The Influence of Web Design and Brand Impression on Purchase Interest of Online Shopping Site OLX.CO.ID

Through the study, the researchers would like to identify the influence of Web Design and Brand Impression on Purchase Interest of the costumers toward the online shopping site OLX.CO.ID. The research is conducted due to the improving interest among the e-commerce businessmen, including the OLX.CO.ID. Within the conduct of the study, the researchers employed two independent variables, namely Web Design and Brand Impression, and one dependent variable, namely Purchase Interest. Then, the respondents for the study were selected by implementing the purposive sampling technique; in total, there were 120 respondents that had been selected for the conduct of the study. The data that had been gathered from the respondents were analysed by using the Partial Least Square (PLS) technique. The results of the study show that Web Design has positive and significant influence on Purchase Interest with the significance value of probability 5.942 and, similarly, Brand Impression has positive and significant influence on Purchase Interest with the significance value of probability 5.059 for the online shopping site OLX.CO.ID.


INTRODUCTION
In this section, an elaboration on the background of the study will be provided altogether with the problem formulation, the objective of the study and the contribution of the study. Each part might be consulted further in the following section.

Background
Online marketplace is predicted to increase in each year altogether with the development of the smartphone industry, the Internet access quality in Indonesia and the wide coverage of each online service provider. In the same time, the three aspects, namely smartphone industry, Internet access quality and wide coverage of online service provider become one of the factors behind the rapid advancement of online business. Such advancement thus has encouraged more and more companies to be the e-commerce players. The ecommerce has been turned into a place where the businessmen develop their business. The statement has been confirmed by the definition of e-commerce according to Kotler (2010): any online company or sites that offers transaction or facilitation for the product and service sale and purchase through online manner.
According to the Association of Indonesian Internet Service Providers (APJII, Asosiasi Penyedia Jasa Internet Indonesia) in their infographic, it is predicted that in 2015 there has been approximately 85% Internet users in Indonesia who accessed websites by using smartphones and 56% of these users access the websites for the purpose of online shopping activities. This implies that the owners of the websites or the online shops (ecommerce) should improve the convenience within the websites, including the convenience on the access by means of both smartphone and computer unit. Furthermore, APJII also states that the society has shown increasing interest to use the Internet. From the data in Figure 1, it is apparent that the number of Internet users in Indonesia has been increasing significantly from 1998 until 2015. APJII projects that the highest increase on the number of Internet users in Indonesia has been found in 2010; by 2015, the number of Internet users in Indonesia has been 139 million people. This finding confirms that the high increase on the number of Internet users will certainly draw the interest of the society in visiting the online shops. In relation to the statement, the results of a survey by Rakuten team shows that the society has high interest to give and to be given gift. The survey was administered to 500 respondents in September 2015 and the results of the survey might serve as the direction for the Rakuten and the other players of e-commerce for this year's activities.

