THE IMPACT OF CUSTOMER INTERACTIONS IN SOCIAL
get additional revenue. Composition of smartphone users at TELKOMSEL mostly comes from smartphones with Android operating systems, followed by iOs, and so on. This information is in line with the information conveyed in the GFK data that the highest sales in Indonesia come from Android smartphones (GFK, 2017).
TELKOMSEL recorded 35% from 139 million subscribers are using smartphone. 45% of them are using android smartphone. These 35% smartphone users are potential to be leveraged using a new device plan (Rachman, 2014). Based on this condition, TELKOMSEL launched an innovation in the form of an-based internet plan IMEI called the TAU Plan.
The TAU plan is a program brand in the form of internet plans offered to smartphone users in the TELKOMSEL with IMEI records of the smartphone has been registered in the database. The TAU plan has several variants, namely:
1. Regular TAU: The plan is offered for smartphone customers with pricing based on the zone scheme
2. TAU Dynamic: This plan is designed for the Business to Business concept, which is offered to partners offered device who will later purchase this plan which bundled with starterpack & preloaded bonus for the customers.
3. TAU Lite 6 & 7: This plan aims to make these 2G & 3G customers migrate to smartphones 4G.
Since then TAU plan is always associated with every device related initiative from TELKOMSEL.
Based on the projected result 2018/19, the number of transactions and the average monthly subscriptions have not yet showed a positive trend indicated by the average number of transactions per user which has shown consistent decrease. Estimated total TAU plan subscribers only around 0.5 - 1% from total smartphone user population in Telkomsel.
Recently Telkom Indonesia also announced that Telkomsel’s total data subscriber decreased 17.4% (111.2 Million subscribers) in Semester I-2019. (Sidik, 2019). Although It’s also reported that smartphone users increased in the same period.
Based on above conditions, there is a potential of more than 100 million (excluding smartphone users who have subscribed to the TAU plan) who may not yet have known or understood the TAU plan. Yet if looking at the potential that exists in TELKOMSEL, then this opportunity is big enough to be maximized.
In addition, it is also known that digital marketing communications through social media for TAU plans are still not widely used by TELKOMSEL. So far, the marketing of TAU plan has always been attached with direct selling advertisements carried out by device partners and is limited to offline (retail) marketing strategies.
The marketing activitiy of TAU plan has never been independent, it always depends on tactical programs initiated by device partners. So that the brand of the TAU plan is not the main message in the promos being run (more as a complementary).
If the current marketing activities are not improved, Telkomsel could lose potential new smartphone users &
additional revenues from TAU plan.
To overcome the lack of exploration on social media platforms to achieve targets in marketing programs, it can be through Social media Marketing Activity (SMMA). With SMMA, we can trace the effect on Purchase Intention on the TAU plan.
Based on the potential of SMMA, it is also the reason for researchers to uncover patterns of social media marketing (Kumar, Bezawada, Rishika, Janakiraman, & Kannan, 2016). Based on the research SMMA has been done known that many different definitions of social media activity that have an influence on brand awareness and Purchase Intention.
There are several customer interactions on perceived SMMA that affect Brand Awareness and WOM which have an impact on increasing Purchase Intention ((Hutter, Hautz, Dennhardt, & Fuller, 2013); (Jamali & Khan, 2018)) namely VIral Advertisement, Opinion Seeking, & Brand Page Commitment.
So that research is needed on customer interactions at Perceived SMMA to Brand Awareness, eWOM and Purchase Intention so that telecommunications companies such as TELKOMSEL can make social media marketing plans more effective and efficient.
This research hypotesess is expected to answer:
1. The influence of Brand Page Commitment on Brand Awareness in the TAU plan 2. The influence of Brand Page Commitment on electronic Word of Mouth in the TAU plan 3. The influence of Opinion Seeking on Purchase Intention in the TAU plan
4. The influence of Viral Advertisment on Purchase Intention in the TAU plan 5. The influence of Brand Page Commitment on Purchase Intention in the TAU plan 6. The influence of Brand Awareness on Purchase Intention in the TAU plan 7. The influence of eWOM on Purchase Intention in the TAU plan
This study uses a quantitative research method with a causative relationship approach. Quantitative research illustrates how the research objectives are achieved through empirical data in the form of measuring numbers.
The correlational approach is used following the objective of this study, namely, to find out the relationship and the level of relationship between the variables studied. Because with knowing the level of relationship, it will help in many decision-making afterward. While according to the time of the study, this study was a cross-sectional study. Cross-sectional is data collection carried out in a coherent period, that is, the data is processed, analyzed, and then drawn conclusions (Indrawati, 2015).
The population in this study are Indonesian people / subscribers of TELKOMSEL using smartphone in Telkomsel network and actively use social media.
