Purchasing Decisions on Blibli e-Commerce: Can Influencer Marketing, Brand Image, and Trust Make an Impact?

This research aims to test and analyze the influence of influencer marketing, brand image, and trust on purchasing decisions on Blibli e-commerce in Batam City. This type of research is quantitative research. The variables in this research are influencer marketing, brand image, trust, and purchasing decisions. The population in this research is the people of Batam City. The data in this study was collected using a questionnaire prepared using a Likert scale measurement scale. The total number of research samples was 204 samples. The data analysis technique uses multiple linear regression with the help of the SPSS program. The findings of this research show that influencer marketing, brand image, and trust positively and significantly influence purchasing decisions on Blibli e-commerce.


INTRODUCTION
Technology is currently developing increasingly rapidly; there is no doubt that the existence of technology has a significant influence on almost all elements of human life, especially in the aspect of trade.In general, consumers tend not to shop on online platforms to fulfil their daily needs.However, nowadays, many people rely on online platforms as a means of shopping to meet their daily needs.
One of the channels used by Indonesian people to meet their needs is electronic commerce (ecommerce).According to Rudiantara, as Chair of the Advisory Board of the Indonesian e-Commerce Association, e-Commerce will support the Indonesian economy in 2023 (Adi Ahdiat, 2023).Currently, in Indonesia, there are five well-known e-commerce applications, namely Bukalapak, Blibli.com,Lazada, Tokopedia, and Shopee.In the first quarter of 2023, the number of Blibli visits will be 25 million, and in the second quarter of 2023, it will be 27 million.Even though it can be said that there was an increase in visits of 2 million from the first quarter to the second quarter, when compared to other e-commerce, Blibli is still in fourth place.If we look at the difference in the number of e-commerce visits, it indicates that Blibli.com is an e-commerce platform that competes with other e-commerce.
Blibli is an e-commerce application made in Indonesia in 2011.To attract attention and increase sales, Blibli carries out various promotions and discounts throughout the year.However, marketing is not only carried out through promotions in the form of advertising.To attract consumer interest in shopping at Blibli, Blibli carries out marketing through influencer marketing.Globally, in 2022, influencer marketing will grow with an approximate market range of US$ 16.4 billion, but in 2023, it is predicted to jump 29% to around US$ 21.1 billion (The State of Influencer Marketing 2023: Benchmark Report, 2023).Suggestions and endorsements from influencers can have the same effect on consumers as traditional advertising ( Mai et al., 2022;Jacobson et al., 2023).It is not only influencer marketing that needs to be considered in marketing, but another aspect that needs to be considered namely Brand Image.
Brand image can influence purchasing decisions at Blibli because it can provide positive or negative perceptions of the Blibli application.Based on the assessment from the official Top Brand Award website (2023) in the Top Brand Index Phase 2 (2023) in the retail category with the online buying and selling site sub-category, it was stated that Blibli was in fourth place with a percentage of 10.60%.If we look at the overall ranking and percentage, it can be seen that Blibli's brand image is still lacking compared to other e-commerce like Shopee, Lazada, and Tokopedia.Blibli e-commerce has been established since 2011, but Blibli was replaced by e-commerce, which was only established several years after Blibli was founded.The brand image of a company is influenced by the trust of consumers who use the company's services.Companies that have an excellent brand image will more easily gain consumer trust and support than companies that have a less good brand image, because a good brand image can increase the perceived value of consumers and increase their intention to buy ( Gronholdt et al., 2000;Fu, 2023).
Trust can be said to be an essential factor in purchasing decisions at Blibli, because trust is one of the critical factors underlying transactions.When consumers feel a higher level of trust in a brand in a social media-based community, this will increase positive society towards the brand, leading to higher brand confidence, commitment and loyalty (Gyori, 2017;Chavadi et al., 2023).Various factors can influence consumer confidence in making transactions at Blibli, namely brand reputation, transaction security, availability of product information, effective communication, and others.However, based on the results of observations made on users, reviews from the Google Play Store found that users felt disappointed with Blibli, so their trust in the Blibli e-commerce application decreased.This is caused by slow service experienced by consumers, a system that often has problems, and complaints responses that are not paid enough attention.Blibli must be able to maintain consumer trust.If consumer complaints are not resolved immediately, consumer trust in Blibli will decrease.This can impact purchasing decisions made by consumers on the Blibli application.
Purchasing decisions can occur if consumers feel confident in making transactions.In the first quarter of 2023, Blibli recorded a net income of 7,775,585,000,000 trillion with a loss of -1,746,505,000,000 trillion (databox); this shows that the Blibli company is still experiencing losses, so it can be said that purchasing decisions at Blibli are still low.The problems that occurred in Blibli e-commerce encouraged researchers to conduct research.This research will examine the effect of Influencer Marketing, Brand Image, and Trust on Purchasing Decisions on Blibli e-commerce.