Source: Kompasiana
The competition among the online shops in Indonesia has been quite intense. Such intense competition is apparent from the number of online shops that strive to gain the highest marketplace in the e-commerce activities. Based on the data from Alexa.Com, the top four online shops in Indonesia from December 2015 until May 2016 are as follows: Source: Alexa Traffic Rank (2016) Figure 2. Traffic of E-Commerce Sites Search in Indonesia Alexa.Com, which is a subsidiary of Amazon.Com, is a company that ranks websites both domestically and globally altogether with other information to more than 30 million websites that have been registered. According to the survey data from Alexa.Com, the top four e-commerce companies in Indonesia are BukaLapak.Com, Lazada.Com, Tokopedia.Com and OLX.Com. These data were taken from December 2016 until May 2016. Apart from the top four e-commerce companies, OLX seems to be the only company that suffers from quite significance decreasing performance. As a result, many people start wondering about the factors behind such decreasing performance.
At the beginning of its establishment, OLX (Online eXchange) was an international company that operates in the C2C Marketplace. Then, OLX itself started to be developed in 2006. The company has claimed to possess networks in the local marketplace from 106 countries throughout the globe. Nowadays, the majority shareholders of OLX are a business group from South Africa named Naspers. In Indonesia alone, the name OLX started to be localized into the domain OLX.co.id in 2010. At that time, OLX entered the Indonesian e-commerce indirectly within the marketplace competition because the company still served as the buffer for the development of the classified ad websites namely TokoBagus.Com. After 2012, OLX officially bought 65% of TokoBagus.Com share that had been very popular at that time. In May 20 th , 2014, TokoBagus.Com announced the change on the name and the URL address of the company from TokoBagus.Com into OLX.co.id. This event has been quite significant in the domain of Indonesian e-commerce, recalling the fact that TokoBagus.Com has occupied the top Journal of Marketing andConsumer Research www.iiste.org ISSN 2422-8451 An International Peer-reviewed Journal Vol.67, 2020 61 ten online shops in Indonesia. Not to mention, for a developing company, the change of the name (the brand) becomes a radical step that might be certainly caused a fundamental reason. Loyal customers will generally continue to use the brand despite being faced with many alternative brands of competing products (Imaningsih, E.S. et al, 2019), at offer superior product characteristics (Astini, Rina, 2016).
As a brand, TokoBagus.Com has certainly been very popular among the service users. Many efforts have been performed in order to boost the brand of TokoBagus.Com especially through the television advertisement so that the brand awareness of TokoBagus.Com might be improved within the society. Consequently, the change of the name might impact the impression of the brand. In turn, the change of the name might certainly impact the purchase interest of the consumers. With regards to the situation, in order to identify further the influence of the web design and the brand impression of the online shopping site OLX.co.id, pre-survey activities were performed to 10 respondents who had been selected randomly on June 25 th , 2016. The results of the pre-survey activities might be consulted in Table 1 below.  (2016) Departing from the results of the pre-survey activities and also based on the above elaboration, a study under the title "The Influence of Web Design and Brand Impression on Purchase Interest of Consumers on the Online Shopping Site OLX.CO.ID" should be pursued in order to describe the case that has proposed in the topic better.

Problem Formulation
Departing from the elaboration on the background of the study, the following problems will be proposed for the conduct of the study: 1. How does the influence of Web Design on Purchase Interest of the consumers in the online shopping site OLX.co.id? 2. How does the influence of Brand Impression on Purchase Interest of the consumers in the online shopping site OLX.co.id?

Objective
Through the conduct of the study, the following objectives are expected to be met: 1. Identifying and analysing the influence of Web Design on Purchase Interest of the consumers in the online shopping site OLX.co.id. 2. Identifying and analysing the influence of Brand Impression on Purchase Interest of the consumers in the online shopping site OLX.co.id.

Contribution
The contribution that is expected to be delivered by the conduct of the study might be divided into two aspects as follows: 1. Academic Contribution The results of the study are expected to serve as in-depth knowledge and review materials for the comparison between the theory and the practice of Web Design, Brand Impression, Purchase Interest and Consumers in the online shopping site OLX.co.id.

Practical Contribution
The results of the study are expected to serve as a matter of consideration for the low, moderate and high class of e-commerce companies in anticipating and designing strategies for dealing with the e-commerce competition.

LITERATURE REVIEW, HYPOTHESIS DEVELOPMENT AND RESEARCH FRAMEWORK
In this section, several literatures will be reviewed with regards to the topic of the study. The literatures will be reviewed in order to provide theoretical foundation for the conduct of the study. Then, along with the literature review the hypotheses within the study will be developed and the framework will be designed in order that the conduct of the study will result in valid and reliable data.

A. Literature Review
This section contains the theoretical foundations that will guide the researcher in conducting the study. In addition, the researcher also explains the results of several studies that have been conducted previously in order to assist the investigation process within the study. The literature review then is related to the concept of Marketing Management, E-Commerce Marketing, Web Design, Brand Impression and Purchase Interest.

Marketing Management
According to Kotler & Armstrong (2008), marketing is a social and managerial process in which individuals or organizations attain what they need and they desire through the creation of one value or the exchange from one value to another. Similarly, Firdauz (2010) states that marketing might be defined as one of the main activities that businessmen, including farmers (agribusinessmen), perform for the sake of their survival. Last but not the least, Machfoedz (2010) proposes that marketing is a process that has been implemented by a company in order to meet the needs and the desires of the consumers by providing products (goods and service).

e-Commerce Marketing
According to Kotler & Armstrong (2008), e-commerce means that a company or a site offers transaction or facilitation of produce and service sale by means of online manner. On the other hand, Sudaryono (2015) states that e-commerce refers to the development and the part of information technology era that is able to create new economy. Last but not the least, Prasetio (2012) states that e-commerce refers to the distribution, the purchase, the sale and the marketing of goods and service though electronic system such as Internet, television, www-based network or another computer network.

