30

1. Introduction

The development of technology in this era nowadays is experienced by cooperatives. The cooperative is a business entity which its activities based on the principle of kinship (in the article 1 Constitutions of Republic Indonesia Number. 25/1992 regarding cooperatives). Some agency is forming the business entity of cooperatives in order to prosper the employees of the agency. The cooperatives are prospering the employees, or in this case the agency employees, by getting the remaining business results and facilitate savings and loans of members to cooperatives owned by the agency. One of the cooperatives in Indonesia is cooperatives owned by Indonesian Army called as Primkopad (Primary Cooperatives Army).

The implementation of the sale and purchase activity in Indonesian Army cooperatives is still using manual transactions which spread thorough Indonesia. The employee of Indonesian Army cooperatives is not only providing the tools or military equipment, but also foods and daily necessity. In its influences this cooperatives developed into minimarket that can be accessed by the entire citizen under the auspices of Indonesia Army. The next obstacle in the process of selling necessity or military equipment is not distributed clearly. Until, the necessity in one region is not much as the spreading necessity and equipment in Java Island or centre of the city. Indonesian Army is also should be able to developed and foster the welfare of wives with facility of UMKM under the auspices the wives union of Indonesian Army (Persit) to developed the business. It is not only as the superior sector in the country, but also as the sector that can compete with foreign parties.

A product can called success when the product is appeared and accepted, this thing cannot be separated of innovation product. The innovation product according to (Kanagal, 2015) is needed by the company to overcome the competitive pressure, taste changes and preference, product life cycle, technological progress (or vice versa, obsolete technology), various patterns of demand, and customer specific requirements. Innovation is the main tools to reaches the position of certain product to implement the function that expected to overcome the problem which appears in Primkopad. Therefore, it needs to do the innovation with utilize the platform e-commerce design.

Internet or World Wide Web (www) has greatly influenced business behavior, such as market, industrial and business. These are changing due to fulfill the economic demands and information technology (IT) which used to encourage business and market activities. In this era, internet becomes a strong communication mechanism and can facilitate the improvement and processing the business transactions. This is caused the changes of substantial in industry. Internet offers the opportunity to sell products of daily necessities directly to customers within the consumer market or consumers in the industrial market. The direct selling of goods and service from internet is called as ‘electronic commerce’. The definition of electronic commerce (e-commerce) according to (Laudon and Laudon, 2009) is a process of buying and selling the products electronically by consumer and from the company to company with computer as the business transaction intermediaries. E-commerce is useful in reducing administrative costs and business process cycle times, and improving relationships with both business partners and customers (Blut et al., 2015).

The existence of an e-commerce platform can create opportunities in terms of substantial sales increase for TNI-AD cooperatives. Planning an e-commerce system at Primkopad can bring opportunities to advance cooperatives and the big profit especially for Indonesia Army or general public. Moreover, in making e-commerce, it needs to have a structured system planning. In carrying out the planning, TNI-AD must collaborate and work in synergy with the application developers or the web developer.

Platform e-commerce which will be designed in this research is one of the business models called as business model canvas. Mapping a business using Business Model Canvas, will use nine contents or nine basic building blocks. (Osterwalder and Pigneur, 2012) stated that business model can be explained very well through nine basic building blocks that show you how to think about how companies make money. These nine building blocks are placed in an arrangement called the Business Model Canvas.

The model is divided into nine main parts which are; Customer Segments, Value Propositions, Channel, Customer Relationships, Revenue Streams, Key Resources, Key Activities, Key Partnerships and Cost Structure. Then, this part is divided into two sides left side (logic) and right side (creativity). In this research the business model canvas is used based on Indonesian Military Army (TNI-AD) since the target is members of the Army in the customer segment and main partnership.

Based on the above background the researchers formulated a research problem, Therefore, the researchers aimed to analyze the business model canvas on e-commerce design system based on TNI-AD.

