SUCCESS FACTORS OF TASK MANAGEMENT APPLICATION IN THE BUSINESS SERVICE DIVISION OF PT TELKOM INDONESIA

: The success of the company's operational management is formed by the success of building and running the information system. PT Telkom Indonesia is using a task management application to manage sales activity. Success assessment of the application uses the Delone & Mclean model. This study aims to analyze factors that influence the success of implementation task management based on the perspective of the direct user called the account manager. Respondents in this study are account managers with a total sample of 150 people. The data collection technique uses a questionnaire and is analyzed using the SEM-PLS method. The results of this study indicate that system quality has a significant effect on usage and user satisfaction, information quality has no significant effect on usage and user satisfaction, and service quality has no significant effect on usage but has a significant effect on user satisfaction. Usage has a significant effect on user satisfaction. Furthermore, user satisfaction and usage have a significant effect on net benefits. Therefore the Delone and Mclean model is confirmed to be used in assessing the success of the implementation task management at the Business Service Division PT Telkom Indonesia.


INTRODUCTION
One of the successes and failures in the business is determined by the ability to properly manage the company's operations.A business's success level is assessed based on how much the company benefits from the business activities in high profits with the selected company strategies (Priatna, 2016).The company's operational management is currently utilizing technology as a support in carrying out business activities.Now, technology has become an enabler and a factor of production that is mandatory for a company or organization (Nugroho, 2021).This technology allows tasks to be completed faster, processes to be more structured, and costs to be cheaper.
Every year large amounts of investment are made by organizations or companies for developing and building information systems, which are expected to support existing business processes (Apriyanto & Putro, 2018).Based on the International Data Corporation report (2020), spending forecast on global technology investment will increase by 20.67% in 2021 and 5.60% in 2022.An information system is said to be successful if it provides benefits to its users, which are measured based on the benefits after using the system.Technology plays a vital role in increasing the efficiency and effectiveness of performance, generating strategic advantages, changing organizational structures, and providing competitive advantages (Kasemin, 2015).Related to the background of the significant investment in this information system, it is necessary to measure the level of success where the purpose of this measurement is to make repairs or increase the performance system.
Many studies have been conducted to identify the factors that influence the success of information systems, one of which was carried out by Delone and Mclean.Delone and Mclean's research (1992) offers a model for assessing the success of an information system called the D&M IS Success Model.This model provides an overview of the success of the information system based on the variables of system quality and information quality affecting use and user satisfaction.The number of uses affects user satisfaction.Use, and user satisfaction affect individuals and the organization.This model was refined by adding a service quality variable, combining individual and organizational impacts into one variable, net benefits (Delone and Mclean, 2003).
The information system an organization uses is task management.The task management system is built by software to organize and manage tasks effectively using task creation, planning, delegation, tracking, and reporting (Riss et al. 2005).Task management is a form of "process-aware information system" in this paradigm that has significantly contributed to increasing employee productivity (Riss et al. 2005).
The novelty in this study is the research object, namely the task management application implemented in Telkom Indonesia.In contrast, in previous studies, the object studied was the information system developed by each company or organization.As one of the state-owned enterprises engaged in information and communication technology, Telkom has implemented technology in all of its operational activities, both internally and in terms of customer service.The forms of the implementation technology are in the sales activity carried out by the Business Service Division (DBS) account manager using the MyTens go Beyond application, which is a digital personal assistant application in supporting business productivity, effectiveness, and transparency, which includes product catalogs, tracking orders, activity reporting, and after-sales service.
The problem that arose in implementing the MyTens Go Beyond application is the level of complaints, which is still high.Based on complaints reports and problem handling in 2022, the average monthly complaint is 57.In addition, the average time for handling problems is still long, with the resolution of disturbances which is still high, around 261 minutes.Furthermore, the order list data in the system did not match the facts, resulting in the unit being unable to control the order.Offline data adjustment activities must be carried out to match actual order data with existing orders in the MyTens go Beyond system, limited order tracking.Besides, this application has not been integrated with the Customer Relation Management (CRM) system.Based on the problems described, this research aims to analyze conditions of the factors that influence the successful implementation of the application and then to ensure how far the application has been successfully implemented by considering the variable system quality, information quality, service quality, usage, user satisfaction, and net benefits based on Delone & Mclean's information system success model.theoretical approach.These variables can be measured and have precise instruments to analyze the data using statistical procedures (Cressel, 2014).
The approach uses the Delone & Mclean success model.The hypothesis testing uses SEM-PLS.Furthermore, the results of hypothesis testing were analyzed to see the factors that influence the successful implementation of task management based on the variables of system quality, information quality, service quality, usage, user satisfaction, and net benefits.The framework of thought in this study describes how thinking about existing research problems starts from the emergence of problems, then explaining the research objectives, solving problems through analytical methods then making recommendations based on research results.Figure 1 shows the framework of this research.