Web Design
Prasetio (2012) states that web design defines the quality and the beauty of a website. Design is highly influential for the assessment of the visitors with regards to the good or the poor layout of the website. On the other hand, Sarwono & Martadiredja (2008) state that web design is a general term that has been used in order to cover how the content of a website is displayed in front of the end users through a web browser or a web-based software.
According to Flavian et al. (2009), there are four factors that infuence the improvement of the ecommerce website design. These factors might be described into the following dimensions: a. Appearance Appearance consists of interesting layout, which is important to attain the high level of credibility within the website. Appearance offers clarity and combination of colors within the website and also balance between the display and the loading speed. The tools that have been usually operated are the visual tools (figures, graphics, icons, animations and colors). b. Navigation Navigation consists of ease of website navigation. The ease of website navigation enables the users to control the website navigation in searching the web content. The tools that have been usually operated are site map of the web, the "Back" or "Return" button and the search engine. c. Content Content consists of clear content arrangement and management. The clear content arrangement and management will ease the information search and thus result in well-qualified content and even information. The tools that have been usually implemented are information tools (update, ease of understanding and relevance) and visual tools (figures of the products with the accurate and well-qualified size). d. Shopping Process Shopping process consists of the offer of easy and efficient shopping process, the prpovision of information on the quality of the product and the service that have been offered within the purchase, the emphasis on privacy and the guarantee on the security aspects. The Shopping Process usually takes less that five stage, is linked to the information that has been useful for purchase activities (given product, general description and additional information) and guarantee on privacy and security. Kotler & Keller (2009) define that brand image is a process in which an individual selects, organizes and interprets information input in order to create a meaningful description. On the other hand, Suryani (2008) states that brand impression is generally defined as all matters that have been related to the brand inside the mind of a consumer. Last but not the least, Rangkuti (2009) states that brand image refers to a group of brand associations that have been established and internalized within the mind of a customer.

Brand Impression
According to Kotler & Armstrong (2008), the measurement on the brand impression might be performed based on the aspects of brand namely: a. Strength Strength refers to the physical prominence of a brand that might not be found in another brand. The physical prominence of the rband refers to its physical attributes so that it might be considered as a competitive edge in comparison to the other brand. Then, the aspects that belong to the strength of a brand consists of the functionality of all facilities within the product, the physical appearance of the product, the price of the product and the layout of the product's supporting facility within the wide coverage of the product in the market. b. Uniqueness Uniqueness refers to the capacity of differentiating one brand from another. This impression appears from the attributes that differentiate the given product from the other products. Then, the aspects that belong to the uniqueness of a brand consists of layout variation or name variation of the brand that is easy to remember or to speak and also the physical appearance of the product itself. c. Favorability The aspects that belong to the favorability of the product are, namely, the name and the capacity of the brand that is easy to remember by the consumers so that the brand might be the favourite alternative of the consumers and also the relevance between the brand impression in the mind of the consumers and the expected impression on the company with regards to the given brand.

Purchase Interest
Purchase interest is described as a situation that a consumer should deal with prior to taking an action, which might serve as the basis for predicting the behaviour or the action. Purchase interest thus might be defined as a behaviour that appears as a response toward the object that displays the desire of the consumers to perform purchase (Kotler & Keller, 2009). On the contrary, Swasta in Kristianto (2009) states that purchase interest refers to the social relationship and action that an individual consumer performs in order to assess, attain and use the goods through a process of exchange or purchase that has been initiated by the decision-making process on the given actions. In other words, purchase interest refers to the tendency of comitting a purchase in the future, but the measurement on the tendency of comitting the purchase in general is conductd in order to maximize the prediction on the actual purchase itself.
According to Ferdinand (2006), purchase interest might be identified based on the following aspects: a. Transactional Interest Transactional Interest refers to the tendency and the willingness of an individual to purchase a product despite the presence of other product alternatives. b. Referential Interest Referential Interest refers to the tendency of an individual to refer the others to a certain product or to share the positive aspects of a product to other people. c. Preferential Interest Preferential Interest refers to the interest that describes the behaviours of an individual who has main preference to the given product such as prioritizing or taking a like on a product or service. The preference might only be replaced if there is something that happens with the preferred product.

d. Explorative Interest
Explorative Interest describes the interest of an individual who always look for information about the products that he or she likes in order to learn about the traits of the product. From the above elaboration on the purchase interest, it might be concluded that purchase interest refers to the stage of purchase tendency that a consumer has toward the product or service in certain period of time since the consumer has positive attitude toward the product or the service based on the purchase experience in the past.