2. Theoretical Review

2.1. Business Model Canvas

Business model canvas is a logic illustration regarding how a business in an organization creates, delivers and captures value (Osterwalder and Pigneur, 2010). This canvas divides the business model into 9 main components which then separated again into right (creative side) and left (logical side) component, just like the human brain. These nine components are consisted of (sorted from right to left). Customer Segment, Customer Relationship, Customer Channel, Revenue Structure, Value Proposition, Key Activities, Key Resource, Cost Structure, and Key Partners.

2.2. Electronic Commerce

Electronic trading or it is also called as e-commerce is the use of communication and computer networks to carry out business processes. A popular view of e-commerce is the use of the internet and computers with Web browsers to buy and sell products, (Pearson, 2008) Mc Leod Person 2008. E commerce is defined as buying and selling, marketing and product service, service and information on various types of computer networks. According to Shely Cashman 2007 (Vermaat, 2007) e-commerce is a business transaction that occurs in an electronic network, such as the internet. Whoever can access the computer, have the connection with internet and have a way to pay for the goods or services they buy, can participate in e-commerce.

2.3. Brand Awareness

Brand awareness plays an important role in buying interest by consumers because consumers tend to buy products that are familiar and popular (Keller, 1993; Macdonald and Sharp, 2000). Brand awareness can help the customer know the brands of product and implement the buying decision (Percy and Rossiter, 1992). Product with high brand awareness will receive choices from consumers which is also higher because it has a higher market share and quality evaluation (Dodds, Monroe and Grewal, 1991; Grewal et al., 1998).

2.4. Purchase Intention (Purchase Interest)

According to (Kotler, 2004), the purchase interest is a decision making to buy a brand among various brands. The buying interest arises through a variety of processes, including: introduction of needs, information search, information evaluation, and finally there will be an interest in buying that exists in consumers. While according to (Sumarwan, 2003), behavioral intentions will make a trust and attitude in conditions of to a particular product through alternative evaluations. While implementing the evaluation, product and brand selection will occur and brand according to consumer desires. In the other word, in those processes, the customer will compare between products.

3. Methodology

The methodology used is quantitative method with explanatory research. Explanatory is explaining the causal relationship between variables through hypothesis testing (Ferdinand, 2006). According to (Sanusi, 2011) causality research design is a research design that is structured to examine the possibility of causal relations between variables. The data type that used is qualitative and quantitative using questionnaire data technique which gives some questions to respondent (Ferdinand, 2006). The data resources or Respondents were consisted of internet and mobile application users among the Indonesian Army and asked to fill out questionnaires. While, the analysis unit in this research is individual because the answers of each respondent can represent their own opinions, the sampling is a purposive sample.

In this research, the analysis is used to identify the problems with quantitative approach and multiple regressions. Multiple regression analysis is the way to analyze using more than one free variable to explain the variant and variable related to the research (Sekaran and Bougie, 2017). The multiple regression formula is seen as:

31

32

4. Results and discussion

4.1. Y variable linear regression analysis

Variable

Independent:

X1: Framing the Market Opportunity

X2: Formulating the Market Strategy

X3: Designing the Customer Experience

X4: Crafting the Customer Interface

X5: Designing the Marketing Program

X6: Leveraging Customer Information through Technology

X7: Evaluating Marketing Program

Dependent:

Y: brand awareness

4.1.1. Classic Assumptions Test

1. Multicolinearity

The test result of multicolinearity is analyzed using SPSS software and obtained data as follow:

33

Based on the table above, it can be seen that VIF value for variables X1 to X7 is smaller than 10. Because the VIF value of all variables is no greater than 10, thus there is no multicollinearity in all of these independent variables.

2.Heteroscedasticity

Heteroscedasticity test is done by making Scatterplot (distribution flow) between residual and predictive value of the standardized dependent variable. Heteroscedasticity test results can be seen in the Scatterplot image, as in the image below:

From the figure above, it is seen that that the distribution of points does not form a certain pattern/path, so it can be concluded that heteroscedasticity does not occur or in other words homoscedasticity occurs, which is free from heteroscedasticity.

