ANALYSIS OF THE EFFECT OF ELECTRONIC CUSTOMER RELATIONSHIP MANAGEMENT (E-CRM) AND BRAND TRUST ON CUSTOMER SATISFACTION AND LOYALTY IN PIXY COSMETIC PRODUCTS

The purpose of this study was to study the Analysis of the Effects Electronic Customer Relationship Management (E-CRM) and Brand Trust to Customer Satisfaction and Loyalty to Pixy Cosmetic Brand. by measuring indicators that influence the variables of E-CRM, Brand Trust, Customer Satisfaction and Loyalty. This research uses descriptive research design using survey methods. Sampling uses certain criteria where respondents have known Pixy Cosmetics products from Key Opinion Leader and have bought Pixy Cosmetic Products via E-commerce and Offline store as well, at least 1 time in as a sampling technique. This study discusses using the structural equation model (structural equation modeling) Lisrel to discuss the significance of the overall model relationship and predetermined pathways. The findings show that the E-CRM variable is positive and significant towards Customer Satisfaction, the Brand Trust is positive and significant towards Customer Satisfaction, and Customer Satisfaction is positive and significant towards the Loyalty.


INTRODUCTION
Indonesia is a market with attractive e-commerce growth from year to year. Since 2014, based on Euromonitor data, online sales in Indonesia have reached US $ 1.1 billion. Based on = Central Statistics Agency (BPS) data also stated that the Indonesian e-commerce industry in the last 10 years has increased by 17 percent with a total number of e-commerce businesses reaching 26.2 million units.
In 2018, e-commerce in Indonesia has experienced very rapid growth, and it is estimated that it will continue to increase as the number of entrepreneurs and entrepreneurs of Micro, Small and Medium Enterprises (MSMEs) in Indonesia grow. Over the past 4 years, e-commerce in Indonesia has increased by up to 500 percent. The latest research by Google and including in the e-Conomy SEA 2018 report shows that Indonesia's digital economy this year has reached US $ 27 billion or around IDR 391 trillion. This figure makes Indonesia's digital economic transactions ranked first for the Southeast Asia region with a contribution of 49 percent.
As reported by TechinAsia.com, the management consulting firm McKinsey & Company in August 2018 released research results regarding the current status of the Indonesian ecommerce industry, as well as projections of its development over the next few years. Currently, the e-commerce industry has had a positive impact on Indonesian employment with an estimated 4 million workers connected to this ecosystem. These new jobs include those that have emerged to support e-commerce activities, such as programming or logistics positions in e-tailing companies and existing jobs that have been renewed by e-commerce developments, such as MSME managers moving from offline to online businesses. producers and distributors, consumer shopping trends will also experience various changes. So far, the ease of transacting and choosing products has driven the number of online consumers projected to increase by around 25% annually and will reach 65 million people by 2022. The increase in popularity of online shopping is also related to the relatively cheaper costs compared to shopping for similar products offline. For example, online consumers in Java, especially urban areas, save 4-14% on average compared to offline shopping. This is because high operating costs make the prices of offline goods more expensive, while a comprehensive distribution network makes online shipping costs cheaper.
This trend change certainly encourages the development of Customer Relationship Management or CRM into E-CRM. The Indonesian economy in the era of globalization was built through the empowerment of micro, small and medium enterprises (MSMEs) and cooperatives. The Electronic Customer Relationship Management (E-CRM) system is customer management for e-Business that must deal with the complexities of attractive management for customers and business partners in a variety of media, including online and offline media, personal contact, and other electronic communication media. (Kalakota and Robinson, 2004: 148-149).
Social media makes it easier for consumers to find information and find their needs. So that the existence of social media makes it easier for consumers to get their needs and wants. Apart from consumers, it is also easier for business actors to get relationships with customers, brand trust from consumers. With a relationship with customers or E-CRM and brand trust, consumer purchasing decisions will increase. This is none other than the role of the development of social media that has developed over the last 5 years. YouTube takes first place with a percentage of 43 percent, Facebook, WhatsApp, and Instagram trailing in second to fourth place respectively. YouTube is a website that facilitates users to share the videos they have or are limited to enjoying various video clips uploaded by various parties. There are various kinds of videos that can be uploaded to this site, such as music videos from certain musicians, short films, television films, film trailers, educational videos, video blogs belonging to vloggers, video tutorials on various activities, and many more. YouTube gets its income from advertisements that are displayed before YouTube videos are played. These ads are called Google AdSense, a Available Online: https://dinastipub.org/DIJMS Page 469 program that offers payments based on the rate at which a video is played. In February 2017, it was noted that a video with a total duration of 400 hours was uploaded on Youtube every minute and a total of one billion hours of YouTube content watched by people every day. This certainly has an impact on the increasing interest of the millennial generation to create a YouTube channel so they can earn additional income, as well as making the profession "YouTuber" as their main income. The increased interest and diversity in these audiences are very beneficial for content creators from Indonesia. The number of Indonesian creators who have earned the Gold Play Button (an award that YouTube gives to creators with more than one million subscribers) has increased rapidly in recent years, indicating the increasing number of Indonesian netizens using YouTube. Also, the quality of content made by Indonesian creators has increased so that it has succeeded in attracting a wider audience.
According to Google, more and more brands are realizing the potential of YouTube to attract consumers through creative messaging. The ad video successfully grabbed the audience's attention through: 1. Innovations in telling stories through advertising, such as serials, 2. Take advantage of moments and everyday life of customers, such as holidays, 3. The use of celebrity figures, and 4. Experiment with the duration that is most effective for conveying the message.
One of the YouTubers whose channel has grown rapidly in a short period is Ririeprams, which focuses on beauty content and family blog videos. Along with the rapid development of the Ririeprams YouTube Channel, more and more beauty brands are inviting Ririeprams to collaborate with the brand. With the hope of increasing engagement and brand awareness for consumers to buy goods from brands that are being promoted by Ririeprams. One of the famous local beauty brands is Pixy. With this phenomenon, the researchers conducted a preliminary survey related to E-CRM, Brand Trust dan Customer Satisfaction. The reason consumers buy pixy cosmetic products is that they trust the reviews of Youtuber and Key Opinion Leader, RiriePrams.
18 12 30 4 The reason consumers buy Pixy cosmetic products is the company that has good relationships with consumers. 9 21 30