B. Hypothesis Development
Based on the framework of the study, in conducting the study there are two hypotheses that will be proposed and each hypothesis will be elaborated briefly as follows.

The Influence of Web Design on Purchase Interest
According to Kotler and Armstrong (2008), online marketing is related to the creation of websites. An online marketer thus should create attactive website and find ways for retaining the visits by the online consumers. The statement is supported by the results of a study by Gunarsih etal. (2012), which show that Web Design has positive and significant influence on Purchase Interest of the consumers. Departing from this elaboration, the first hypothesis for the conduct of the study will be proposed as follows: H1 : Web Design has positive influence on Purchase Interest of the consumers.

The Influence of Brand Impression on Purchase Interest
The results of a study by Octavia & Mashariono (2015) show that Brand Impression has significant influence on the Purchase Interest of Flexitemp One Med digital thermometer. Similarly, the results of a study by Desi (2011) show that Brand Impression has significant influence on Purchase Interest of Telkom Speedy. Departing from this elaboration, the second hypothesis for the conduct of the study will be proposed as follows: H2 : Brand Impression has positive influence on Purchase Interest of the consumers.

Research Framework
Departing from the theories that have been reviewed, the research framework for the conduct of the study might be designed as follows. Note: 1. The independent variable or the exogenous variable refers to the variable which value is not dependent on the value of the other variable and is symbolized by the letter X. Within the conduct of the study, there are two independent or exogenous variables namely Web Design (X1) and Brand Impression (X2). 2. The dependent or endogenous variable refers to the variable which value is dependent on the value of the other variable and is symbolized by the letter Y. Within the conduct of the study, the dependent or the endogenous variable is Purchase Interest.

METHOD
Prior to the conduct of the study, the method that had been used during the conduct of the study might be described in the following sections. Through the description, it is expected that the overall process within the conduct of the study might be well described.

A. Site and Period
The study was conducted in four areas of South Jakarta namely Kebayoran Lama, Cipulir, Petukangan and Gandaria. In order to attain the data for the conduct of the study, the object of the study as the data source should be defined first. Then, the study took place from March 1 st , 2016 until the data had been completely gathered.

Design
The design that had been implemented in the study was the associative causal design using qualitative approach. According to Sugiyono (2008) question the relationship between two variables or more namely the relationship between the exogenous variable (the influencing variable) and the endogenous variable (the influenced variable). Specific to the context of the study, the causal design was implemented in order to identify the influence of Web Design and Brand Impression on Purchase Interest of the consumers in OLX.co.id online shopping site.

Variable Definition and Variable Operationalization
In order to guide the conduct of the study, the variable definition and the variable operationalization should be outlined. The outline of both aspects might be consulted in the following sections.

Variable Definition
According to Sugiyono (2008), research variable refers to the attributes, the traits or the values of an individual, an object or an activity that has certain variation which a researcher has assigned in order to be studied and be verified. Specific to the context of the study, there were two types of variables namely the exogenous variable and the endogenous variable. The exogenous variable was Web Design (X1) and Brand Impression (X2). The exogenous variables like these ones are known as independent variables (stimulus, predictor or antecedent). Independent variable refers to the variable that is able to influence or trigger the creation of the endogenous (dependent) variable. On the other hand, the endogenous variable within the conduct of the study was Purchase Interest (Y) of the consumers. This kind of variable is often termed as dependent variable (output, criteria or consequence). In other words, independent variable refers to the variable that has been influenced or that has been impacted by the independent variables.

Variable Operationalization
The variable operationalization for the conduct of the study might be consulted in Table 2, Table 3 and Table 4 below.    (2006) In addition to the variable operationalization, the criteria of the Liker Scale Instrument is also provided in Table 5 below.

