Determinant Factors of Customers Switching Behavior to Customer Satisfaction and Loyalty in Online Transportation Users in Bandung

This study was aimed to find out which factors that become the most influence on customers switching behavior for online transportation and how the impact on their satisfaction and loyalty for future consumption. Transportation service is one of the service industry sectors that play a strategic role in human life. The intense competition in the online transportation industry and the various choice of brands in the market make the consumers easy to switch from their current product to other brand products. The research method used in this study was a quantitative method, with Structural Equation Modeling (SEM) analysis technique using SMART PLS 2.0 software. The sampling method used was accidental sampling with 400 respondents. The results of the study showed that the contribution of price, promotion and eservice quality simultaneously influenced on which directly affected customer satisfaction was 64.9%. Whereas, the results of the study also showed the contribution of price, promotion, e-service quality and customer satisfaction simultaneously influenced on which directly affected customer loyalty was 48.3%. E-service quality has the biggest impact on customer satisfaction by 30.69%; meanwhile, promotion has the biggest impact on loyalty by 3.17%. Info Article History Article: Submitted 12 October 2019 Revised 2 December 2019 Accepted 4 December 2019


INTRODUCTION
According to the Indonesian Mobile Phone Association, or Asosiasi Penyelenggara Telepon Seluler Seluruh Indonesia (ATSI) in Antara News (2010) stated that 180 million Indonesian people became the regular customer of mobile phone service, which means it was about 60% of Indonesian population already owned a telecommunication device. Based on an APJII survey in 2017 showed that 67.8% of internet users in Indonesia used their smartphone to access the internet, -14.7% computer or PC users, and 3.8% tablet users.
Meanwhile, based on a survey conducted by Internet Service Users Association, or in Indonesia called Asosiasi Pengguna Jasa Internet (APJII) in 2016, the penetration of internet users in Indonesia reached 132.7 million people, from the total population 256.2 million people, with 52.5% was men and 47.5% was women, where 65% was in Java island. Thus, it means 52% of the total population in Indonesia already used the internet in their daily life. From the explanation above, it can be concluded that internet can not be separated from the daily life activities of Indonesian society. In addition to data from "We Are Social" in January 2018, the average time used by Indonesian people to access the internet from all media was 8 hours 51 minutes (Global Web index Q2 and Q3, 2017).
The most mobile application accessed by people in Indonesia, according to data "We Are Social" in January 2018, can be categorized based on types, such as messenger, social media, browser and online transportation. In the category of chatting application, Whatsapp Messenger was on the first top list. Meanwhile, for the social media category, Facebook was on the first top list. In addition to online transportation category, GO-JEK was the online transportation application mostly accessed by people in Indonesia.
The loyalty of online transportation users is influenced by swinger behavior (the shift of brand usage quickly), which is a phenomenon that cannot be avoided by an online transporta-tion service provider. According to the Jakpat survey, it was stated that 40.96% of online transportation service users used Gojek and Grab at the same time; therefore, the users always compared to several providers before deciding to choose the provider.
Transportation service is one of the strategic components in economy growth distribution. The functions of the transportation service industry such as to help the mobilization of humans and goods, and the information and financial flow, which need to be managed fast and accurately to fulfill the punctuality demand. The growth of the transportation sector will directly show economic growth; thus, transportation has a crucial and strategic role, either in macro or micro. With the development of technology and the internet makes business people utilize it as an opportunity to support dynamic people mobility, which demands a practical and fast transportation mode by providing transportation service only through a smartphone. At the beginning of 2015, GO-JEK began booming in Indonesia as the innovation of online transportation and firstly launched its application on Android and IOS.
Brand switching is one of customer behavior phenomena where a customer switches his/ her loyalty from one brand to other brands. The intense competition in the industry and various choices in the market make the customers easily switch from the previous brand to other brands. Brand switching occurs because of some factors, either internal or external. Internal factors are such as the change of customer interest, consumptive lifestyle, and the change of customer preference to the product. Meanwhile, the external factors such as various brand choices, promotions from other brands, social factors, recommendation factors from the reference group, and the development of technology and media.
Online transportation users tend to do brand switching from one service provider to another service provider to make themselves more satisfied. Based on the result analysis of ecommer-ceIQ released by the Jakarta Post, the users of online transportation application was dominated by women (54% from the total users). Meanwhile, the reasons why they chose online transportation provider were such as safety (26%), easy to find driver (23%), promotion and discount (22%), user-friendly application (16%), payment choice (5%), food delivery (3%), customer service (3%) and customer bonus (2%) (www.goodnewsfromindonesia.id, 2018).
This research was aimed to find out which factors were the most influencing on customers switching behavior for online transportation, and how its impacts on their satisfaction and loyalty for future consumption. The researchers focused on analyzing external factors; therefore, it was expected that through the results of this study, it helps the service providers to increase their competitive advantages. To determine external factors of online transportation switching behavior, the researchers conducted a pre-research survey in Bandung community. The results of a pre-research survey showed that price (38%), sales promotion (29%), service quality (17%), user-friendliness of application (9%), and brand trust (7%).
Based on the causative factors of switching behavior in online transportation customers, it can be analyzed the level of customer satisfaction and the tendency to switch to other brands. If the customers feel satisfied with the product and service, then the loyalty possibility will be higher. Customer satisfaction will positively correlate to customer loyalty but negatively correlate to intention to switch (Mohzan et al., 2011). The more intense the competition of online transportation service in Indonesia, especially in Bandung, the more vigorous the promotion from the online transportation industry is, which will trigger the customers easily switch to other brands. Therefore, this study was aimed to study further about customers behaviors to online transportation, especially in Bandung.