METHODS
This research was conducted from October -December 2022 at the Business Service Division PT Telkom Indonesia with account managers as respondents.The data used in this research is primary data obtained from the collection technique using a questionnaire and analyzed using the Structural Equation Modeling-Partial Least Square (SEM-PLS) method involves structural models and measurement models (Sarstedt, 2019) and provides solutions for small sample sizes with many constructs (Hair et al. 2018) with 150 respondents.Other data were obtained from various literature sources such as previous research journals, company documents, and websites.The method used in this research is quantitative research that examines the relationship between variables using an objective  has a significant effect on user satisfaction, and user satisfaction has a significant effect on net benefits.Wara et al. (2021) researched the success of the examination application system (SIAP) at BPK North Sulawesi using System quality, information quality, service quality, user satisfaction, usage, and net benefits.System quality and service quality results have a positive effect on usage.Information quality has a positive effect on user satisfaction, and user satisfaction positively affects net benefits.Zibak (2021) conducted research on the success of cyber threat management platforms using variables adapted to the research object, namely content quality, system quality, service quality, trust, usage, user satisfaction, and net benefits.Quality of content and trust in the platform is among the most critical success factors and should be considered a priority for operating a successful cyber threat management platform.

Description of the Conditions of Success Factor Implementation of the Task Management Application (MyTens go Beyond)
In this study, the application implementation is described by research variables through the loading factor value and the median value.Loading factor reflects the degree of association between the observed variables and the factors or indicators.Furthermore, the median explains the level of understanding of the variables studied based on the user's point of view.These variables include system quality, information quality, service quality, usage, user satisfaction, and net benefits.The following are the results of respondents' responses to the research variables.
Based on the information system success model of Delone and Mclean (2003), the following is an initial research hypothesis to serve as a reference in measuring and analyzing the successful implementation of task management applications (Mytens go Beyond).The initial research model based on the hypothesis above can be seen in Figure 2. The hypothesis for this study is formulated as follows: H1 : System quality affects usage H2 : Information quality affects usage H3 : Service quality affects usage H4 : System quality affects user satisfaction H5 : Information quality affects user satisfaction H6 : Service quality affects user satisfaction.H7 : Usage affects user satisfaction H8 : Usage affects net benefits H9 : User satisfaction affects net benefits A popular theory of the information system success model is Delone McLean's theory.In Delone McLean's (1992) success model, the success of a system is measured using system quality, information quality, usage, user satisfaction, individual impact, and organizational impact.Delone McLean's theory was then developed with the renewal model in 2003 by adding service quality variables and changing individual variables and organizational impact into net benefits.Iivari (2005) conducted an empirical test by Delone and Mclean 1992 for mandatory or mandatory system information using the PLS-SEM.The results showed that system quality and information quality affect user satisfaction, and user satisfaction has a more decisive influence on individual impact.Wang and Liao (2008) tested the success of information systems using Delone McLean's 2003 theory to measure the success of the e-Gov application.The results obtained were that information quality and system quality significantly influence user usage and satisfaction.Use

Usage
The highest loading factor value is shown on the indicator frequency of use with a median value of 4, which is included in the good category.It shows that account managers often use the application.The reason for application usage is mandatory.At the same time, the loading factor with the lowest value is seen in the daily used time indicator with a median value of 2, which is in the unfavorable category.It shows that the daily use of the account manager is considered less.

User Satisfaction
The highest loading factor value is seen in the information satisfaction indicator with a median of 4 which is included in the good category.It shows that the output information generated by this application benefits its users.The information in question is related to the output data the account manager needs, such as the project data list, order tracking, performance, and others.At the same time, the lowest loading factor value is seen in the system quality indicator with a median value of 4, which is included in the good category.It shows that the application system can support how the account manager works.

Net Benefit
The highest loading factor value is shown on the work indicators with a median value of 4, which is included in the good category.It shows that this application makes the account manager's job easier.The activities are appropriately recorded, the number of orders can be monitored, and performance can be seen directly.Meanwhile, the lowest loading factor is seen in valuable indicators in work with a media value of 4, which is included in the good category.It shows that the application as a whole helps support activities for its users.