3. Normality

Normality test result can be seen from the figure of Normal P-P Plot below. It should be reminded that the normality assumption referred to in the classical assumptions of the OLS approach is the residual (data) formed by the linear regression model normally distributed, not the independent variable or the dependent variable. The criteria for a residual (data) are normally distributed and not using Normal P-P Plot approach but can be done by seeing the distribution points in the picture. If the distribution points approaching the straight line (diagonal) then it is said that (data) residual distributed normally, but if the distribution of these points away from the line thus it is not normally distributed.

Multicolinearity test

34

The distribution of points from the figure Normal P-P Plot above is relatively close to a straight line, then it can be concluded that (data) residual is distributed normally.

4.1.2. Feasibility Test Model

1. Model Reliability Test (F Test)

Reliability test model or feasibility test model or known as F test (some call it a simultaneous test model) is the initial stage of identifying a regression model that is estimated to be feasible or not. Decent (reliable) means the estimated model is appropriate to be used to explain the effect of the independent variables on the dependent variable. The name of this test is called as F test, because it follows the F distribution which criteria is tested as are like One Way Anova.

If the value of count prob. F (SPSS output is shown in column sig.) is smaller than the error rate (alpha) 0.05 (predetermined), then it can be said that the estimated regression model is feasible, whereas if prob. F value is bigger than error rate of 0,05 then the regression model is not feasible. The F test results can be seen in the table ANOVAa below. The value of Prob. F count, it can be seen in the last column (sig.).

Simultaneous Test

35

The value of prob. F count (sig.) in the table above is 0,000 which is smaller than the 0.05 significance level so it can be concluded that the estimated linear regression model is feasible to use to explain the effect of X1 to X7 on the Y variable (brand awareness).

2. Regression Coefficient Test (t Test)

T test in the Multiple linear regression is intended to test whether the parameters (regression coefficients and constants) which are thought to estimate equations/models of multiple linear regression are appropriate parameters or not. In this section, the t test is focused on the slope parameter (regression coefficient) only. So the intended t test is the regression coefficient test. The test results can be seen in the Coefficienta table as shown below:

If the value of prob. t counts (SPSS output is shown in column sig.) is smaller than the error rate (alpha) 0,05 (which has been specified). It indicated that independent variable (from the t count) has significant effect on the dependent variable, while if prob. t count is bigger than the error rate 0,05 then the independent variable has not the significant effect on the dependent variable.

From those table above, the value of prob. t count from independent variable X1, X3, X4, X5, and X6 is smaller than 0,05. Therefore, the independent variable of X1, X3, X4, X5 and X6 has significant effect on the Y dependent variable on 5% alpha or those variables has significant effect on the Y value 95% confidence level. While the variables X2 and X7 are significantly greater than 0.05 so the variables do not have a significant effect on the Y variable.

3. Determination coefficient

Determination coefficient explains the variation of the influence of independent variables on the dependent variable or the proportion effect of all independent variables on the dependent variable. The coefficient of determination can be measured by the value of R-Square or Adjusted R-Square. R-Square is used when there is only 1 free variable (commonly called Simple Linear Regression).

Determination Coefficient

Partial Test

36

If can be seen from the R-Square value of 0.980 that the proportion of the influence of the variables X1 to X7 to the Y variable is 98%. That is, the variables X1 to X7 have a proportion of influence on the Y variable by 98% while the remaining 2% (100% - 98%) are influenced by other variables that are not in this multiple linear regression models.

4.1.3. Models interpretation

1. Unstandardized coefficients

The interpretation that done in the regression coefficients is included into two things; sign and magnitude. The sign shows the direction of the relationship. Signs can be positive or negative. Positive sign shows the direct effect between independent variables on the dependent variable, while negative sign shows the opposite effect. The same direction means, if the independent variable increases, the dependent variable will experience the same increase.