Hypothesis Development The Relationship of Switching Behavior to Customer Satisfaction
Brand switching is a behaviour that is performed by a customer because of some certain factors or can also be defined as a vulnerability where the customer will switch to other brands (Althonayan et al., 2015). The occurrence of brand switching behavior can be influenced by some factors, such as price, sales promotion, and service quality (Althonayan et al, 2015;Wimalasiri, 2017;Ghasrodashti, 2018). Through this study, it was expected to reveal the degree of brand switching within the online transportation services in Bandung, as well as to examine the activities on brand switching behaviour.
Switching behavior can also be influenced by motivation, either intrinsic or extrinsic factors (Shukla, 2009). Satisfaction comes from the differences in a person's initial expectations and product performance (Sharma & Panga, 2018). In addition, dissatisfied customers can degrade the customer base and the organization's image (Levesque & McDougall, 1996;Calvo-Porral & Lévy-Mangin, 2015;Pantano & Vasilios, 2016). On the other hand, it is crucial to study the relationship between customer switching intention and customer satisfaction, because the result of previous studies showed that the dimension of switching behavior influenced on customer loyalty.
Price determination is a reason makes customer switch to other brands, that consists of price comparisons, fees, charges, penalties and agreement on promotion price (Kotler & Amstrong, 2016). Price has been considered to be a marker of quality in service businesses (Wilson et al., 2012). The previous studies confirmed that price is considered to be key element in determining customer satisfaction (Shin & Kim, 2008;Mohsin et al., 2012). Therefore, in this study was proposed the following hypothesis: H1 : There is a significant influence of price on customer satisfaction.
Sales promotion is a way to influence customers to directly purchase goods with certain brands; thus through a good sales promotion will influence the customer in brand switching (Kotler & Amstrong, 2016). Sales promotion is related to contact between the company and its customers (Al-Kwifi & Ahmed, 2015). Whereas, the main factors such as sales promotion offer is crucial to determine the consumer behavior. In addition, if the sales promotion factor is missing, it may lead to dissatisfaction (Chen & Chuang, 2008;Elder & Krishna 2011). Therefore, the following hypothesis was proposed: H2 : There is a significant influence of sales promotion on customer satisfaction.
Service quality in brand switching behaviour is also influenced by the quality perception of customer. The better quality of a new product offered can influence on the customers to do brand switching. Service quality is an essential element to succeed in the competitive environment (Kuo et al. 2009;Bigné et al. 2011;Calvo-Porral & Lévy-Mangin, 2015). Service quality is positively correlated with customer satisfaction, although service recovery that helps to address service failures and customer complaints appears to be the most attractive antecedent for customer satisfaction (Qin et al., 2010). Based on the explanation above, the hypothesis is as follows: H3 : There is a significant influence of e-service quality on customer satisfaction.