Validity and Reliability Testing
The first stage is to assess convergent validity.An indicator validates reflective latency if it has a loading factor value greater than 0.70 (> 0.70).While the loading factor of 0.50 to 0.60 can still be maintained for models in the development stage.If the loading factor value is less than 0.5, the construct is declared invalid and must be eliminated from the model and recalculated.Figure 3 below presents a result of calculating the loading factor after the elimination process.

System Quality
The highest loading factor value is shown on the indicator access convenience with a median value of 4, which is included in the excellent category where this indicator shows that the application is comfortable to use by its users.This convenience can be seen in that the application system can be accessed through websites and mobile apps that can be used anywhere and anytime.Meanwhile, the lowest loading factor value looks at the system integration indicator with a median value of 4, which is included in the excellent category.The MyTens go Beyond application can be integrated with other applications such as Customer Relationship Management (CRM).The lowest loading factor value is seen in the response time indicator with a median value of 4 which is included in the excellent category.It shows that the application can quickly process data input by the account manager.

Information Quality
The highest loading factor is seen on the indicator, constantly updated/up to date, with a median value of 4, which is included in the good category.It shows that the application provides up-to-date information following the conditions.It is information related to activity stages, performance, and others.At the same time, the lowest loading factor value is seen in the period indicator with a median value of 4, which is included in the good category.It shows that the indicator can provide the information the account manager desires, such as filtering the time period as needed.

Service Quality
The highest loading factor is seen on the indicator empathy with a median value of 4, which is included in the good category.It shows that application providers understand the needs of users.The function of the application is to support operational activities, especially at the pre-sales stage, by presenting this application.Meanwhile, the lowest loading factor value is seen in the application display indicator with a median value of 4, which is included in the good category.It shows that the application must have an attractive display, simple and easy to understand for the user.The indicators referred to in the diagram above are response time (X13), access convenience (X14), functionality (X15), reliability (X16), usability (X16), completeness (X21), information accuracy (X22), up to date (X23), Format Output (X24), Time Period (X25), Relevance (X25), Responsive (X31), Security assurance (X32), Application Views (X33), Empathy (X34), Service feature security (X35), Frequency of use (Y12), System satisfaction (Y21), Information satisfaction (Y22), Service satisfaction (Y23), Faster for work (Y31), Improved performance (Y32), increased productivity (Y33), increased effectiveness (Y34), ease of work (Y35 ), and useful in work (X36).Furthermore, the average variance extracted (AVE) test will be carried out to further strengthen the convergent validity results with the criterion that if the AVE value is > 0.5, then the construct used in the study is valid.

R Square analysis
The structural or inner model is constructed based on Partial Least Square (PLS).It was evaluated using R 2 with values of 0.67, 0.33 and 0.19 for each endogenous latent in the structural model, which can be interpreted as strong, moderate and weak substances (Chin, 1998).
It is known that the R-Square for the use variable (Y1) of 0.051 is included in the weak category.R-Square for user satisfaction (Y2) of 0.576 and net benefits (Y3) of 0.511 is included in the moderate category.Table 3 is the R 2 result for the variable usage, user satisfaction and net benefits.