This study uses SEM as a tool to analyze. The most popular estimation method used in SEM research is the Maximum Likelihood. The recommended amount so that the results can be trusted when the variables used is four variables at 100-200 (Ghozali, 2014). Because this study uses SEM, the number of samples that researchers use is a minimum of 200 data. While the sampling technique used in this study is Purposive sampling, which is a technique of determining samples with certain considerations. The reason for purposive sampling is because of the research target smartphone users who use social media.
This study received 230 samples from Indonesia.
The data collection technique used in this study is to use primary and secondary data, with the following explanation:
a. Primary data
Primary data is data can be obtained from respondents through multiple tools such as questionnaires, panels, and focus groups, and data from interviews with resource persons, Sujarweni (2015: 89). The primary data in this study is gathering the opinion of respondents through a questionnaire on google form.
b. Secondary Data
According to Sujarweni (2015: 89), secondary data is data obtained from records, books, articles or magazines in the form of financial reports published by a company, government reports, theory books and so on.
The data collected will be tested on its validity & reability with the following explanation:
a. Validity
Validity is the degree of the legality of an indicator in measuring the concept to be measured. In research using the SEM method, two stages must be passed to assess validity and reliability (Hair et al, 2010):
i. With the T-test
The indicator is said to be valid if the -T value is from λ ≥1.96 ii. With average Variant Ext (AVE)
The indicator is said to be valid if AVE ≥ 0.6
AVE=
b. Reliability
The purpose of a reliability test is to ensure consistency of indicators on each measurement. Reality is done by measuring Construct Reliability. The indicator is said to be consistent with the value of Construct Reliability (CR)
≥ 0.7
CR = ; CR > 0.7
CR : Construct Reliability
: Loading Factor setiap indikator : Error pengukuran pada indikator
The researcher used the LISREL 8.80 application to measure the validity and reliability of this study and will be carried out on 230 samples.
The technical data analysis that the researcher uses is SEM (Structural Equation Modeling) using the LISREL 8.80 application. This is what researchers do because the research framework consists of a lot of funding and relationships between complex constructs. Also, with SEM it is possible to analyze a series of relationships simultaneously to provide statistical efficiency.
The analysis technique using the SEM model starts from determining what variables will be used and the indicators used to measure each variable in this study. The variables and indicators taken for research can see research models from previous studies that can be changed by researchers. After determining variables and indicators, relationships between variables can be described as a research model or framework.
The next step is this research will perform Validity & realibility testing to explain the relationship between these variables. Validity test can be done by looking at the results of the loading factor on each indicator and the results of the AVE value for each variable. While for reliability testing can be done by looking at the results of the value of construct reliability on each indicator. The research model used must be tested also for compatibility to be able to show the covariance matrix on each indicator used. The compatibility of the research model can be seen from the GOF assessment which has several criteria.
According to Hair et al (2010: 672), χ2, DF, CFI, and RMSEA have enough information to be able to evaluate the research model. The next step is to describe the research model and its indicators after the indicators have passed the validity, reliability, and GOF tests. Testing variables at this stage is also to see how the influence between variables in the research model is following the hypothesis that was made before. The last step is if the overall research model is good, the SEM method can be ended by drawing a conclusion and giving recommendations.
OS1 OS2 OS3 VA1 VA2 VA3 BPC1 BPC2 BPC3
OS
VA
BPC
eWOM
BA
PI eWOM1 eWOM
2 eWOM
3
BA1 BA2 BA3
PI1 PI2 PI3 λ4
λ7
λ10
λ16 λ13
λ1
δ2 δ3 δ4 δ1
δ6 δ7 δ8 δ5
δ9
ε1 ε2 ε3
ε4 ε5 ε6
ε7
ε8
ε9
Figure 1. Structural Drawing & Equation
The suitability of the model or Goodness of Fit can be used to assess a research model whether the model falls into the category of fit or does not see the indicators in the LISREL 8.80 software. The model that is stated to be good on one of the Goodness of Fit criteria will not necessarily be stated as well on other Goodness of Fit criteria.
This study observes the loading value and significant parameters as a way of testing hypotheses. In LISREL, there is no significant value that can directly indicate the relationship between a variable and other variable is significant.
However, in each LISREL estimate, there are three useful information to measure the accuracy of each parameter estimate. Under the standard error is the T value obtained from the comparison between the estimated value and the standard error:
T table with a confidence interval of 5% is .
Validity tests in this study will use the second order confirmatory analysis test technique (Second order CFA) (Ghozali & Fuad, 2014). The first level of analysis is carried out from the dimensions to the indicators. The data from the processing of the author shows that:
• All indicators have a loading factor value (> 0.7 and AVE> 0.5, therefore it can be concluded that each indicator on each of these latent variables has been proven valid for use as a measuring instrument.
• All indicators have a T-Value> 1.96, therefore it can be concluded that each of these indicators is a significant indicator related to being used as an indicator
• All dimensions have a value of CR> 0.7, therefore, it can be concluded that all indicators can be considered reliable.