The Relationship of Customer Satisfaction to Customer Loyalty
Customer satisfaction is a feeling of someone, either happiness or disappointment, resulted from product comparison or service experience based on the expectation (Kotler & Keller, 2016). If service or experience is far from the expectation, then the customer will not feel satisfied. Meanwhile, if it is as what is expected, then the customer will feel satisfied.
Customer satisfaction is a level of someone feeling after comparing service or result from what is felt compared to what is expected (Tjiptono, 2014). Customer loyalty is as a customer commitment to a brand, store or distributor based on very positive behaviour and displayed in a consistent repurchasing.
Satisfaction is an important step to form loyalty. To increase customer loyalty, satisfaction is one of the factors that need to be considered (Akbar & Parvez, 2009;Kassim & Asiah, 2010;Van-Vuuren et al., 2012;Rimawan et al., 2017). If the customers feel satisfied then it will increase their loyalty. Based on the studies previously explained, it can be concluded that customer satisfaction is a feeling experienced by a customer, after comparing what is expected to what is resulted. If the service is more than what is expected, the customer will feel very satisfied, which then will impact on loyalty.
A customer can be identified as a loyal customer to a company when the customer shows a certain purchasing behaviour that is marked by an interval based on a series of decision. Loyalty is marked by a situation that shows how long the customer uses the product accompanied by the purchasing more than twice. Based on explanation above, the hypothesis is assume as: H4 : There is a significant influence of customer satisfaction on customer loyalty.

The Relationship of Switching Behavior to Customer Loyalty
In brand switching, a customer does not pass the phases of conviction, and a normal attitude or behavior. The customer does not extensively find out about the brand, but act as a passive information catching. The customer does not create a brand conviction but choose the brand because of brand familiarity. The relationship between price and brand loyalty receives great considerable attention (Ferreira & Coelho, 2015). The study focused on appropriate additional questions about prices such as price elasticity and brand loyalty. From the statements above, it is clearly stated that one of switching behavior factors, -price, can influence a customer to keep using a product. Based on the explanation above, the hypotheses drawn is as follows: H5 : There is a significant influence between Price to Customer Loyalty.
Brand switching is a moment where a customer or group switches loyalty from a certain brand product to other brands. Thus, in other words, brand switching is an action that is performed by a customer to switch from the current brand to another new brand. Promotion can be seen as a way of closing the information gap bet-ween potential sellers and buyers ( Johnson et al., 2008). Promotion is a part of a specific effort to encourage customers to tell others about their services. Moreover, there is also a strong and positive influence of sales promotion on customer loyalty (Zeithaml et al., 1985;Owomoyela et al., 2013;Ameur et al., 2015). Thus, the hypothesis proposed is as follows: H6 : There is a significant influence of sales promotion on customer loyalty.
The approach of switching behavior operationalizes loyalty in four ways. First, through action based on the actual consumption of goods or services. This approach usually combines volume and frequency of purchasing during one certain period (Lin et al., 2000;Veloutosou et al., 2004). Second, steps that are aimed to a proportion of consumption in a set of certain good or service is in certain market or even retail location assigned. The concept of brand loyalty is included in this type. Third, steps that are based on the probability of repeat buying. Fourth, the steps of testing the point where the customer switches to another brand (Tsao & Li-Wei, 2005). Therefore, Concluded that service quality has a significant effect on gaining higher customer satisfaction to achieve more loyal customers in target markets (Hussain, 2016;Ashraf et al., 2018). Thus, the following hypothesis is stated: H7 : There is a significant influence of e-service quality on customer loyalty.
As for the framework used in this study can be seen in Figure. 1.

METHOD
This study is included in descriptive verification with a quantitative approach. The main aim of descriptive study is designed to gain data that describes characteristics of interesting topics in a study (Zikmund, 2015). Verificative research is a study that aims to tests the truth of something been created in hypotheses through statistical calculations to find out the relationship between two variables or more (Sekaran & Bougie, 2013). The population of this study was all online transportation services users living in Bandung.To determine the proportion of sample size, the Bernoulli formula was used (Zikmund, 2010) with the level of error 5%. It was calculated using the Bernoull formula, then the sample size in this study was 384,16 (»400) respondents.