Population and Sample
According to Sugiyono (2008), population refers to the area of generalization that consists of objects or subjects with certain qualities and characteristics that a researcher has assigned for the purpose of his or her study and verification. As a result, population might not only be people but also matters. In the same time, Sugiyono (2008) states that sample is part of the qualities or the characteristics that the given popultion has. In the context of the study, the samples were gathered by using the non-probability sampling technique. The nonprobability sampling technique was used because the size of the opportunity for the element to be selected again as the subject of the study had been unknown.
Then, the sample gathering method that had been implemented in the study was the purposive sampling method. Purposive sampling method refers to the sample gathering method that involves certain consideration. In defining the size of the sample, Sugiyono (2008) explains that the number of sample members has been widely known as the sample size. The expected number of sample size is 100% representative for the population so that there should not be any errors of generalization. In order word, the sample size should be equal to or representative for the population.
Within the conduct of the study, the number of sample that had been planned was 120 people. The consideration was that this number had been representative for the appropriate size of the sample. According to Hair et al. (2010), the minimum sample size was 5-10 observations for each estimated parameter. Specific to the context of the study, the number of the indicators was 24 and therefore the number of the sample should be 5 x 24 = 120 people.

Data Gathering Method
The data in the study were directly attained from the completion of the questionnaire that had been delivered to the consumers or the respondents with regards to their view or perception toward the influence of Web Design and Brand Impression on Purchase Interest in OLX.co.id online shopping site around South Jakarta.

Data Analysis Method
The data analysis method that had been implemented in the study was the Partial Least Square (PLS) Analysis and the Structural Model Analysis. Tthe procedures of each method might be consulted in the following section.

Partial Least Square Analysis
The data analysis method that had been implemented in the study was the component-based or the variance-based SEM in which the data processing relies on the use of Partial Least Square (Smart-PLS) Version 3.0. As having been proposed by Ghozali (2014), Partial Least Square (PLS) is a powerful analysis method because this method does not assume that the data should be in certain measurement scale with little number of sample. The other name for this Partial Least Square Analysis method is the evaluation measurement or the outer-model measurement. Outer model measurement (outer relation measurement or measurement model) defines how each indicator block establishes relationship to its latent variable. The formula for the block with its reflective indicator is as follows: According to the above formula, x and y refer to the indicator ot the manifest for the exogenous and the endogenous latent variable and , whereas and refers to the loading matrix that describes the coefficient of simple regression which links the latent variables to their indicators. The residuals that have been measured by means of and might be interpreted as the measurement errors or noise (Ghozali, 2014). Then, within the study the sequence of the Partial Least Square analysis consisted of the following stage: a. Convergent Validity The convergent validity of a measurement model with indicator reflection should be assessed based on the correlation between the item score/the component score that has been calculated by means of PLS. According to Ghozali (2014), the reflexive individual measures will be considered high if the correlation of the reflexive individual measures has been higher than 0.70 in relation to the construct under measurement. However, from the preliminary to the development stage of the study the loading score 0.50 until 0.60 has been considered sufficient.

b. Discriminant Validity
The discriminant validity tst from the measurement model with the reflective indicators might be assessed based on the cross-loading value between the indicator and its construct. An indicator will be considered valid of the indicator has the highest loading factor value relative to the targeted construct than the other construct; thereby, the constructs of the latent variables will able to predict the measures in their block better than the measures in the other block. Another method for viewing the discriminant validity was comparing the Square Root of Average Variance Extracted from each construct and the correlation from one construct to another within the model. If the square root of AVE for each construct is higher than the correlation value from one construct to another within the model, then the measurement model has good discriminant validity. The formula for calculating the discriminant validity is as follows: c. Composite Reliability The composite reliability test aims at testing the reliability of an instrument within a research model. The construct will be consideed to have good reliability, or the questionnaire that has been implemented as the research instrument is already consistent, if both of the composite reliability score and the Cronbach's alpha score has been equal to or higher than 0.70 for all variables.

