately, Hidayatullah et al. (2021) and Malarvizhi et al. (2022) exhibit a meaningless inclination.

Although demonstrating a positive relationship between brand awareness and image, this related research is few, such as those conducted by Bilgin (2018) and Salsabillah et al. (2024). A similar situation happens with the studies proving brand awareness to mediate the association between SMM and brand image, like Bilgin (2018) and Salsabillah et al. (2024).

Based on these inconsistent proofs of the relationship of social media marketing with brand awareness (BA) and image (BI), the infrequent tendency of BA toward BI, and the rare studies proving the mediating effect of BA on the impact of SMM on BI, this study aims to test these relationships based on the perspective of Kopitagram customers in Bandung City. In this city, according to a Google Map, Kopitagram is on Cianjur Street No.3, Kacapiring Sub-Distric, Batununggal District. Using consumers of coffee shops in this investigation differs from Bilgin (2018), employing the customers of the top five brands; Faisal and Ekawanto (2022) and Ayuningtyas and Sijabat (2022), studying the customers of entrepreneurial products; Hidayatullah et al. (2021), learning about the tourists purchasing mementos; Purwianti and  Fionna (2023) investigating the shoppers of fashion goods; Yanti and Muzammil (2023) with the customers of skincare; Malarvizhi et al. (2022), employing the users of portable tech gadgets; Lim et al. (2024) and Salabillah et al. (2024), surveying cosmetics customers.

Additionally, this study uses the structural equation model (SEM) based on covariance as Bilgin (2018), Faisal and Ekawanto (2022) and Park and Namkung (2022) utilize, differing from Hidayatullah et al. (2021) and Lim et al. (2024) employing the regression model, Ayuningtyas and Sijabat (2022), Malarvizhi et al. (2022), Purwianti and Fionna (2023), Yanti and Muzammil (2023), and Salsabillah et al. (2024) applying variance-based SEM. Indeed, this study is relevant for coffee shop owners or managers using social media to make customers aware of their brands and have a strong image in the competitive market by innovating their goods and services. 

2. Literature Review and Hypothesis Development

Social media is the device for the marketing strategy utilizing the online platform to interact with consumers directly to enlarge the brand radius and awareness (Dwivedi et al., 2021). In their study, Malarvizhi et al. (2022) utilize five dimensions of social media marketing, i.e., entertainment (ENT), interactivity (INT), trendiness (TREND),  customization (CUST), and electronic word of mouth (E-WOM) to be associated with brand awareness. As a result, ENT, TREND, CUST, and E-WOM positively affect BA, but INT does not. Using the exact dimensions, Lim et al. (2024) demonstrate that CUST, INT, and E-WOM positively influence BA. Unfortunately, ENT and TREND do not. Through their investigation, Faisal and Ekawanto (2022), Park and Namkung (2022),  Purwianti and Fionna (2023), and Salsabillah et al. (2024) demonstrate that marketing through social media marketing can elevate the brand awareness of the consumers. Based on this confirmation, the first hypothesis is like this:

H1: The relationship between social media marketing and brand awareness is positive.

In their study, Malarvizhi et al. (2022) utilize five dimensions of social media marketing, i.e., entertainment (ENT), interactivity (INT), trendiness (TREND), customization (CUST), and electronic word of mouth (E-WOM) to be associated with brand image. As a result, TREND, CUST, and E-WOM positively affect BA, but ENT and INT do not. Through their investigation, Faisal and Ekawanto (2022) and Park and Namkung (2022) reveal that marketing through social media positively creates the brand image. Similarly, Yanti and Muzammil (2023) and Salabillah et al. (2024) affirm that using social media with engaging content will make the customer always remember the brand. Based on this confirmation, the second hypothesis is like this.

H2: The relationship between social media marketing and brand image is positive.

Brand awareness is the salient reflection of the brand in the customer's mind and has six phases: (1) Consumers recognize the brand, (2) they can recall it, (3) make the brand in their top of mind, (4) mention the brand when asked by others, (5) understand what the brand is for, and (6) have the opinion (Aaker, 1996). In their study, Bilgin (2018) and Salsabillah et al. (2024) demonstrate that brand awareness positively affects brand image. Also, they prove that brand awareness can mediate the association between social media marketing and BI. Based on this evidence, this study proposes the third and fourth hypotheses like this:

H3: The relationship between brand awareness and image is positive.