The estimated linear regression model (equation) is as follows:

Y = -2,770 + 0.122X1 + 0,005 X2 +0,219 X3 + 0,122 X4 + 0,574 X5 + 0,372 X6 + 0,065 X7 + e

Positive X1 regression coefficient value indicated when X1 raises the value then Y will also increase its value. This was also similar to positive X2 regression coefficient values. When X2 increase so Y value will increases. The addition of X2 value of one unit may increase Y value of 0,005 units and vice versa, the decreasing of X2 value of 1 unit will decrease Y value of 0,005 units.

While for positive X3 regression coefficient value means when X3 increase the Y value will also increased. X4 regression coefficient with positive value has the same meaning with X4 regression coefficient. For X5 regression coefficient has positive value when X5 increase the Y value is also experience increase. X6 regression coefficient has positive value has the same meaning with X6 regression coefficient. X7 regression coefficient has positive value and same meaning with X7 regression coefficient.

2. Z Variable Linear Regression Analysis

Variable Independent:

X1: Framing the Market Opportunity

X2: Formulating the Market Strategy

X3: Designing the Customer Experience

X4: Crafting the Customer Interface

X5: Designing the Marketing Program

X6: Leveraging Customer Information through Technology

X7: Evaluating Marketing Program

Variable Dependent:

Z: purchase intention

a. Classic Assumption Test

1. Multicollinearity

The multicolliniearity test result using SPSS software is obtained as follow:

37

Based on the table above it can be seen that VIF value for X1 to X7 variable is smaller than 10, because the VIF value from all variables is not bigger than 10. Therefore, the multicollinearity did not happen to the entire independent variable. One of the conditions of classic assumption of regression linear with OLS is a good linear regression model which free from multicolinearity.

2. Heteroscedastiscity

The heteroscedasticity test result can be seen in the figure Scatterplot, as in the figure below:

From the figure above it is seen that distribution of point cannot making a certain pattern, until it is concluded as there is no heteroscedasticity or in the other words there is the appearance of homoscedasticity.

3. Normality

The normality test result can be seen from the figure Normal P-P Plot below. It needed to realize that normality assumption within the classic assumption approach of OLS is the residual data which formed by a normally distributed linear regression model, it is not an independent variable or dependent variable. If the distribution of the points approach on a straight line (diagonal) then it is said that the residuals data are normally distributed, but if the distribution of these points are away from the line then it is not normally distributed.

Multicollinearity Test

38

The distribution of points from the Normal P-P plot above is relatively close to a straight line, so it can be concluded that data are normally distributed.

b. Feasibility Test Model

1. Reliability Test Model (F Test)

Reliability test Model or feasibility test model or referred as F test (some call it a simultaneous test model) is the initial stage of identifying a regression model that is estimated to be feasible or not.

If the value of count prob. F (SPSS output is shown in column sig.) is smaller than the error rate (alpha) 0.05 (predetermined), then the regression model is feasible, whereas if prob. F value is bigger than error rate of 0,05 then the regression model is not feasible. The F test results can be seen in the table ANOVAa below. The value of Prob. F count, it can be seen in the last column (sig.).

The table above showed the value of prob. F count (sig.) is 0,000 smaller than the 0.05 which is significance level. So it can be concluded that the estimated linear regression model is feasible and can be used to explain the effect of X1 to X7 on the Z variable (purchase intention).

2.Regression coefficient test (t test)

The t test in multiple linear regression is intended to test whether the parameters (regression coefficients and constants) which are thought to estimate equations/models of multiple linear regression are the right parameters or not. Parameters that estimated in linear regression include the intercepts (constants) and slopes (coefficients in linear equations). In this section, the t test is focused on the slope parameter (regression coefficient) only. So the intended t test is the regression coefficient test. The test result can be seen in the table of coefficienta as below:

Simultaneous Test

39

If the prob. t count value (SPSS output is shown in column sig.) is smaller than the error level (alpha) 0,05 (which has been determined) then it can be said that independent variable (from the t count) has significant effect on the dependent variable, while if the independent variable is bigger than error level 0,05 then it can be said that independent variable has not significant effect on dependent variable.