Literature Review Influencer Marketing
Companies generally use marketing influencers with many followers to market via social media.Influencer marketing can be interpreted as an effort to promote products or increase brand awareness in a rapidly developing industry through content distributed by social media users who are considered to have an influence (Carter, 2016;Ashill et al., 2023).In this context, influencer marketing indicates a contractual relationship in which the marketing influencer advertises a brand or product in exchange for payment, free products, or invitations to exclusive events (De Veirman et al., 2019;Rohde et al., 2021) .Chen and Yuan's research in the journal (Taylor, 2020) provides new insight into how followers view a 'trusted' influencer, which is seen through the authenticity of their posts and based on these posts being associated with the influencer's high credibility.Besides that, influencer marketing plays a role in improving brand image and increasing brand awareness effectively (Hariyanti & Wirapraja, 2019).

Brand Image
Brand image results from perception or research conducted on a brand, and consumers decide whether the brand is good or bad.Brand image can be explained as things related to a particular brand according to consumer image (Ryu et al., 2019;Manansala et al., 2022).A brand image that is unique and distinct from competitors will provide a competitive advantage for the company (Bruer et al., 2021;Manansala et al., 2022).The primary marketing strategy is to build a positive brand image, which will help maintain existing customers and appeal to potential new customers (Yi et al., 2018;Fu, 2023).Iwan and Saputra, in the journal (Afna & Khoiri, 2023), state that brand image is the image people or consumers think of when they want or need a product.Brand image is the overall customer response to the product formed in their minds (Eze et al., 2012;Anggadwita et al., 2020).
Kotler and Keller (Darmansah & Yosepha, 2020) argue that there are several relationships in brand associations which can be indicators of brand image formation, namely, 1) brand identity, 2) brand personality, 3) brand associations, 4) brand attitudes and behavior, and 5) brand benefits and advantages.Badir & Andjarwati (2020) in their research results stated that brand image has a significant impact on purchasing decisions, while sales promotions do not have a significant impact on purchasing decisions.If the brand image in consumers' minds improves, then purchasing decisions will increase (Gumelar, 2016;Zukhrufani et al., 2019).

Trust
In e-commerce, trust plays an essential role because consumers will not make transactions or shop online if they do not trust the seller.Trust is the agreement of one party to be adequate to his actions towards another party in the hope that the other party will take essential actions against the person giving the trust (Mayer et al., 1995, p. 712;Zhang, 2020).Trust can also be interpreted as someone's hope for another person's actions to obtain positive results (Ababneh, 2020;Arthur et al., 2023).Trust can be defined as a person's level of confidence and desire to act based on the words, actions and decisions of others (McAllister, 1995;Verianto, 2019).Individuals or groups will not contribute to an organization and something they do not believe in (Syafira et al., 2020).Anderson and Narus (1990), in the journal (Bernarto et al., 2019), emphasized that customers can only give trust if the customer believes that what other people do will have a positive impact on them.
Dharma (2019) also believes that there are several indicators of trust: 1) probability, 2) equity, and 3) reliability.Badir & Andjarwati (2020), in their research results, stated that EWOM does not influence purchasing decisions, while Ease of Use and Trust influence purchasing decisions.

Purchasing Decisions
Basically, before purchasing a product, consumers will look for information before choosing a product that will be the product chosen for purchasing decisions.Purchasing decisions are fulfilling an identified need or desire by selecting a product (Kumar et al., 2012;Rodrigues et al., 2023).Song et al. (2019), in the journal (Iskuntianti et al., 2020), their research results, stated that brand image influences brand trust; besides that, trust can influence the desire to buy the same product.This shows that trust can mediate among brand image and purchasing decisions (Iskuntianti et al., 2020).Quoting the opinion of Kotler & Keller in (Sudirjo et al., 2023), there are five purchasing decision indicators, namely, 1) product selection, 2) purchase channel selection, 3) determining purchase time, 4) purchase amount, and 5) payment method.
Based on the literature review in the research, the hypothesis in this research is: H1 : It is suspected that the influencer marketing variable has a partially significant impact on purchasing decisions on Blibli e-commerce in Batam City.H2 : It is suspected that the brand image variable has a partially significant impact on purchasing decisions on Blibli e-commerce in Batam City.H3 : It is suspected that the trust variable has a partially significant impact on purchasing decisions on Blibli e-commerce in Batam City.H4 : It is suspected that the variables influencer marketing, brand image and trust simultaneously have a significant impact on purchasing decisions on Blibli e-commerce in Batam City.