H4: Brand awareness mediates the association between social media marketing and brand image.

3. Research Methods

This investigation has three variables. The first is social media marketing, which acts as the exogenous, where its items are from Park and Namkung (2022). The second and third are brand awareness and image, and their indicators come from Park and Namkung (2022). Furthermore, the details are in Table 1.

Table 1: The indicators of social media marketing, brand awareness, and brand image

Source: Park and Namkung (2022)

This study uses purposive sampling by setting some criteria when taking the samples, as Hartono (2014) explains. The criteria are as follows. Firstly, customers must domicile in Bandung. Secondly, they access Kopitagram's Instagram account before buying the products. Finally, they must purchase Kopitagram-related products for at least three months. Based on these criteria, this study uses the survey to collect the responses with a five-point Likert scale, as Sugiyono (2022) explains. For theoretical verification, the samples collected are at least 200 (Ghozali, 2021b; Mariana et al., 2024).

Because of the survey, validity and reliability examination is needed (Mariana et al., 2021). Confirmatory factor analysis is utilized when testing validity by comparing the loading factor (LF) with 0.5. When LF exceeds 0.5, a valid response occurs (Ghozali, 2017). Composite reliability and Cronbach Alpha are employed to detect reliable responses by comparing them with 0.7. A reliable response happens if they are above 0.7 (Junaidi, 2021).

Following Park and Namkung (2022), this study uses covariance-based SEM with the AMOS program to analyze the data, where the model is in the first and second equations:

        BA = γ1SCM + ξ1 (Equation 1)

        BI = γ2SCM + β1BA +  ξ2  (Equation 2)

When using the CBSEM, some measures exist to detect the goodness of fit: Chi-square divided by degree of freedom (CMIN/DF) (Dash & Paul, 2021), root mean square residual (RMSR) (Junaidi, 2021), parsimonious goodness of fit indexes (PGFI), parsimonious normed and comparative fit index (PNFI and PCFI) (Dash & Paul, 2021), where the cut-off points exist in Table 2.

Table 2: The cut-off point of some goodness of fit model measures

Finally, the first, second, third hypotheses are examined by comparing the critical ratio probability with a 5% significance level. Each hypothesis is acceptable if the probability is lower than 5% (Junaidi, 2021). For the fouth hypothesis, this study uses Sobel test, as suggested by Ghozali (2021a), by comparing the probability (2-tailed) of Z-statistic for Sobel with 5%. The mediating effect exist if this value is below 5% (Sahabuddin & Hadianto, 2023).

4. Results

4.1. Respondent Features

Based on a survey in July 2024, this study obtained 203 respondents, 100% all of whom meet the criteria, namely giving consent to fill out the questionnaire, having seen the Instagram account of Kopitagram in the past two years, and residing in Bandung. Table 3 displays the features of respondents buying coffee and its associated products in Kopitagram at least three times a month:

a.           The foremost respondents are male (51.23%), from 20 to 29 (81.28%), and undergraduate students (75.37%) based on gender, age, and current schooling level. Based on the occupation, the most is a private employee (45.32%).

b.          At the minimum, the respondents are female (48.77%), from 40 to 49 (2.56%), and senior high school (7.39%) based on gender, age, and current schooling level. By mentioning the occupation, the influencer (0.49%) and senior high school student (0.49%) are the least.

Table 3: The features of the respondents

4.2. The instrumental testing results

In the first step of confirmatory factor analysis (CFA), the response to IMA1 is inaccurate because their loading factor is lower than 0.5: 0.414. Thus, this indicator is removed. After that, the CFA is performed again; however, the loading factor of IMA2 is still lower than 0.5: 0.448; hence, IMA2 is deleted. Then, CFA is carried out again, and the result is in Table 4. In the fourth table, all loading factors exceed 0.5: 0.557, 0.633, 0.642, 0.613, 0.662, 0.610. 0.804, 0.789, 0.633, 0.654, 0.669, and 0.673 for SMM3, SMM4, SMM5, SMM6, SMM7, SMM8, BA1, BA2, BA4, BI1, BI2, and BI3. Fortunately, the accurate response is consistent, reflected by composite reliability and Cronbach Alpha above 0.7 for SMM: 0.789 and 0.793, BA: 0.788 and 0.781, and BI: 0.704 and 0.704.