Respondents Characteristics
In this study, the respondents taken were the customers of online transportation service in Bandung, that were about 400 respondents. The characteristics of respondents in this study are; based on gender is women by 59%, with age range is between 20-25 years old by 48%, current education is Senior High by 49%, the most occupation is student by 48%, with income per month < Rp2.000.000 by 38%, and the frequency of using online transportation is 1-3 times in a week by 67%.

Testing Result of Outer Model
The outer model was implemented to define how each indicator relates to their latent variables. In testing this outer model, it is needed to do the test of validity and reliability for statement indicators used. The testing was conducted using SmartPLS 2.0 software. In addition, the outer moder in this study can be seen in Figure.2.
The test of convergent validity will be fullfilled if the value of loading factor in each indicator > 0.5. The following is the result of convergent validity test from all indicators: Based on Table 1, it can be seen that all indicators used were valid because they have the value of loading

Figure 2. Outer Model Structural Equation Modelling
To conduct a validity test, there are two types of tests conducted, namely convergent validity and discriminant validity. The convergent validity test will be met if the loading factor value on each indicator is > 0.5. The indicator used to declare the validity in the discriminant test is if the value of the loading factor obtained is the highest value to the intended variable compared to the loading factor to other variables. Following are the result of the convergent validity test conducted on all indicators. Table 2 shows that all indicators were valid. It was performed by looking at the value of the loading factor that exists on each indicator of each variable. If the loading factor value achieved is the highest value on a predetermined variable compared to the loading factor on other variables, then the indicator is declared valid. For example, in the reference to others indicator on the customer loyalty variable, it showed that the value (0.8414) was higher compared to the value in the variable price (0.4994), to variable promotion (0.5166), to variable e-servqual (0.5648), and customer satisfaction (0.5905).
In addition to the validity test, the test was also performed to all variables in this study. The test was conducted by comparing the value of composite reliability and Cronbach's alpha contained in each variable. The value that must be met to make all variables are declared to be reliable is > 0.7, either for composite reliability or Cronbach's alpha.  Table 3 shows that all variables used were reliable because the values met the requirement, -all values of composite reliability were greater than 0.7, and all variables had value of Cronbach's alpha were greater than 0.7. Thus, all indicators and variables, either for composite reliability and cronbach alpha, can be used for the testing of inner model.

Testing Result of Structural Inner Model
The inner model test was performed to find out the relationship between construct, significance value and R 2 of the research model conducted. It was processed by taking into account the value of R 2 in endogenous latent variables and t-value of each exogenous latent variables to the bootstrapping result of endogenous latent variables. Then, the next process was to see whether there was a significant and close relationship between independent variables and dependent variables. The path diagram of inner model can be seen in following Figure 3.

Figure 3. Inner Model of Structural Equation Modelling
Based on Figure 3, it can be seen the t-value of each exogenous latent variables to endogenous latent variables. To conduct this test, twotail test with 5% error level was implemented. Thus, critical value that must be met in the hypothesis test is 1.96 (Ghozali, 2011). If t-value is greater than critical value (1.96), then there is significant influence between exogenous latent variables to endogenous latent variables. The values result can be seen in Table 4.

SEM Model 1
In this model, there are three independent variables, namely price (X 1 ), promotion (X 2 ), e-service quality (X 3 ) and one dependent variable, customer satisfaction (Y). The equation model of first sub-structure can be seen in the following Figure 4. For the equation of first model, the formula used is as follows: Y = 0.286 X 1 + 0.065 X 2 + 0.554 X 3 + 0.351 Based on the result of analysis calculation (Table 5), then it can be taken some objective information as follows: The contribution of price (X 1 ) that has a direct influence on customer satisfaction (Y) was 0.286 2 = 8.18%; The contribution of promotion (X 2 ) that has a direct influence on customer satisfaction (Y) was 0.065 2 = 0.42%; The contribution of e-service quality (X 3 ) that has a direct influence on customer satisfaction (Y) was 0.554 2 = 30.69%; The contribution of price (X 1 ), promotion (X 2 ), and e-service quality (X 3 ) that has a direct and simultenous influence on customer satisfaction (Y) was 0.649 = 64.9%; The rest of 35.1% was influenced by other factors that can not be explained in this study.