Source: Preliminary Observation Results of Research (2019)
Based on the pre-survey result, it provides that most of the respondents answered "No" to the statements submitted. For the first question "Pixy cosmetic products always provide a good relationship with customers" as many as 10 respondents said "Yes" and as many as 20 respondents said "No. For the second question "Pixy cosmetic products make consumers believe in the Pixy brand" as many as 8 respondents answered "Yes" and as many as 22 respondents answered "No. For the third question "The reason consumers buy Pixy cosmetic products is that they believe in the reviews of Youtuber and Key Opinion Leader, RiriePrams", as many as 18 Available Online: https://dinastipub.org/DIJMS Page 470 respondents answered "Yes" and as many as 12 respondents answered "No". In the fourth question "The reason consumers buy Pixy cosmetic products is the company has a good relationship with consumers" as many as 9 respondents answered "Yes" and as many as 21 respondents answered "No". This means that the E-CRM and Brand Trust built by Pixy are still unable to increase Satisfaction and Loyalty towards Pixy's products. So based on this, the researcher felt the need to conduct research related to the influence of E-CRM, Brand Trust, on Customer Satisfaction and Loyalty on Pixy Products.

LITERATURE REVIEW E-CRM
Customer-focused management of the entire e-business relationship with each customer, to measure, create and increase revenue and reduce costs for each customer & segment and thus to generate greater lifetime positive value (Michael, 2015(Michael, : 1080. (2005) the dimensions of E-CRM are as follows:

According to Ab Hamid and McGrath
1) Information Quality Based on the information systems literature, the quality of information is a reflection of its relevance, up-to-dateness, adequacy, consistency, and understanding. As the efficiency of consumer decision-making improves when the search is simplified, the information presented on the site must be easy to understand and update. 2) Easy of Navigation, includes both the organization (format) of the layout content as well as easy-touse navigation (ease of use). The 14 site is easy to navigate and offers quick access to the information needed with minimal effort. 3) Customer Service Efficiency, Consumers constantly demand careful, continuous, and useful communication with company representatives. Since this attribute is often identified as a salient dimension of in-store selection behavior, company representatives must have the basic technical knowledge and skills to answer online questions. They must understand the specific needs of the consumer, can deal with problems as they arise, and address consumer complaints in a friendly manner.