Table 4: The loading factor, composite reliability, and Cronbach Alpha

4.3. The goodness of fit model detection results

Table 5 exhibits the acceptable result of the goodness of fit model, as demonstrated by CMIN/DF lower than 3 (Dash & Paul, 2021): 2.414 and RMSR below 0.05 (Junaidi, 2021): 0.023 and PGFI, PNFI, and PCFI higher than 0.5 (Dash & Paul, 2021): 0.595. 0.659, and 0.701.

Table 5: The result detection of CMIN/DF, RMSR, PGFI, and PCFI

4.4. The model estimation results

Table 6 demonstrates the estimation results of CBSEM, i.e., the probability of critical ratio for SMM  BA, SMM  BI, BA  BI is  0.000, 0.000, and 0.000 with a positive coefficient. Because these values are less than 5%, the first, second, and third hypotheses are recognizable.

Table 6: The result of the CBSEM

Table 7 exhibits the Sobel testing result with probabilty (2-tailed) of 0.003. Because this value is lower than 5%, brand awareness can mediate the influence of social media marketing and brand image. 

Table 7: The mediating testing result of brand awareness

5. Discussion

In this research, the first hypothesis is confirmed well. It indicates that social media marketing can effectively increase consumers' mindfulness of the brand. According to Kotler et al. (2022), this mindfulness enables customers to easily recognize the company's goods and services; hence, it can strengthen the relationship between consumers and the brand. Therefore, this study aligns with Park and Namkung (2022), using coffee shop customers as their samples. Although using different consumers as their objects, this study confirms Faisal and Ekawanto (2022), investigating the 311 customers of entrepreneurial products, Purwianti and Fionna (2023), after researching 360 shoppers of fashions in Batam City, and Salsabillah et al. (2024) after learning about 258 cosmetic consumers from Bandung, Jakarta, Semarang, Surabaya, Yogyakarta, and other cities.

Also, the second hypothesis is affirmed well. It indicates that social media marketing can make the customer always remember the brand in their mind. Brand image builds the customers' perspective on quality, satisfaction, and value (Raji et al., 2019). Therefore, this research aligns with Faisal and Ekawanto (2022) and Park and Namkung (2022), who reveal that the more aggressively the businesses utilize social media, the more aggressively the customers have the brand image. Similarly, Yanti and Muzammil (2023) and Salabillah et al. (2024) affirm that social media marketing can shape the brand image.

Additionally, the third hypothesis is acknowledged well. Consumers with the top level of brand mindfulness have a specific perspective that explains the brand benefits well. Indeed, they have a positive image. Therefore, this study confirms Bilgin (2018) and Salsabillah et al. (2024), revealing a positive association between brand awareness and image based on the viewpoint of 547 followers of the top five brands in Turkey and 258 customers of cosmetics in Indonesia, respectively.

Finally, the fourth hypothesis is well received: Brand awareness can mediate the effect of social media marketing on brand image. This circumstance means that marketing via social media, especially Instagram, effectively makes the customers mindful of the brand, leading to elevating its image. Thus, this evidence supports Bilgin (2018) and Salsabillah et al. (2024), proving the partial mediation effect of this brand awareness.

6. Conclusion

This study aims to prove two relationships. The first is the association between social media marketing (SMM) and brand awareness (BA). The second is the association between SMM and brand image (BI). Furthermore, this study applies 203 Kopitagram users in Bandung as the sample to answer this purpose. After statistically running and examining data, this study demonstrates a positive propensity of SMM toward BA and BI, indicating the effectiveness of SMM in enhancing BA and BI of the customers. Besides, BA mediates the impact of SMM and BI. By mentioning this evidence, this study suggests that Kopitagram facilitates social interaction among customers through its social media marketing presence. Besides, Kopitgram should provide its Instagram with updated, informative, and beneficial content, such as special prices at special events, discounts, and live music. Academically, this research toughens the positive inclination of SMM toward BA and BI with large samples. However, there are still restrictions, such as the consequences of BA and BI. Thus, the following scholars should add brand trust, loyalty, and purchase intention and decision by keeping SMM as the exogenous variable in their model.

7. Acknowledgement

The authors gratefully acknowledge Universitas Kristen Maranatha for providing partial financial support for this study.

Author Contributions: All authors contributed to this research.

Funding: Not applicable.

Conflict of Interest: The authors declare no conflict of interest.

Informed Consent Statement/Ethics Approval: Not applicable.