Structural Model or Hypothesis Test
In addition to the outer model measurement, the inner model measurement is the development of the theory-based and the concept-based model in analysing the relationship between the exogenous variable and the endogenous variable that have been elaborated in the conceptual framework. The test toward the structural model was evaluated by viewing theR-square value, which has been resulted from the goodness-of-fit model test. Then, the sequence of the outer model measurement was as follows: In performing the Goodness of Fit Model, the predictive-relevance (Q 2 ) was assigned. In terms of principle, the Q-square value that equals to zero or null indicates that the structural model has predictive relevance value. c. Path Coefficient Test The estimate value for the path relationship within the structural model should be significant. The significance value might be attained by means of bootstrapping procedure. Then, the significance of the hypothesis might be viewed by observing the parameter coefficient value and the T-statistics significance value on the algorithm bootstrapping report. The requirement is that the T-statistic significance value should be higher than 1.96.

RESULTS AND DISCUSSIONS
In this section, elaboration on the general description of the object altogether with the descriptive analysis on both the respondents and their responses will be provided. This elaboration might be consulted in the following section.

A. General Description of Research Object
OLX is the biggest platform of classified ad in the world tha provides a markeplace with high rate and dynamic growth for the local communities. OLX links the local communities to the online marketplace by means of smartphone or website for the interest of sale, purchase and exchange on the unused goods or service. In each month, hundred millions of people from any local community use OLX to search and sell numerous products, including computers, cellphones, furnitures, cars, properties, sport tools, service and many more (www.OLX.co.id).

Descriptive Statistics
The descriptive analysis within the study is conducted toward to aspects namely the respondents' characteristics and the respondents' responses in the questionnaire. The analysis on the two aspects might be consulted in the following sections.

Respondents' Characteristics
The characteristics of the respondents who have participated in the study are described based on four points namely gender, age, educational background, occupation status and monthly expenditure. The characteristics of the respondents based on the gender might be consulted in the Figure

Respondents' Responses
After the respondents' characteristics have been analysis, the respondents' responses should be analysed. The analysis on the responses is based on the three variables within the study namely Web Design, Brand Impression and Purchase Interest. The descriptive analysis on the responses of Web Design might be consulted in Table 6 below. Then, the descriptive analysis on the responses of Brand Impression might be consulted in Table 7 below. Last but not the least, the descriptive analysis on the responses of Purchase Interest might be consulted in Table 8 below.

Results of Data Analysis
As having been explained, for the data processing technique the Partial Least Square (PLS)-based Structural Equation Modelling (SEM) is implemented. Partial Least Square (PLS) is a Variance-Based Structural Equation Modelling analysis method in which the data processing relies on the use of Partial Least Square (PLS) Version 3.0 software. In addition, structural model test is also implemented in order to identify the relationship among the construct, the significance value and the R-square value for the construct of the dependent variables in the t-test as well as the significance of the structural path parameter coefficient. The results of each test might be consulted in the following sections.