Brand Trust
Chaudhuri and Holbrook (2001) argue that brand trust is defined as the willingness of consumers to rely on the brand's ability to follow the functions that the brand has shown. Brand Trust can be defined as the feeling of satisfaction and trust that consumers have in their communication with companies which is now commonly practiced by social media. One of the main goals to build brand trust is to achieve a competitive advantage and thus develop company performance (Khadim, et al. al, 2018: 3).
According to Delgado (2004), the dimensions of Brand Trust are:

1) Brand Reliability
Based on consumer confidence that the product can meet the promised value or in other words the perception that the brand can meet needs and provide satisfaction. Brand reliability is essential for creating trust in a brand because the brand's ability to meet the value it promises will make consumers feel confident about getting what is needed in this case the need to get out of feeling threatened.

2) Brand Intention
Based on consumer belief that the brand can prioritize the interests of consumers when problems in product consumption arise unexpectedly. Both components of brand trust rely on subjective consumer assessments or are based on each consumer's perception of the benefits that the product/brand can provide.

Customer Satisfaction
According to Kotler (2000), customer satisfaction is the result felt by buyers who experience the performance of a company that matches their expectations (Dadang, 2010, p. 38). According to Zeithaml et al. (2009: 120), customer satisfaction can be divided into two, namely the satisfaction of each customer interacting with the provider or every Moment of truth or service encounter and second from the accumulation of each moment of truth or service encounter. From this, it can be interpreted that customer satisfaction can be felt both in every interaction process and after finishing consuming a good or service.
According to Irawan (2004: 37), there are five main dimensions of customer satisfaction, namely:

1) Price
For sensitive customers, usually low prices are an important source of satisfaction because they will get a high value for 10 money. This price component is relatively insignificant for those who are not pricing sensitive.

2) Service Quality
Service quality depends on three things, namely systems, technology, and people. Customer satisfaction with service quality is usually difficult to imitate because the formation of attitude and behavior that is in line with the company's wishes is not an easy job. Improvement must be carried out starting from the recruitment process, training, and work culture.

3) Product Quality
Customers feel satisfied after buying and using the product if the product quality is good.

4) Emotional Factor
Shown by consumers on the satisfaction they get in using a product/service that creates a sense of pride and self-confidence.

5) Efficiency
Ease of obtaining these products or services and ease of payment can make customers more satisfied if it is relatively easy, comfortable, and efficient to get a product or service Available Online: https://dinastipub.org/DIJMS Page 472

Loyalty
According to Sutisna (2003: 41), loyalty is a pleasing attitude towards a brand which is presented in a consistent purchase of that brand all the time. Meanwhile, Fandy Tjiptono (2000: 110) states that customer loyalty is a customer commitment to a brand, store, supplier based on a very positive attitude and is reflected in consistent repeat purchases. Customer loyalty is a behavioral drive to make repeated purchases and to build customer loyalty to a product/service produced by this business entity takes a long time through a repetitive purchasing process (Gibson, 2005: 75). The customer is said to be loyal or loyal if the customer shows regular buying behavior or there is a condition that requires the customer to buy at least twice in a certain time interval. Efforts to provide satisfaction are made to influence customer attitudes, while the concept of customer loyalty is more related to behavior than customer attitudes (Griffin, 2005: 46).
According to Griffin (2005) the dimensions of Loyalty are as follows: 1) Makes regular repeat purchase (make repeat purchases regularly) 2) Purchase across product and service lines (purchase other product lines from your company) 3) Refers to others (provide references to others) 4) Demonstrates immunity to the pull of the competition (shows immunity to the pull of a competitor or is not easily influenced by the persuasion of competitors).