Hypothesis Test Results
Hypothesis testing in this study used the path coefficient and p-value.A positive value coefficient indicates the influence of latent variable constructs.Furthermore, There has no influence indicated by a negative value coefficient.Assessment of this hypothesis uses a significance level of 95% or a standard error of 0.05.
Based on the value of the p-value of a hypothesis has a significant effect if the p-value <0.05 and vice versa.A hypothesis has no significant effect if the p-value > 0.05. Figure 4 presents the results of the path coefficient.
The path coefficient describes the magnitude influence of the construct on the latent variable.Meanwhile, the influence between latent variables is described by the p-value.The significance value used is 95% or a standard error of 0.05.Based on the result testing hypothesis can be concluded as significant or not significant.The final hypothesis testing in this study is related to variable system quality, information quality, service quality, usage, user satisfaction and net benefit.The path coefficient value is negative -0.180, which shows that the direction of the relationship between Information Quality (X 2 ) with Usage (Y 1 ) is negative or reversed.If Information Quality (X 2 ) increases, then Usage (Y 1 ) will decrease, and vice versa.The p-value results show 0.097 (>= α), which shows that Information quality (X 2 ) has no significant effect on Usage (Y 1 ).It shows the quality of the information presented by the MyTens go Beyond application.In this case, the application's output does not affect usage because the application is mandatory, so the information produced by the application is not helpful or still used by the account manager.These results follow the results of research from Afnan (2018), Dewantoro (2019), andWara et al. (2021).
Hypothesis 3: Service Quality (X 3 ) affects Usage (Y 1 ) The path coefficient value negative -0.038 shows that the direction of the relationship between Quality of Service (X 3 ) with Usage (Y 1 ) is negative or reversed.
the information generated from the application does not add to the user's satisfaction with the application.The account manager focuses on the level of usability of the application as a whole system so that the information features provided in the application, which provide an accurate picture of the activities carried out by the account manager, do not directly affect user satisfaction.These results follow Dewantoro (2019) and Akrong (2022) research.
Hypothesis 6: Quality of Service (X 3 ) affects User Satisfaction (Y 2 ) The path coefficient value positive 0.431 shows that the direction of the relationship between Quality of Service (X 3 ) with User Satisfaction (Y 2 ) is positive or unidirectional.It means if the Quality of Service (X 3 ) increases, then User Satisfaction (Y 2 ) will increase too, and vice versa.The p-value results show 0.000 (<= α), which shows that the quality of service (X 3 ) has a significant effect on user satisfaction (Y 2 ).It shows that the quality of service from the comprehensive support of application providers influences user satisfaction.It is in line with the application provider's commitment that the development of the application meets the expectations of the company and the users, both in terms of service features and the responsiveness of the application provider.These results follow research from Afnan (2018), Wang andLiao (2008), andDewantoro (2019).

Hypothesis 7: Usage (Y 1 ) affects User Satisfaction (Y 2 )
The path coefficient value positive 0.172 shows that the direction of the relationship between Usage (Y 1 ) with User Satisfaction (Y 2 ) is positive or unidirectional.It means if Usage (Y 1 ) increases, then User Satisfaction (Y 2 ) will increase too, and vice versa.The p-value results show 0.004 (<= α), which means that usage (Y 1 ) has a significant effect on user satisfaction (Y 2 ).It shows that user activity reflected in the application usage level shows that user satisfaction is directly proportional.The indicator reflected in the loading factor is shown by the Frequency of use of the MyTens go Beyond application by a mandatory account manager.The results of this study follow research from Costa (2020), Wara et al. (2021), Dharmawan (2022), andAkrong (2022).
If the Quality of Service (X 3 ) increases, then Usage (Y 1 ) will decrease, and vice versa.The p-value results show 0.768 (>= α), which shows that the quality of service (X 3 ) has no significant effect on Usage (Y 1 ).It shows that the MyTens go Beyond application developer's comprehensive support does not affect the application's use.The highest loading factor value is found in the empathy indicator with a score of 0.817, meaning that application developers understand what users need regarding service improvement, adding features, and other compliant services.This result is directly proportional to the research results by Wang and Liao (2008) and Afnan (2018).
Hypothesis 4: System Quality (X 1 ) affects User Satisfaction (Y 2 ) The path coefficient value positive 0.240 shows that the direction of the relationship between System Quality (X 1 ) with User Satisfaction (Y 2 ) is positive or unidirectional.It means if System Quality (X 1 ) increases, then User Satisfaction (Y 2 ) will increase too, and vice versa.The p-value results show 0.007 (<= α), which indicates that the quality of the system (X 1 ) has a significant effect on user satisfaction (Y 2 ).It shows that the quality of a good application will increase the application users' satisfaction.The application is developed with relevant features that can accommodate the needs of the user so that the application can make it easier for users to carry out their activities, where the purpose of developing the application is for better account manager governance and record keeping.This result is directly proportional to the results of research from Iivary (2005), Wahyuni (2011), Afnan (2018), Dewantoro (2019), Costa (2020), Zibak (2021), and Dharmawan (2022).

Hypothesis 5: Information Quality (X 2 ) affects User Satisfaction (Y 2 )
The path coefficient value positive 0.141 shows that the direction of the relationship between Information Quality (X 2 ) with User Satisfaction (Y 2 ) is positive or unidirectional.It means if Information Quality (X 2 ) increases, then User Satisfaction (Y2) will increase too, and vice versa.The p-value results show 0.147 (>=α), which shows that the quality of information (X 2 ) has no significant effect on user satisfaction (Y 2 ).It shows that and increase the net benefits through improving the quality of application information by providing output data that is faster (up to date).In addition, performance information must provide an accurate picture because the information in this application is used to carry out performance appraisals that impact the amount of remuneration the account manager receives.In addition, the improvement of service quality must also be carried out by application developers.An indicator that needs to be improved is empathy.Application providers must be able to understand application users' needs, such as solving problems quickly and making continuous application improvements.