Results of Outer Model Measurement
There are three sequences or three aspects within the conduct of outer model measurement namely convergent validity, discriminant validity and composite reliability. For the conduct of convergent validity test, the loading factor limit that has been applied is 0.50. Then, the results of the convergent validity test might be consulted in Table 9 below.  (2016) Based on the results in Table 9 above, it is apparent that two indicators are not valid within the variable Brand Impression (BI4) and Purchase Interest (PI6) because the loading factor of both indicators has been lower Journal of Marketing and Consumer Research www.iiste.org ISSN 2422-8451 An International Peer-reviewed Journal Vol.67, 2020 72 than 0.50. As a result, the data should be recalculated by dropping BI4 and PI6 from the model. After the first modification, it turns out that there is still one construct which loading factor value has been lower than 0.50 on its Average Variance Extracted (AVE) value. This construct is found in the variable Brand Impression (BI); consequently, second modification should be performed by dropping several indicators so that the loading factor value of Brand Impression might be higher than 0.50.
During the second modification, the first action that should be taken is recalculating the PLS Algorithm. Within the recalculation, there are two indicators that should be dropped from the construct of Brand Impression namely BI5 and BI8 so that the Average Variance Extracted (AVE) value might be higher than 0.50. The results of the final modification might be consulted in Figure 9 below. Source: PLS Processing Output Figure 9. Structural Model of Convergent Validity Then, for the discriminant validity test, the reflexive indicators might be observed in the cross-loading value between the indicators and the construct. An indicator will be considered valid if the indicator has the highest loading factor value relative to the targeted construct in comparison to the highest loading factor value relative to the other construct. The results of the discriminant validity test might be consulted in Table 10 below.  (2016) From the results in Table 10 above, it is apparent that the correlation between the constructs and the indicators of Web Design has been higher the correlation between the constructs of Web Design and the indicators from the other variables (Brand Impression and Purchase Interest). Similarly, it is also apparent that the correlation between the constructs and the indicators of Brand Impression has been higher than the correlation between the constructs of Brand Impression and the indicators from the other variables (Web Design and Purchase Interest). Last but not the least, it is apparent as well that the correlation between the constructs and indicators of Purchase Interest has been higher than the correlation between the constructs of Purchase Interest and the indicators of the other variables (Web Design and Brand Impression).
Another method for viewing the discriminant validity is an observation on the Square Root of Average Variance Extracted (AVE) for each construct and the correlation from one construct to another within the model. A construct is said to have high reliability if the AVE value of the construct is higher than 0.50. The results of the AVE test might be consulted in Table 11 below. Departing from the results in Table 11 above, it is apparent that all constructs of the variables have already met the requirements of reliability.
In addition to the construct validity test, the construct reliability test is also performed. The measurement tool for the conduct of the construct reliability test is composite reliability and Cronbach's alpha from the block of the indicators that measure the construct. In other words, a construct is measured from the block of the indicators that measure the construct. A construct, thus, is said reliable if both of the reliability score and the Cronbach's alpha score are higher than 0.70. The results of the composite reliability test might be consulted in Table 12 below.  (2016) Departing from the results in Table 12 and Table 13 above, it is clear that both of the composite reliability score and the Cronbach's alpha score for the variables Web Design, Brand Impression and Purchase Interest have been higher than 0.70. Therefore, it might be concluded that all of the constructs within the three variables have good reliability and thus the questionnaire that has been implemented for the data gathering activities in the study is already reliable or consistent.

Results of Inner Model Measurement
In addition to the outer model measurement, the inner model measurement is also conducted in order to view the relationship among the constructs, the significance values and the R-square values of the construct for the t-test dependent construct and also for the significance of the structural path parameter. The results of the Rsquare test might be consulted in Table 14 below. Journal of Marketing andConsumer Research www.iiste.org ISSN 2422-8451 An International Peer-reviewed Journal Vol.67, 2020 (2016) From the results in table 14 above, it is apparent that the R-square value is 0.576. The implication of such value is that the model has the moderate goodness-of-fit model. In other words, it might be further implied that the 57.60% variability on the Purchase Interest might be explained by Web Design and Purchase Interest whereas the remaining 42.40% variability on the Purchase Interest might be explained by the variables that have not been included into the study.
The goodness-of-fit model test within the structural model or the inner model measurement relies on the predictive relevance value (Q 2 ). The Q-square value that is higher than 0 (zer0 or null) indicates that the model has good predictive relevance value. On the other hand, the R-square value for the endogenous variable within the study might be attained from the following formula: The above calculation shows that the predictive relevance value is 0.576 or higher than zero (0.576 > 0.). The implication is that 57.69% variance on Purchase Interest (the endogenous variable) might be explained by the variables that have been assigned (Web Design and Brand Impression). In other words, the model is feasible to attain the relevant predictive value.
Last but not the least, the parameter significance that has been estimated in the hypothesis test provides very useful information with regards to the relationship among the research variables. The estimates value for the path relationship within the structural model should be significant. In order to identify whether the value is significant or insignificant, the bootstrapping method might be implemented and the bootstrapping procedures might be viewed from the t-statistics in the algorithm bootstrapping report. Specifically, the significant or the insignificant value might be traced into the t-table value at alpha 0.05 (5%) = 1.96 and its comparison on the tstatistics value. The final results of bootstrapping output for the structural model test might be consulted in Table  15 below.  (2016) From the results in Table 15, the t-statistics value from both independent variables are significance since the t-statistic value of Web Design (5.942) and the t-statistic value of Brand Impression (5.059) has been higher than the t-table value (1.960) at alpha 5%. Therefore, it might be concluded that both Web Design and Brand Impression has positive and significant influence on Purchase Interest.

Discussions
Based on the tests that have been administered, the influence of Web Design and Brand Impression on Purchase Interest of the consumers in OLX.co.id will be explained in the following sections.