From the table above it can be seen that prob. t count value from the independent variable X1, X3, X4, X5, X6 and X7 is smaller than 0,05 and indicate that the independent variables have significant effect towards the dependent variable X2, the significant value is bigger than 0,05 until it can be said that the X2 variable is not give the significant effect on variable Z.

3. Determination coefficient

The coefficient of determination explains the variation of the influence of independent variables on the dependent variable. The determination coefficient value can be assessed to R-Square value or Adjute R-Square. R-Square is used when only 1 independent variable (commonly called Simple Linear Regression), whereas Adjusted R-Square is used during the independent variable is more than one. To counting the determination coefficient value, the researcher prefers to use R-Square rather than Adjusted R-Square, even though the independent variable is more than one.

40

If seen from the R-Square value of 0.961, it shows that the proportion of the influence of the variables X1 to X7 to the Z variable is 96.1%. it means, variable X1 until X7 has the proportion effected on the variable Z of 96,1% whereas the rest of it is 3,9 % (100% - 96,1%) is influenced by another variable which is not in this multiple linear regression model.

c. Interpretation Model

The interpretation that done through the regression coefficient is included in two things, sign and magnitude. The sign shows the direction of the relationship. It can be positive or negative. Positive sign shows the direct effect between independent variables on the dependent variable, while negative sign shows the opposite effect. The same direction means, if the independent variable increases, the dependent variable will experience the same increase.

The estimated linear regression model (equation) is as follows:

Z = 1,532 + 0,079X1 + 0,041 X2 +0,320 X3 + 0,206 X4 + 0,289 X5 + 0,580 X6 + 0,146 X7 + e

The regression coefficient X1 is positive, meaning that when X1 goes up, the value of Z will also increase. The positive regression coefficient X2 has the same meaning as the regression coefficient X1. When X2 increases, the value of Z will increase. Adding an X2 value of 1 unit will increase the Y value by 0.041 units and vice versa, decreasing the X2 value by 1 unit will decrease the Z value by 0.041 units.

For the regression coefficient X3 is positive, meaning that when X1 goes up, the value of Z will also increase. The positive regression coefficient X4 has the same meaning as the regression coefficient X4. For the regression coefficient X5 is positive, meaning that when X5 goes up, the Z value will also increase. The positive regression coefficient X6 has the same meaning as the regression coefficient X6. The positive regression coefficient X7 has the same meaning as the regression coefficient X7.

3.Multiple Linear Regression Analysis of Variables Y and Z

Variable

Independent:

X1: Framing the Market Opportunity

X2: Formulating the Market Strategy

X3: Designing the Customer Experience

X4: Crafting the Customer Interface

X5: Designing the Marketing Program

X6: Leveraging Customer Information through Technology

X7: Evaluating Marketing Program

Y: Brand Awareness

Dependent:

Z: Purchase Intention

a. Classic Assumption

1.Multicolinierity

The multicolinearity test is using SPSS software is obtained as follow:

41

Based on the table above, it can be seen that VIF value for the variable of X5, X6 and Y is bigger than 10 so those variables are contain multicollinearity and must handled. The X1, X2, X3, X4 and X7 variables with small VIF value which is lower than 10 did not have multicolinearity. One of the classical assumptions of linear regression with OLS is a good linear regression models it is independent from the multicolinearity.

2. Heteroscedasticity

Heteroscedasticity test is done by making Scatterplot (distribution flow) between residual and predictive value of the standardized dependent variable. Heteroscedasticity test results can be seen in the Scatterplot image, as in the image below:

From the figure above, it is seen that that the distribution of points does not form a certain pattern/path, so it can be concluded that heteroscedasticity does not occur or in other words homoscedasticity occurs, which is free from heteroscedasticity.