In addition to validity and reliability tests (VR Test), it is also necessary to ensure the model used is a model that is fit for the available data or called Goodness of fit (GoF). According to Hair et al. (2010), enough indicators to be reported are the chi-square value (𝜒 2), degrees of freedom (DF), comparative fit index (CFI), and root mean square error of approximation (RMSEA), but researchers will add other indicators as a companion and strengthen results. It can be concluded that the results of data processing above that of the 14 indicators, 10 indicators are stated Goodfit from a minimum of five indicators that are required to be fit, therefore the model used can be declared as a FIT model and can be used for research.
The hypothesis test in this study is determined by the direction of the relationship and the significance of the relationship. As for the assessment of significance, this study uses T-Value. This study sets a trusted standard of 95% or a hypothesis about a significant relationship will be accepted if the value of T> 1.96 or smaller than -1.96.
The results of processing by the LISREL program can be seen in the picture as follows:
eWOM
BA
PI OS
VA
BPC
0.53 0.68 0.18
0.18 0.12
0.15 0.29
0.51
0.72 0.53
1.00
1.00 1.00
Figure 3. Structural Regression Coefficient Model
eWOM
BA
PI OS
VA
BPC
8.58 11.65 2.64
1.78 1.37
2.10 3.55
8.69
11.11 7.45
13.11
11.04 11.86
Figure 4. Value of T Structural Model
Opinion Seeking
Viral Advertisement
Brand Page Commitment
eWOM
Purchase Intention
Brand Awareness
Social Media Marketing Activity
Figure 5. Hypothesis Development by Researcher (2019) To conclude the hypothesis that is built, it will be indicated by the table as follows:
Table 1. Hypothesis Testing Hypothesis Regression
Coefficient T-Value Validation Result
H1 : BPC+ BA 0.53 8.58 H1 Accepted
H2 : BPC+ eWOM 0.68 1.65 H2 Accepted
H3 : OS+ PI 0.18 2.64 H3 Accepted
H4 : VA+ PI 0.12 1.37 H4 Rejected
H5 : BPC+ PI 0.18 1.78 H5 Rejected
H6 : BA+ PI 0.15 2.10 H6 Accepted
H7 : eWOM+ PI 0.29 3.55 H7 Accepted From table 1 it can be concluded that:
1. H1: BPC has a positive and significant effect on the development of Brand Awareness 2. H2: BPC has a positive and significant effect on eWOM
3. H3: OS has a positive and significant effect on Purchase Intention 4. H4: VA has a positive but insignificant effect on Purchase Intention 5. H5: BPC has a positive but not significant effect on Purchase Intention 6. H6: BA has a positive and significant effect on Purchase Intention directly 7. H7: eWOM has a positive and significant effect on Purchase Intention
The relationship between variables contained in a research model will form a structural equation. This research produces structural equations:
After this research was conducted, it was found that there were in accordance with previous studies. The following are the results of analysis by researchers:
1. H1: The result is in accordance with previous research in the smartphone industry (Jamali & Khan, 2018) and the car industry (Hutter, Hautz, Dennhardt, & Fuller, 2013). Brand Page Commitment on Telkomsel social media such as Facebook fan pages, official Instagram and Twitter can increase Brand Awareness for the TAU plan.
2. H2: The result is in accordance with previous research on the Samsung telecommunications industry (Jamali
& Khan, 2018) and on research conducted (Hutter, Hautz, Dennhardt, & Fuller, 2013) and (Bamini KPD, Dahnil, &
Yi, 2014). The more interaction between Telkomsel and customers on Telkomsel’s social media, the more eWOM that customers will do.
3. H3: The result is in accordance with previous research, namely OS plays an important role in influencing customer purchase intentions in research (Farki, 2016), (Ichsan, 2018), (Elwalda & Lu, 2014), (Wijiseno & Ariyanti, 2015), (Bamini KPD, Dahnil , & Yi, 2014) & (Mulyati & Gesitera, 2020)
4. H4: The result is in accordance with previous research on the Samsung telecommunications industry (Jamali &
Khan, 2018) where VA does not affect the intention of potential customers to buy certain products. This is also supported by a statement from Abigail Narulita Rahajeng (PressPlay Advertising Agency) that advertisements that are viral at least make people familiar with the product. an increase in sales can sometimes be at the start of a viral increase. However, over time, in reality it could only be temporary. Depending on the product, it is of good quality or not, if not, it will decline (Khoiri, Agniya; CNN Indonesia, 2018). So that the Viral Advertisement on the TAU plan / product does not necessarily increase purchase intention, another thing that needs to be considered in this case could be quality (plan price and quota of the plans offered).