SEM Model 2
In this model, there are five independent variables, namely price (X 1 ), promotion (X 2 ), eservice quality (X 3 ), customer satisfaction (Y) and one dependent variable, customer loyalty (Z). The equation model of second model can be seen in the Figure 5.
For SEM Model 2, the formula used is as follows: Z = 0.137 X 1 + 0.178 X 2 + 0.005 X 3 + 0.472 Y + 0.517 Based on the analysis calculation in Table  6, it can be taken some objective information as follows: The contribution of price (X 1 ) that has a direct influence on customer loyalty (Z) was 0.137 2 = 1.88%; The contribution of promotion (X 2 ) that has a direct influence on customer loyalty (Z) was 0.178 2 = 3.17%; The contribution of e-service quality (X 3 ) that has a direct influence on customer loyalty (Z) was 0.005 2 = 0.003%; The contribution of customer satisfaction (Y) that has a direct influence on customer loyalty (Z) was 0.472 2 = 22.28%; The contribution of price (X 1 ), promotion (X 2 ), e-servqual (X 3 ), and customer satisfaction (Y) that has a simultaneous and direct influence on customer loyalty (Z) was 0.483 = 48.3%; The rest of 51.7% was influenced by other factors that cannot be explained in this study.
The following is the summary of indirect influences between independent variables (X 1 , X 2 , X 3 , Y) and dependent variable (Z). The indirect influence of price (X 1 ) on customer loyalty through customer satisfaction (X 1 , Y, Z) was 6.47%. In addition, the indirect influence of promotion (X 2 ) on customer loyalty through customer satisfaction (X 2 , Y, Z) was 8.40%. Meanwhile, the indirect influence of e-service quality (X 3 ) on customer loyalty through customer satisfaction (X 3 , Y, Z) was 0.002%.

The Influence of Price on Customer Satisfaction (H1)
The direct relationship between price and customer satisfaction was obtained is that H0 was rejected, which means H1 was accepted. The t-value was 6.158, which was greater than the critical value (1.96); thus, it can be concluded that the price variable influenced significantly on customer satisfaction. The value of the coefficient parameter of price to customer satisfaction was 0.286, which means price has a positive influence on customer satisfaction that obtained about 0.286. The results of this study support a previous study (Althonayan et al., 2015), who stated that service pricing has a positive association with customer satisfaction. Besides, some studies also showed that the price of a product has been considered as one of the important variables to satisfaction (Shin & Kin, 2008;Mohsin et al., 2012;Sharma & Panga, 2018).

The Influence of Promotion on Customer Satisfaction (H2)
The direct relationship between promotion and customer satisfaction was obtained is that H0 was accepted, which means H2 was rejected. It was a result of the t-value was 1.490, which was lower than the critical value (1.96); thus, it can be concluded that promotion did not influence significantly on customer satisfaction. The value of the coefficient parameter of promotion to customer satisfaction was 0.065, which means promotion has a positive influence on customer satisfaction, which obtained only 0.065. The previous studies, which support this study result stated that promotion variable does not affect the customer satisfaction (Kaura et al., 2015;Gulla et al., 2015;Rasmansyah, 2017).

The Influence of E-Service Quality on Customer Satisfaction (H3)
The direct relationship between e-service quality and customer satisfaction was obtained is that H0 was rejected, which means H3 was accepted. It was as a result of the t-value was 12.448, which was higher than critical value (1.96); thus, it can be concluded that e-service quality influenced significantly on the customer satisfaction. The value of the coefficient parameter of e-service quality to customer satisfaction was 0.554, which means e-service quality has a positive influence on customer satisfaction, which obtained about 0.554. This finding is consistent with the results of previous studies which stated that service quality is a very important variable and its role cannot be ignored in retaining customers by increasing the level of satisfaction (Chen, & Chuang, 2008;Elder & Krishna, 2011;Sharma & Panga, 2018).