3. Normality

The normality test result can be seen from the figure Normal P-P Plot below. It needed to realize that normality assumption which means in classic assumption approach of OLS is residual data which formed a normally distributed linear regression model, which is not an independent variable or dependent variable. If the distribution points approach or meet on a straight line (diagonal) then it is said that the data are normally distributed, but if the distribution of these points away from the line then it is not normally distributed.

42

The distribution points from the Normal P-P plot above is relatively close to a straight line, so it can be concluded that the residuals data are normally distributed.

b. Feasibility Test Model

1.Reliability test Model (F test)

Reliability test Model or feasibility test model or known as F test (some call it a simultaneous test model) is the initial stage of identifying a regression model that is estimated to be feasible or not.

If the value of count prob. F (SPSS output is shown in column sig.) is smaller than the error rate (alpha) 0.05 (predetermined), then can be said that the estimated regression model is feasible, whereas if prob. F value is bigger than error rate of 0,05 then that regression model is not feasible. The F test results can be seen in the table ANOVAa below. The value of Prob. F count, it can be seen in the last column (sig.).

The value of prob. F count (sig.) 0,000 is smaller than the 0.05 significance level which indicated that the estimated linear regression model is feasible to use to explain the effect of X1 to X7 on the Z variable (purchase intention).

2. Regression coefficient test (t test)

The t test in multiple linear regression is intended to test whether the parameters (regression coefficients and constants) which are thought to estimate equations/models of multiple linear regression are the right parameters or not. The estimated parameters in linear regression are consisted of intercepts (constants) and slopes (coefficients in linear equations). In this section, the t test is focused on the slope parameter (regression coefficient) only. So the intended t test is the regression coefficient test. The test result can be seen in the table of coefficienta as in the table below:

If the prob. t count value (SPSS output is shown in column sig.) is smaller than the error level (alpha) 0,05 (has been determined) then it can be said that independent variable (from the t count) has significant effect on the dependent variable, while if the independent variable is bigger than error level 0,05 then is showed that independent variable has not significant effect on dependent variable.

From the table above, the prob. t count value from the independent variable of X4, X5 and Y is smaller than 0,05 which lead to the significant effect on the dependent variable Z in alpha 5% or in other word X4, X5 and Y have significant effect toward the Z value within the confidential standard of 95%. Whereas the variables of X1, X2, X3, X6, and X7 have its own significant value bigger than 0,05which showed that those variables did not give significant effect on Z variable.

3.Determination coefficient

The determination coefficient explains the variation of the influence of independent variables on the dependent variable. The determination coefficient value can be assessed to R-Square value or Adjusted R-Square. R-Square is used when there is only 1 independent variable (commonly called Simple Linear Regression), whereas Adjusted R-Square is used during the independent variable is more than one. To counting the determination coefficient value, the researcher prefers to use R-Square rather than Adjusted R-Square, even though the independent variable is more than one.

43

If seen from the R-Square value of 0,980, it shows that the proportion of the influence of the variables X1 to X7 to the Z variable is 98%. it means, the variable X1 until X7 has the proportion effected on the variable Z for 98,1% whereas the rest of it is 2 % (100% - 98%) is influenced by another variable which is not in this multiple linear regression model.

c. Interpretation Model

The interpretation that done through the regression coefficient is included of two things, namely; sign and magnitude. The sign shows the direction of the relationship on whether it is positive or negative. The positive sign shows the direct effect between independent variables on the dependent variable, while the negative sign shows the opposite effect. The direct direction means that there is similar increase on both independent and dependent variable.