Influence of Service Quality on Consumer Loyalty
The results of the hypothesis test within the study show that Web Design has positive and significant influence on Purchase Interest of the consumers in OLX.co.id online shopping site. The significance influence of Web Design might be traced from the highest t-statistic score namely 5.942. This t-statistic value is also higher than the t-table value 1.96 (H1 is accepted). In other words, it might be concluded that good and effective web design highly influence Purchase Interest rather than Brand Impression. However, it is found that the dimensions within the indicators of Web Design, specifically in the component "Appearance," has low score. Consequently, the web layout of OLX.co.id shopping site has not been completely internalized among the consumers.

Influence of Consumer Trust on Consumer Loyalty
The results of the hypothesis test within the study show that Brand Impression has positive and significant influence on Purchase Interest of the consumers in OLX.co.id online shopping site. This finding thus confirms that Brand Impression is indeed one of the important factors that influence the Purchase Interest of the consumers in the given online shopping site. The significance influence of Brand Impression might be traced from the t-statistic value that has been higher than the t-table value of the variable (5.059 > 1.960). Consequently, the second hypothesis that has been proposed in the study is accepted.
Then, departing from the description on the responses of the respondents, the highest score is found in the dimension of Brand Strength. The consumers perceive that this brand is able to be the market leader since this brand has competitive price and well-qualified brand, which have been one of the main factors that a company should have. In other words, it might be concluded that the consumers demand quality from a product or a service. However, based on the existing data, it is found that the scores of the Brand Impression in the indicators under the dimension of Uniqueness is lower than the scores of the Brand Impression in the indicators of the other dimensions. The implication is that OLX.co.id has not been an online shopping site that holds good brand uniqueness and product quality in comparison to the competitors.

CONCLUSIONS AND SUGGESTIONS
In this section, the conclusions of the study along with the suggestions for the given party will be provided. In the same time, several implications for the future researchers who would like to pursue any study on the similar topic will also be provided. Then, both of the conclusions and the suggestions might be consulted in the following sections.

A. Conclusions
The objective of the study is to analyse the influence of Web Design and Brand Impression on Purchase Interest of the consumers in the OLX.co.id online shopping site. Within the conduct of the study, the Partial Least Square (PLS) has been implemented. Then, based on the results of the test and the discussions within the study, the researcher would like to draw two conclusions. First, the results of the hypothesis test for the first hypothesis show that Web Design has positive and significant influence on Purchase Interest. The implication is that the better and the more interesting the Web Design of OLX.co.id is the higher the level of Purchase Interest will be in the OLX.co.id. Second, the results of the study for the second hypothesis show that Brand Impression has positive and significant influence on Purchase Interest. The implication is that an e-commerce company should have good impression among the consumers since Brand Impression highly influences Purchase Interest of the consumers within the utilization of the product or the service. Consequently, the company should create good Brand Impression in order to draw the attention of the consumers.

Suggestions
Departing from the discussions and the conclusions within the study, and also based on the descriptions of the responses provided by the respondents, several suggestions that might serve as a matter of suggestions for OLX.co.id online shopping site will be proposed for both the company and the future researchers. These suggestions might be consulted in the following sections.

For the Company
First of all, the OLX.co.id should focus on the improvement of the Web Design so that the consumers will always be retained in terms of Purchase Interest. The reason is that consumers perceive that OLX.co.id has not met their desires in comparison to the other online shopping sites. Then, from the calculation results on the index of Web Design (X1), it is found that WD1 has the lowest score from all indicators. Therefore, the OLX.co.id online shopping site should improve the design, the colour and the level of effectiveness so that the website will be more interesting so that the Purchase Interest of the consumers might be improved.
Furthermore, the OLX.co.id online shopping site should focus on the creation of good Brand Impression in front of the consumers. The reason is that many customers are still unsure about the change of the name and this situation certainly impacts their perception on the trust and the product, which have direct relationship to the Brand Impression of the OLX.co.id. As a result, the good Brand Impression of OLX.co.id does not always guarantee that the OLX.co.id brand will always be trusted by both the consumers and the potential consumers. In relation to the situation, from the calculation results on the index of Brand Impression (X2), it is found that BI4 has the lowest score from all indicators. The consequence is that the OLX.co.id should pay attention more on the sellers who will joint the brand so that the sellers will provide the well-qualified products in order to improve the brand quality. By doing so, the consumers will be satisfied and thus their purchase interest might be improved.