RESEARCH METHODS
This research uses a descriptive type of research with a quantitative approach, namely research that only describes situations or events. This study does not seek or explain relationships, does not test hypotheses, or make predictions (Roswita, 2015). Descriptive research in this research is intended to get a description and information about the effect of E-CRM, Brand Trust, and Customer Satisfaction on Loyalty.
This very large and diverse population cannot be known with certainty, as for the determination of the sample size in this study using the Rao Purba formula in Sujarweni (2015: 69) : Information: n = Sample Size Z = 1.96 at a certain level of significance (90% confidence level) Moe = Margin of error, the maximum error rate is 10% By using the formula above, the following calculations are obtained: The sampling technique in this study uses simple random sampling techniques or what is commonly referred to as simple random techniques.
This research also uses Path Analysis. According to Robert D. Retherford (1993) Path analysis is a technique for analyzing causal relationships that occur in multiple regression if the independent variable affects the dependent variable not only directly but also indirectly. David Garson of (North Carolina State University: 2003) defines path analysis as a regression extension model used to test the alignment of a correlation matrix with two or more causal models being compared by researchers. In its current development, path analysis has been expanded and deepened into the form of "Structural Equation Modeling" analysis or known as SEM. Structural Equation Modeling (SEM) is a causal model analysis of the independent (exogenous) variable, the intermediate variable (endogenous), the dependent variable (endogenous), and all unmeasured variables. This immeasurable variable is called a latent variable which is measured through its indicators (which consists of component components and each component consists of items). In SEM, there are two components, namely the structural model and the formation of latent variables.  Based on Table 2, the characteristics of the 96 respondents who are customers who use Pixy Products are: a. Based on the gender of the respondents, the majority of respondents were female, that is, 66%. This is because, for women, cosmetic products have always been a part of everyday life, to get and maintain beauty from time to time. Also, the decision of women to buy cosmetics is that makeup helps women increase self-confidence, women feel that makeup is their place to be creative and show their identity, the use of makeup as a way to look more perfect, and the demands of work are often one of the reasons for women. must dress up. Besides that, the trend also has an important role in women's habit of using makeup. b. Based on the age of the respondents, the majority of respondents were between 15 and 25 years old, which is 73%. This is because in the productive age range and at this age respondents are aware of the impact of premature aging due to the many activities carried out outside the home such as sports, going to school, meetings, work, and others. There is a risk of making skin patches and burning due to sun exposure. To keep protecting the skin from the sun, women are starting to look for makeup that suits their skin condition. c. Based on the length of time being a Pixy user, the results processed by Pixy respondents are 58% new consumers who have used it for less than 1 year and 42% of old consumers who have used it for more than 1 year. Where Pixy must maintain and increase brand trust and Electronic Customer Relationship Management so that consumers become loyal to Pixy products.
The overall model fit test is made to find out how well the resulting model describes the actual conditions. Research data processing was carried out using the maximum likelihood method which was carried out with the Lisrel 8.80 application. The results of processing this data produce goodness of fit which can be seen in Table 3., From Table 3, it can be seen that the model fit value shows a good value, namely good fit and marginal fit, which means that the overall value of the fit shows a good fit. Figures 3. and 3. represent the results of the overall model testing as follows:

B. Measurement Model Test
The validity test uses the Factor Analysis model. In this test, it is necessary to pay attention to the Standardize Loading Factor (SLF) value. Questions will be considered valid if they have SLF and Extraction values greater than 0.5.
The reliability test shows the extent to which a measuring instrument can provide relatively the same results when re-measured on the same object. Reliability is calculated by the Variance Extract and Construct Reliability formula with the formula:

Construct Reliability Formula : Variance Extract Formula:
The greater the value of the calculation results, indicating that the constituent indicators for a latent variable are reliable indicators of measuring these latent variables. The suggested construct reliability value is greater than 0.7. Meanwhile, the recommended measure of variance extract feasibility is greater than 0.5.

C. Validity and Reliability Test a. E-CRM Construct Analysis
In this test, 6 indicators were observed from E-CRM that had been tested, as follows: The results of data processing analysis Table 3.3. shows that the indicators forming latent variables (constructs) show good validity, namely the value of standardized loading factor (SLF) ≥ 0.50. The value of construct reliability (CR) E-CRM 0.96 is greater than 0.70 and the variance extracted (VE) value of 0.82 is greater than 0.50 which indicates that each indicator forming latent variables has good reliability.

b. Brand Trust Construct Analysis
In this test, 4 indicators were observed from the brand trust that had been tested, as follows: The results of data processing analysis Table 3.4. shows that the indicators forming latent variables (constructs) show good validity, namely the value of standardized loading factor (SLF) ≥ 0.50. The value of construct reliability (CR) brand trust 0.94 is greater than 0.70 and the variance extracted (VE) value of 0.79 is greater than 0.50 which indicates that each indicator forming latent variables has good reliability.

c. Customer Satisfaction Construction Analysis
In this test, 10 indicators were observed from Customer Satisfaction that had been tested, as follows: Results of data processing analysis Table 3.5. shows that the indicators forming latent variables (constructs) show good validity, namely the value of standardized loading factor (SLF) ≥ 0.50. The value of construct reliability (CR) Customer Satisfaction 0.97 is greater than 0.70 and the variance extracted (VE) value of 0.78 is greater than 0.50 which indicates that each indicator forming latent variables has good reliability.

d. Loyalty Construction Analysis
In this test, 7 indicators were observed from the loyalty that had been tested, as follows: Results of data processing analysis Table 3.6. shows that the indicators forming latent variables (constructs) show good validity, namely the value of standardized loading factor (SLF) ≥ 0.50. The loyalty construct reliability (CR) value of 0.94 is greater than 0.70 and the variance extracted (VE) value of 0.68 is greater than 0.50 which indicates that each indicator forming latent variables has good reliability.