The estimated linear regression model (equation) can be seen as follows:

Z = 4,392 - 0,046X1 + 0,036 X2 +0,094 X3 + 0,080 X4 - 0,303 X5 + 0,196 X6 + 0,080 X7 + 1,033 Y + e

The regression coefficient X1 is negative, meaning that when X1 increases, the value of Z will decrease. The positive regression coefficient X2 has the same meaning as the regression coefficient X1. For the regression coefficient X3 is positive it means that when X3 raises the value of Z will also increase. The positive regression coefficient X4 has the same meaning as the regression coefficient X4. For the regression coefficient X5 is negative, meaning that when X5 rises, the value of Z will decrease.

X6 regression coefficient is positive value means that when X1 raises the value of Y will also increase. The positive regression coefficient X6 has the same meaning as the regression coefficient X6. The positive regression coefficient X7 has the same meaning as the regression coefficient X7. Y regression coefficient is positive values that have the same meaning with Y regression coefficient. When Y increases so Z value will increases. The addition of Y value of1 unit will increasing Z value of 1,033 unit and vice versa, the decreasing of Y value of 1 unit will decreasing Z value of 1,033 units.

4. Conclusion

a) Internet marketing analysis of brand awareness (variable Y)

1. There is a significant and positive joint effect of variables in the Framing the Market Opportunity, Formulating the Market Strategy,: Designing the Customer Experience, Crafting the Customer Interface, Designing the Marketing Program, Leveraging Customer Information Through Technology and Evaluating Marketing Programs affecting brand awareness.
2. There is a significant effect of Framing the Market Opportunity on brand awareness.
3. There is no significant effect of Formulating the Market Strategy on brand awareness.
4. There is a significant effect of Designing the Customer Experience on brand awareness.
5. There is a significant effect of Designing the Marketing Program on brand awareness.
6. There is a significant effect of Leveraging Customer Information Through Technology on brand awareness.
7. There is no significant effect of Evaluating Marketing Program influences brand awareness.

b) Internet marketing analysis on purchase intention (variable z)

1. There is a significant and positive joint effect of variables in Framing the Market Opportunity, Formulating the Market Strategy, : Designing the Customer Experience, Crafting the Customer Interface, Designing the Marketing Program, Leveraging Customer Information Through Technology and Evaluating Marketing Program affecting purchase intension.
2. There is a significant effect of Framing the Market Opportunity on purchase intension.
3. There is no significant effect of Formulating the Market Strategy on purchase intension.
4. There is a significant effect of Designing the Customer Experience on purchase intension.
5. There is a significant effect of Crafting the Customer Interface on purchase intension.
6. There is a significant effect of Designing the Marketing Program on purchase intension.
7. There is a significant effect of Leveraging Customer Information through Technology on purchase intension.
8. There is a significant effect of Evaluating Marketing Program affecting purchase intension.

c) Internet marketing analysis on brand awareness &purchase intention (Y & Z)

1. There is a significant and positive joint effect of variables in Framing the Market Opportunity, Formulating the Market Strategy, : Designing the Customer Experience, Crafting the Customer Interface, Designing the Marketing Program, Leveraging Customer Information Through Technology and Evaluating Marketing Program affecting purchase intension.
2. There is no significant effect of Framing the Market Opportunity on purchase intension.
3. There is no significant effect of Formulating the Market Strategy on purchase intension.
4. There is no significant effect of Designing the Customer Experience on purchase intension.
5. There is a significant effect of Crafting the Customer Interface on purchase intension.
6. There is a significant effect of Designing the Marketing Program on purchase intension.
7. There is no significant effect of Leveraging Customer Information through Technology on purchase intension.
8. There is no significant effect of Evaluating Marketing Program on purchase intension.
9. There is a significant effect of brand awareness on purchase intension.

Based on the matter above, the theory concept of e-marketing on brand awareness and purchase intention has been proofed to be applied in army. This happens because the Army organization is the government organization and the cooperatives inside are still selling the product to a limited circle. So it needs to be done a test and an application on the concept of internet marketing which can expand the market share and affect people’s purchasing power on the product sold at Indonesian Military Army which then it can add profit and value for organization.

44

45

46