The Influence of Customer Satisfaction on Customer Loyaty (H4)
The direct relationship between customer satisfaction and customer loyalty was obtained is that H0 was rejected, which means H4 was accepted. It was as a result of the t-value was 6.097, which was higher than the critical value (1.96); thus, it can be concluded that customer satisfaction influenced significantly on customer loyalty. The value of the coefficient parameter of customer satisfaction to customer loyalty was 0.472, which means customer satisfaction has a positive influence on customer loyalty, which obtained about 0.472. The result is supported as previous studies revealed that customer satisfaction turns out to be affecting customer loyalty, that satisfaction is an important step in forming loyalty (Akbar & Parvez, 2009;Kassim & Asiah, 2010;Van-Vuuren et al., 2012;Rimawan et al., 2017).

The Influence of Price on Customer Loyalty (H5)
The direct relationship between price and customer loyalty was obtained that H0 was rejected, which means H5 was accepted. It was as a result of the t-value was 2.086, which was higher than the critical value (1.96); thus, price influenced significantly on customer loyalty. The value of the coefficient parameter of customer satisfaction to customer loyalty was 0.137, which means price has a positive influence on customer loyalty, which obtained about 0.137. These results are equivalent to previous studies conducted by Khan & Hameed (2017), that price has shown a dominant relationship between customer satisfaction and loyalty. The relationship between price and loyalty receives great considerable attention too (Ferreira & Coelho, 2015) it is clearly stated that switching behavior factors, can influence the customer to keep using a product.

The Influence of Promotion on Customer Loyalty (H6)
The direct relationship between promotion and customer loyalty was obtained is that H0 was rejected, which means H6 was accepted. It was as a result of the t-value was 2.690, which was higher than the critical value (1.96); thus, promotion influenced significantly on customer loyalty. The value of the coefficient parameter of promotion to customer loyalty was 0.178, which means promotion has a positive influence on customer loyalty that obtained about 0.178. Based on a study (Ameur et al., 2015), showed that sales promotion plays an important role in customer loyalty. Moreover, there is also a strong and positive influence of promotion on customer loyalty (Zeithaml et al., 1985;Owomoyela et al., 2013).

The Influence of E-Service Quality to Custom er Loyaty (H7)
The direct relationship between e-service quality and customer loyalty was obtained is that H0 was accepted, which means H7 was rejected. It was as a result of the t-value was 0.072, which was lower than the critical value (1.96); thus, e-service quality did not influence significantly on customer loyalty. The value of the coefficient parameter of e-service quality to customer loyalty was 0.005, which means e-service quality has a positive influence on customer lo-yalty, which obtained about 0.005. The previous studies that support this result are studies (Rasmansyah, 2017;Dewi et al., 2014) that state service quality does not have a significant influence on customer loyalty.

CONLUSION AND RECOMMENDATION
Based on the study conducted to 400 respondents, it can be concluded to answer all research questions proposed in problem identification in this study. The conclusions are, from the result of the study showed that the contribution of the price (X1), promotion (X2), and e-service quality (X3), which have a simultaneous and direct influence on customer satisfaction (Y). The minor influence caused by other factors that cannot be explained in this study. Meanwhile, from study results also showed that the contribution of the price (X1), promotion (X2), e-service quality (X3) and customer satisfaction (Y), which have a simultaneous and direct influence on customer loyalty (Z). Another factor have slight influence however cannot be explained in this study.
This study has several limitations. First, this research analyzes the effect of price, promotion, and e-service quality on customer loyalty, either directly or through customer satisfaction as a mediator in online transportation services. Perhaps, there are some other variables, which can affect customer satisfaction and customer loyalty, the researchers strongly suggest that in future studies, these variables should be explored and analyzed. Second, this study was conducted in the online transportation services industry; therefore, the results and conclusion might not be generalizable to other industries. Thus, it is strongly recommended for the next study, this topic is also conducted in other industries.
For the next researchers, it is expected to add another determinant variables, which influence on customer switching behavior of online transportation service. Besides, it is also suggested to increase the sample size of the study to make the result more accurate.
The suggestion for online transportation providers is to improve their e-service quality. As this study was found out that service quality has a high impact on customer satisfaction. But this factor does not enable yet to increase customer loyalty. Meanwhile, the pricing strategy has a high impact on loyalty.
Lastly, based on the result of the study, we can conclude that people in Bandung are very sensitive to price change. They tend to switch from one provider to other providers if the service price increases.