Effects of E-Marketing and Social Media Marketing on E-commerce Shopping Decisions

This research analyzes marketing through electronic media and social media as well as purchasing decisions on e-commerce sites; testing whether marketing through electronic media affects shopping decisions on e-commerce sites; testing whether marketing through social media affects shopping decisions on e-commerce sites; and testing whether marketing through electronic media and social media marketing affects shopping decisions on ecommerce sites. This research method is quantitative with a survey approach. The research population is ecommerce consumers in the city of Bandung. The sample size used was 125 samples. Path analysis is used as an analysis technique. The findings show that marketing through electronic media and social media and consumer decisions are in the high category. Marketing through electronic media affects consumer decisions to make purchases of 45.35%; Marketing through social media affects consumer decisions to make a purchase of 35.73%. In addition, marketing through electronic media and social media affect buying decisions of 81.08%


I. INTRODUCTION
There has been a significant increase in the number of Internet users worldwide. From We Are Social data in 2019, there are 4.3 billion active Internet users globally and the number of Internet users continues to increase by 9.1% annually on average. This growth has had an impact upon the growth of mobile and social media users. Active social media users and mobile users amount to 3.4 billion and 3.2 billion users, respectively. Social media users via mobile grew by 10%. Facebook continues to be the most popular social media platform with 2.2 billion users (wearesocial.com, 2019).

77
In Indonesia, the e-commerce business reached 74 million consumers in 2015, which increased to 102 million users in 2016. Compared to other countries, the average Internet penetration in the Asia-Pacific region was 32% while Indonesia only achieved 29%. Among the 74 million people who have access to e-commerce, only 7% participate in online transaction activities. Indonesian people have started to like shopping for expensive products on e-commerce sites (Noviandri, 2015).
Currently, the development of e-commerce businesses is strongly influenced by the e-marketing activities (Strauss and Frost, 2012) and social media marketing (Malita, 2011) conducted by e-commerce sites, considering that both strategies influence the attitudes and behaviors of consumers while shopping. E-commerce companies that utilize e-marketing and social media marketing strategies include Tokopedia, Shopee, Lazada, Zalora, Blibli.com, JD.id, Zilingo Shopping, and AliExpress (iPrice.co.id, 2018). However, the current study will only focus on Lazada, Zalora, and Tokopedia because these three e-commerce sites comprise different market segments, each with their own uniqueness. Lazada focuses on providing various consumer goods to meet every day needs, Zalora focuses on providing a variety of fashion products, and Tokopedia focuses on providing a diverse range of payment transaction tools.
This research was conducted in Bandung since it is a city wherein the number of e-commerce consumers continues to rise rapidly (iPrice.co.id, 2018). This research seeks to examine e-marketing, social media marketing, and e-commerce shopping decisions and determine the extent of the influence that e-marketing and social media marketing have on consumers' shopping decisions on e-commerce sites, both individually and together.

A. Purchase Decision
Consumers' online shopping habits can be gauged only with an understanding of their buying behavior. Engel et al. (2006) and Schifman and Kanuk (2012: 34) referred to consumer behavior as the behaviors involved in searching for, buying, utilizing, evaluating, and consuming products that consumers expect will meet their needs. Kotler and Armstrong (2012:177) stated that purchase decisions form a part of the buying decision process. In the final stage, consumers form preferences for various brands from a collection of available alternative choices as well as develop an intention to buy products they like. Kotler and Armstrong (2012:183) explained that consumer decisions comprise five decisions, namely, choice of product, choice of brand, choice of distributor, choice of time of purchase, and choice about the quantity to be purchased; these five aspects are used to measure consumers' decision to shop online. These dimensions can be empirically measured and researched from e-commerce sites.

B. E-Marketing
E-marketing refers to marketing via electronic communication technology, specifically the Internet. The Internet is a physical network of various connected computers comprising server and communication networks that are used to store information and disseminate it (Chaffey, et al. 2013). Kotler and Armstrong (2012) described e-marketing as marketing using various online electronic media. Chen-Ling and Lie Ting (2006) and El-Gohary (2010) referred to e-marketing as marketing using web media and various related transaction activities that can be easily accessed by consumers.
Chen-Ling and Lie Ting (2006) stated that e-marketing comprised e-mail marketing, e-mail newsletters, viral e-mail, and banner ads; all four channels were examined in this study to empirically measure the effect of emarketing on e-commerce shopping decisions. Furthermore, they discussed that e-marketing is the process of marketing products and services to customers by utilizing various web media. E-marketing creates fundamental attitudes and behavioral changes in e-commerce consumers; thus, consumers who shop online are influenced by e-marketing. Therefore, our first research hypothesis (H1) is as follows: E-marketing influences buying decisions on e-commerce sites.

C. Social Media Marketing
According to Gurnelius and Susan (2011: 10), social media marketing refers to online publication that utilizes communication technology such as Web 2.0. It involves conversation, involvement, and participation. Evans and McKee (2011: 34) and Paramitha (2011) explained that social media transforms people from content readers into content publishers. This involves a change from the typical broadcast mechanism to a one-to-many 78 model through conversation development. Ramadani et al. (2014) stated that social media marketing was a multi-channel media outlet that can influence the consumers' interest in making purchases. Gurnelius and Susan (2011: 15) and Puntoadi (2011) found that the purpose of social media marketing is relationship building, brand development, promotion, publicity, and market research. Social media marketing is used as a product marketing tool and aims to strengthen brands.
Moreover, Gurnelius and Susan (2011: 15) observed that the most popular social media platforms are Instagram, Facebook, Twitter, WhatsApp, and Line; these are the five tools used in this study to empirically evaluate the consumers of e-commerce sites. Andersson andEkman (2009), Keltner et al. (2014), Permatasari and Kuswadi (2017), and Permatasari and Laydi (2018) confirmed that social media marketing influences shopping decisions. Therefore, our second research hypothesis (H2) is as follows: Social media marketing influences buying decisions on e-commerce sites.