CONCLUSION AND SUGGESTION a. Conclusion
Based on the data analysis and discussion described in the previous chapter, several research conclusions can be drawn as follows: 1) E-CRM variable has a positive and significant effect on Customer Satisfaction Available Online: https://dinastipub.org/DIJMS Page 480 2) Brand Trust variable has a positive and significant effect on Customer Satisfaction 3) Customer Satisfaction variable has a positive and significant effect on Loyalty

b. Suggestion
By analyzing the research results, some suggestions that can be considered as an input for Pixy's management and subsequent researchers are as follows: 1. Advice for companies PT. Mandom Indonesia Tbk, especially Pixy's management, must have big data for Indonesian consumers who use cosmetics from various brands so that they can analyze the needs of cosmetics such as what Indonesian consumers need and expect. In the E-CRM factor, respondents gave the lowest response regarding the completeness of the information provided by Pixy.
To overcome this, Pixy can evaluate the accuracy or relevance of the content of the Pixy website. Information on the Pixy website must have a clear message, complete information can help consumers and potential consumers in solving problems, such as consumers no longer hesitate to buy the Pixy product they want. Completeness of information can increase customer satisfaction.
In addition to the information dimension, the indicator with a low value is "On the official Pixy cosmetics website page, it has explained in detail about Pixy's cosmetic products" that management must start paying attention, improving and completing features in providing information to consumers or potential consumers by providing complete and complete information. details on each Pixy product, besides that the Pixy website can add Q&A features in providing additional information about Pixy products, presenting content that can provide more information to consumers related to Pixy products in collaboration with more influencers making Pixy products known to many people without taking time long. Film actors, singers, directors, writers, bloggers, Youtubers, or even celebrities who take part in creating content that is relevant to Pixy's products and helps potential customers learn more about Pixy's products. So that it can generate interest to all consumers or potential customers who visit the Pixy website which has an impact on increasing website traffic and customer satisfaction and potential customers who have visited the Pixy website. In addition to providing special information through websites, consumer e-mail, social media such as Facebook, Instagram and Twitter, call centers, and personalization and instant messaging services such as Whatsapp, Line, Kakao Talk, Telegram and WeChat, so that customers can always be interested in using them. products from Pixy, passing a special message sent to consumers shows Pixy's concern for consumers, to maintain good relationships with consumers.
Furthermore, the navigation dimension, the indicator with a low value is the appearance of the Pixy website which is easy to understand, where Pixy's management must update the appearance of the website page which is more mobile-friendly because not all consumers and prospective kinsmen who access the Pixy website use a PC because most consumers and potential consumers will access it via a mobile phone which results in a different web page display, making Pixy's website easy to load or access on a mobile phone. Another important thing for an easy-to-understand web page is easy navigation, which makes it easier for incoming visitors to enter the next page. Pixy can add breadcrumbs which are effective visual aids to increase the ability to find the desired sections and pages on a Pixy website.
Pixy must increase brand trust, consumers and potential consumers because they don't like the formulation of the Pixy brand cosmetics, this needs to be reviewed in the ingredients for making Pixy products, especially beauty products, where the products produced by Pixy must withstand the tropical Indonesian climate. Because different countries have different climates as well as the product formulation, it must be different, it must be adapted to the different climates, weather, and skin conditions of the Indonesian people. This can increase the satisfaction of potential customers and consumers of Pixy products to remain loyal.
Besides, consumers feel that the Pixy brand has not met expectations of good product quality, Pixy must continue to improve and improve product quality to satisfy the ever-changing desires of consumers. Not only has cosmetic products that are popular for an affordable price should provide a quality product that is comparable to the price of the product. By educating potential consumers that Pixy products are cosmetic products that have health benefits that not all people know about. This will encourage buyers to continue to use Pixy products over new products from other brands and will expand the existing market to bring in more sales impacting the satisfaction and loyalty of Pixy consumers.