Research Design
This research uses a quantitative approach.Quantitative research methods are built on the philosophy of positivism, which helps research a particular population or sample to examine predetermined hypotheses (Sugiyono, 2019, p. 8).This research is descriptive, where this research will describe particular objects accompanied by explanations regarding matters related systematically to the facts or characteristics of specific populations in certain aspects factually and carefully.Apriliawati (2020) explains that when a researcher wants to see a picture of a condition or certain variables in a population using data obtained from a particular sample representing the population, they generally use a survey method.Primary data in this research uses a questionnaire technique.Furthermore, this research will also use secondary data originating from local and international jou,rnals which can be used as references for this research.

Population and Sample
Sugiyono (2019:80) believes that a population is an object or subject determined by researchers in a generalized area with specific integrity and characteristics to be studied and analyzed and then draw findings.The population of this research is the people of Batam City who use the Blibli application, the exact number of which is unknown.
The sample selection used the purposive sampling method, as explained by Sugiyono (2019:85), as a method used to ensure that research illustrations with specific considerations confirm the information obtained to make it more representative.By using a purposive sampling technique, sampling has criteria, namely that the respondent lives in Batam City and the respondent uses the Blibli application.Because the exact number of Blibli application users in Batam City cannot be known, the researcher will use the Jacob Cohen formula to determine the sample size and obtain 204 respondents.

Data Analysis Methods
This research data analysis technique contains validity and reliability tests, multiple linear regression analysis, and coefficient of determination analysis.To obtain a description of respondents' answers to the research variables this will be done with the help of the SPSS (Statistical Package for the Social Science) program to assist the data analysis process.
The validity test is a measurement to determine the accuracy and accuracy of the instrument's validity level.The validity test will be executed by comparing the calculated r with the r table.The reliability test is proof that the instrument used is reliable as a tool for collecting research data.The reliability test is reliable if Cronbach's alpha is > 0.6 (Budi Darma, 2021, p. 17).
Sujarweni, in the journal (Mappigau, 2022), explains that the multiple linear test aims to measure the influence among the independent variable and the dependent variable based on the ratio in the linear equation.Sujarweni, in the journal (Ayaumi & Komariah, 2021), explains determination analysis (R2) as a description of the ability of the dependent variable in regression measurements.Based on Table 1, regarding gender, the most significant number of respondents were male, namely 107 people, with a percentage of 52.5% and only 97 female respondents, with a percentage of 47.5%.In terms of age, the most common are aged 17 -15 years, with a percentage of 54.4%, aged 26 -35 years with a percentage of 29.9%, aged 36 -45 years with a percentage of 11.3% and aged over 45 years with a percentage 4.4%.In terms of income, the highest income earned by respondents was from IDR 1 million per month to IDR 5 million per month.Furthermore, the longest respondent used the application for less than seven months.Based on Table 2, there are variables and their indicators; it can be concluded that the calculated r is greater than the table r so that all items or indicators are declared valid.Based on Table 3, the influencer marketing variable shows a Cronbach's Alpha of 0.682, the brand image variable shows a Cronbach's Alpha of 0.743, the trust variable shows a Cronbach's Alpha of 0.633, and the purchasing decision variable shows a Cronbach's Alpha value of 0.727.Based on the data above, it shows that all these variables are said to be reliable because all variables have a Cronbach alpha value > 0.60.Each variable in this research can be said to be reliable because the measurements provide consistency.

R-Square
The coefficient of determination (R 2 ) measures how far the model can illustrate variations in the dependent variable.The results of this test can be seen from the table as follows:

Hypothesis Testing
The T-test is the first part of hypothesis testing.With the T-test, the criteria used are t count > t table and a significance value <0.05.It is known that the t table at a significance of 0.05 is 1.97190 (from the t table).Based on the calculated t of the Influencer Marketing variable of 12,173 > from the t table of 1.97190, the significance value of the Influencer Marketing variable is 0.00 < 0.05.Therefore, this means that Influencer Marketing has a significant impact on Purchase Decisions, so H1 is accepted.Based on the calculated t of the Brand Image variable of 13.817 > from the t table of 1.97190, the significance value of the Brand Image variable is 0.00 < 0.05.Therefore, this means that Brand Image has a significant influence on Purchasing Decisions, so H2 is accepted.Based on the calculated t of the Trust variable of 14,131 > from the t table of 1.97190, the significance value of the Trust variable is 0.00 < 0.05.Therefore, this means that Trust has a significant influence on Purchase Decisions, so H3 is accepted.The F test is the second part of hypothesis testing.With the F test, the criteria used are calculated f > f table and a significance value < 0.05.It is known that the f table at a significance of 0.05 is 2.65 (from the f table).Based on the results of data processing, the calculated f value is 100.279> from the f table of 3.04, while the significance value is 0.00 < 0.05, so it can be concluded that Influencer Marketing, Brand Image and Trust simultaneously have a significant impact regarding the Purchase Decision, then meaning H4 is accepted.

DISCUSSION
The results of the hypothesis test (H1) have statistically provided evidence that Influencer Marketing has a positive and significant impact on Purchasing Decisions on Blibli e-commerce in Batam City.These results support research (Nurhamidah, 2022), stating that Influencer Marketing has a positive and significant impact on Purchasing Decisions.The output of the hypothesis test (H2) has statistically provided evidence that Brand Image has a positive and significant impact on Purchasing Decisions on Blibli e-commerce in Batam City.These results support research (Darmansah & Yosepha, 2020), stating that Brand Image has a positive and significant impact on employee performance.The output of the hypothesis test (H3) has statistically provided evidence that Trust has a positive and significant impact on Purchasing Decisions on Blibli e-commerce in Batam City.These results support research (Istiqomah et al., 2019), stating that trust has a significant effect on purchasing decisions.The output of the hypothesis test (H4) has provided evidence that Influencer Marketing, Brand Image and Trust simultaneously have a significant impact on Purchasing Decisions on Blibli e-commerce in Batam City.

CONCLUSION
Influencer Marketing has a positive and significant impact on purchasing decisions, so every marketing influencer who collaborates with Blibli can increase purchasing decisions at Blibli.However, conversely, if the influencer marketing figure who collaborates with Blibli is considered less attractive by consumers, the level of purchasing decisions will be low.Brand image has a positive and significant impact on purchasing decisions, so the better the Blibli brand image, the higher the level of purchasing decisions.If the Blibli brand image in the eyes of consumers is not good, then purchasing decisions will be low.Trust has a positive and significant impact on purchasing decisions; if consumers have great trust in Blibli, then the level of purchasing decisions that occur will be higher.If consumer trust is low, consumers will not make purchasing decisions on Blibli e-commerce but will switch to other e-commerce applications.
Suggestions for e-commerce Blibli, using influencer marketing in marketing is essential.Blibli must consider influencer marketing, which will work with them to build brand image and company trust.Influencer marketing must have credibility and attractiveness to influence consumers to transact on the Blibli application.If influencers can influence consumers to make transactions, then indirectly, the company can instill a brand image in consumers' minds.It will be easier for consumers to remember the Blibli application as a trusted e-Commerce.If consumers feel that Blibli e-commerce has a positive brand image for society, the level of consumer trust will increase, and this can encourage purchasing decisions in the Blibli application so that Blibli can increase transactions in the application and increase profits for the Blibli company.Suggestions for future researchers can use this research as a reference for further research, adding other factors and variables that are related or related to purchasing decisions.

Table 1 .
Demographic Profile

Table 3 .
Reliability Test

Table 4 .
Multiple Linear Analysis Results The constant has a regression coefficient of 2,658; this shows that the Purchase Decision has not been influenced by the Influencer Marketing, Brand Image and Trust variables, so the Purchase Decision has a value of 2,658.The Influencer Marketing variable has a regression coefficient of 0.270 or 27.0%; this shows that the other independent variables have constant or fixed values.If the influencer marketing variable increases by 1%, it can cause an increase in the purchasing decision variable by 0.270.The Brand Image variable has a regression coefficient of 0.322 or 32.2%; this indicates that the other independent variables have a constant or fixed value.If the brand image variable increases by 1%, it can cause an escalation in the purchasing decision variable by 0.322.The Trust variable has a regression coefficient of 0.471 or 47.1%; this shows that the other independent variables have constant or fixed values.If the trust variable increases by 1%, it can cause an escalation in the purchasing decision variable by 0.471.

Table 5 .
R-Square ResultsBased on table5, it shows that the R Square value is 0.601.It can be concluded that the Purchasing Decision variable can be illustrated by the Influencer Marketing, Brand Image, and Trust variables of 60.1%, and other variables outside this research influence the remaining 39.9%.