Conclusions
The factors that influence the application's success are very well understood by the respondents, which are reflected in the variables of system quality, information quality, service quality, usage, user satisfaction, and net benefits.The presence of this application provides net benefits to users and organizations through the indicators it builds.System quality has a significant effect on usage and user satisfaction.Information quality has no significant effect on usage and user satisfaction.Service quality has a significant effect on user satisfaction but has no significant effect on usage.Usage has a significant effect on user satisfaction, and also usage and user satisfaction have a significant effect on net benefits.Application users can feel net benefits, namely account managers, especially indicators of ease of work where the application benefits account managers in carrying out their activities.The application's better system quality, information quality, and service quality will impact the net benefits the account manager can feel.Recommendations for improving application performance are carried out by improving system quality, information, and service quality.Indicators that need to be improved are reliability, where the application must be able to answer the user requirement, the indicator is constantly updated by providing output data that is faster (up to date), and empathy, where Application providers must be able to understand what the needs of application users, such as solving problems quickly and making continuous application improvements.The path coefficient value negative -0.153 shows that the direction of the relationship between Usage (Y 1 ) with Net Benefit (Y 3 ) is negative or reversed.It means if Usage (Y 1 ) increases, Net Benefit (Y 3 ) will decrease, and vice versa.The p-value results show 0.016 (<= α), which means that usage (Y 1 ) has a significant effect on net benefits (Y 3 ).The only indicator with this latent variable is the Frequency of use.It shows that the level of use of this application is directly proportional to the net benefits received.Customers use the application on a mandatory basis, whereas a whole, the customer is assisted by the MyTens go Beyond application.This application can make the account manager's work easier to be more structured and recorded in one application that can be accessed anywhere, either by mobile apps or websites.The results of this study follow the results of previous studies by Wang and Liao (2008), Afnan (2018), Zibak (2021), andDharmawan (2022).The path coefficient value positive 0.733 shows that the direction of the relationship between User Satisfaction (Y 2 ) and Net Benefit (Y 3 ) is positive or unidirectional.It means if User Satisfaction (Y 2 ) increases, Net Benefit (Y 3 ) will increase too, and vice versa.The p-value results show 0.000 (<= α), which means user satisfaction (Y 2 ) has a significant effect on net benefits (Y 3 ).The level of satisfaction with the information on the data presented in the application, both in terms of content and features, increases the overall benefits, where the most visible is that it makes the work of the account manager easier.It follows the results of research from Dewantoro (2019), Wara et al. (2021), Zibak (2021), andDharmawan (2022).

Managerial Implications
Based on the results of this study, efforts to increase net benefits must be carried out by focusing on improving system quality, information quality, and service quality.In the system quality section, the developer must pay attention to one of the indicators that must be improved, namely reliability, where the application must be able to answer the user requirement.The application must be able to answer the user's needs, simplify the way of work and provide overall benefits when using the application

Figure
Figure 1.Research framework

Figure 3 .
Figure 3. Diagram of loading factor value outer model evaluation after elimination

Figure
Figure 4. Path coefficient value

Table 1 .
Table 1 presents the results of the average variance extracted test.The next step is to assess Cronbach's Alpha and Composite Reliability criteria.Each construct is reliable if it has Cronbach's Alpha and Composite Reliability more significant than 0.7.Table 2 below presents the reliability test results using the Smart PLS 3.0 program.It is known that all latent constructs have Cronbach's alpha and composite reliability values more significant than the critical value.The condition indicates that all latent constructs have good reliability.AVE result

Table 2 .
Cronbach's alpha and composite reliability values Table4results from the hypothesis test and path coefficient value.It shows that a good system in the MyTens go Beyond application will make it easier for the account manager to work in sales activities.The indicator that has the highest loading factor is reliability.It shows that this application is reliable because it is the only presales tool used by account managers, so the application made by the developer has been considered ideally so that account managers can use it.At the same time, the indicator value with the lowest loading factor is response time, where application response in terms of data processing is still considered low, where fast things regarding data that the account manager has input are the primary goal.Therefore the better the system quality of the MyTens go Beyond application, the better the use of the application.