III. RESEARCH METHODS
This research is a descriptive verification study (Sarwono, 2008;Noor, 2011;and Sanusi, 2011). The descriptive research aspect involves describing the variables affecting e-marketing, social media marketing, and e-commerce shopping decisions, while the verification research aspect involves determining how e-marketing and social media marketing influence e-commerce shopping decisions. This research was designed to have a causal timeframe, which is limited to one point or is cross-sectional. (Sugiyono, 2016;Sudjana, 2012).
The sample population includes the e-commerce site customers in Bandung city, who use Lazada, Zalora, and Tokopedia but whose numbers are not known with certainty. Since the population size is not known with certainty, Malhotra (2011: 291) states that sample size can be determined by taking at least four or five times the number of statement items in the questionnaire. In this study, there are 25 statement items; hence, the study sample size is 25 × 5 = 125 consumers. However, in our implementation, we extended the sample size to 150 consumers, assuming that 25 respondents may be mistakenly included in the sample or that the questionnaire distributed may be unsuitable for processing. The research data were obtained using a Likert-scale choice of five alternative answers or scores ranging from 5 (largest) to 1 (smallest). Data collection was conducted using a questionnaire instrument (Sugiyono, 2016).
The descriptive analysis involved first converting the questionnaire results into quantitative data based on the weight of predetermined categories and then creating a general description of the respondents' responses to each variable, including statistical processing by reviewing the number, frequency, and standard deviations for each item and indicator, which were summarized into tabulations and graphs. Path analysis was used as the statistical data analysis technique.

IV. RESULTS AND DISCUSSION
From the summary of results, the average e-marketing score was 4.07, the average social media marketing score was 4.05, and the average shopping decision score was 4.09-all of which can be classifed as high scores.

A. Correlation Test Results
There was significant correlation among all variables. The correlation test results indicate that the correlation between e-marketing (X1) and social media marketing (X2) was 0.660 (strong), e-marketing (X1) and shopping decisions (Y) was 0.832 (very strong), and social media marketing (X2) and shopping decisions (Y) was 0.946 (very strong)

B. Significance Test Results
The significance test results of the simultaneous influence of e-marketing and social media marketing on shopping decisions are shown in Table 1. The results indicate that the F count was greater than the F value. The value of sig. was 0.00, which is less than 0.05; thus, the hypothesis is proven.  (2019) The significance test results regarding the influence of e-marketing and social media marketing on shopping decisions on e-commerce sites are partially presented in Table 2.  (2019) The results indicate that the e-marketing had a significance value of 0.00; thus, ho is rejected. Its t-value is greater than the table value, thus, the correlation significance test conducted by using both, a significance value and a t-test, prove the acceptance of H1. Therefore, e-marketing has a positive and significant effect on shopping decisions on e-commerce sites. Social media marketing had a significance value of 0.00 and a t-value greater than the table value; hence, H2 accepted. The correlation significance test conducted by using both a significance value and a t-test, thus, social media marketing has a positive and significant effect on shopping decisions on e-commerce sites.

80
As per Figure 2, the results of the path analysis test are described in Table 3.  (2019) Thus, the path equation formula is given by: Y = 0.574.X1 + 0.483.X2 +  The results of the path coefficient test are listed in Table 4.  The total influence of e-marketing on shopping decisions amounted to 45.35%, with 32.95% arising from the direct effect of e-marketing and 12.4% from the indirect effect of social media marketing. The total impact of social media marketing on shopping decisions amounted to 35.73%, with 23.33% arising from the direct influence of social media marketing and 12.4% arising from the indirect influence of e-marketing. Thus, the combined effect of e-marketing and social media marketing on shopping decisions, amounts to 81.08%.  Table 5 shows that the coefficient of determination (R 2 ) is 0.810. This value shows that e-marketing and social media marketing simultaneously influence 81.0% of shopping decisions, while the remaining 19.0% is influenced by other factors not discussed in this study, such as sales promotion, brand image, etc. Based on the influence classification guidelines, presented in Table 5, the magnitude of these simultaneous effects can be classified as very strong.

D. Determination Test Results
These results indicate that both e-marketing and social media marketing have a positive and significant effect on consumers' decisions to shop on e-commerce sites, especially Lazada, Zalora, and Tokopedia. These online shopping sites use e-mail marketing, e-mail newsletters, viral e-mail, and banner ads as strategies to encourage consumers to shop. These strategies are implemented regularly by Lazada, Zalora, and Tokopedia so that consumers have adequate information to support their purchasing decisions. Meanwhile, the social media marketing activities are regularly conducted by Lazada, Zalora, and Tokopedia via Instagram, Facebook, Twitter, WhatsApp, and Line. These five social media sites are used to inform consumers about various offers and discounts that can influence them to shop on e-commerce sites.

V. CONCLUSION
The hypothesis testing results indicate that e-marketing, social media marketing, and consumer e-commerce shopping decisions belong to the high classification. E-marketing affects consumers' e-commerce shopping decisions, both positively and significantly. The total contribution of e-marketing's influence on consumers' decisions to shop on e-commerce sites, was 45.35%. Similarly, social media marketing also influences consumers' e-commerce shopping decisions, both positively and significantly. The total contribution regarding social media marketing's influence on consumer decisions to shop on e-commerce sites was 35.73%. Thus, emarketing and social media marketing together influence consumer decisions to shop on e-commerce sites, both positively and significantly, with a combined contribution of 81.08%. Thus, the management of Lazada, Zalora, and Tokopedia must maintain and increase the frequency of marketing activities via e-marketing and social media marketing.