5. H5: The result is in accordance with previous research on the Samsung telecommunications industry (Jamali
& Khan, 2018) and in the e-commerce industry (Ramadhita, 2019) where activities at BPC such as Facebook fanpages, official instagram and others do not directly affect potential customers’ intentions to buy products.
certain.
6. H6: The result is in accordance with previous research on different industries (Bamini KPD, Dahnil, & Yi, 2014), (Hutter, Hautz, Dennhardt, & Fuller, 2013) & (Jamali & Khan, 2018), where activities in BA affect the intention prospective customers to buy a certain product. Strengthening brand awareness on TAU products / plans will help generate purchase intention.
7. H7: The result is in accordance with previous research in different industries (Bamini KPD, Dahnil, & Yi, 2014), (Hutter, Hautz, Dennhardt, & Fuller, 2013), (Jalilvand & Samiei, 2012) & (Jamali & Khan, 2018), where activity at eWOM affects potential customers’ intentions to buy certain products. So that Telkomsel needs to consider increasing eWOM through interaction with customers on Telkomsel social media, using celebrities or influencers who have many subscribers / followers and optimizing promoters to increase discussion on personal social media and Telkomsel.
The conclusions that can be drawn from this study are:
1. Brand Page Commitment have been able to expand and increase Brand Awareness to customers / users of Telkomsel smartphones.
2. Brand Page Commitment have been able to trigger eWOM or discussions / discussions & reviews among Telkomsel customers / users regarding the TAU plan.
BA = 0.53*BPC, Errorvar.= 0.72 , R² = 0.28 EWOM = 0.68*BPC, Errorvar.= 0.53 , R² = 0.47 PI = 0.29*EWOM + 0.15*BA + 0.18*OS + 0.12*VA + 0.18*BPC, Errorvar.= 0.51 , R² = 0.49
3. Opinion Seeking’s such as the results of customer reviews or feedback are very influential in generating purchase intentions among Telkomsel customers when discussing / discussing the TAU plan.
4. Viral Adverstiment is not a major consideration / influence of Telkomsel customers in generatingPurchase Intentionfor TAU plans.
5. The role of BPC such as the Facebook fanpage, Instagram official page, and others does not have a direct influence in generating purchase intention (purchase intention) of Telkomsel customers for the TAU plan.
6. Brand Awareness has an effect on creating purchase intentions for Telkomsel customers. the wider the customer’s brand awareness of the TAU plan, the easier it is for customers to recall the plan when the customer has an interest / intention to buy it.
7. eWOM in the form of discussion or discussion has a direct effect on Telkomsel’s customers in generating Purchase Intention for TAU plans.
Theoritical Aspect Recommendation:
1. For further research, you can examine plans or other products owned by Telkomsel.
2. Subsequent research can also examine customer behavior / responses to other operators in order to draw more accurate conclusions about the characteristics of customer service providers in response to digital marketing activities.
Recommendation for Practitioner Aspects:
1. To increase purchase intention for the TAU plan, Telkomsel can do the following:
a. Through strengthening Brand Awareness and increasing eWOM on the TAU plan. This can be achieved through the following:
i. Brand Awareness: Increasing activities on the Brand Page Commitment such as increasing Telkomsel’s interactions (represented by social media admins) with customers on Facebook fan pages, Instagram official pages and others. Interactions can be in the form of posting the latest advertising materials, holding plan promos that only apply to members / followers of the social media account. This will strengthen Brand Awareness in TAU’s product / plan. Positive Brand Awareness will make it easier for customers to remember the TAU plan and make decisions when buying the plan.
b. eWOM: The more interactions on the Brand Page Commitment will automatically lead to eWOM among followers on Telkomsel’s social media. Telkomsel can also periodically use celebrities / influencers to invite their followers to do special promos for followers who review, repost, retweet, regram promo material on the Telkomsel fanpage.
c. Take advantage of the influence of Opinion Seeking:
i. Telkomsel can use competent influencers / celebrities to review TAU plans. This will make it easier for customers to find out, opinions, recommendations and references related to TAU plans, so that customers will be more easily influenced when they want to buy TAU plans.
ii. Another thing that Telkomsel can do is use partner promoters, Telkomsel employees and their families to offer and provide reviews regarding the TAU plan through their respective social media.
2. Through this research it is also known that Indonesian people, especially smartphone users, are easily exposed to information obtained through digital marketing activities, especially through social media. This is in line with Telkomsel’s digital marketing strategy which focuses more on digital channel campaigns such as advertising on social media and websites, no longer relies on conventional promotional media such as advertisements in newspapers, billboards, flyers, balighos or promos via SMS.
There are several limitations in this research. First, it focuses only on TELKOMSEL prepaid smartphone users, especially those with social media. Second, the research didn’t compare to other similar data plan from TELKOMSEL or other operators. The next phase of the research can explore more on comparison of similar data plan from Telkomsel & other operators.