Customer Satisfaction Analysis of Supervisory Consultant Construction Performance at Badung Regency Public Works and Spatial Planning

Performance to maintain competitiveness in the construction sector. Based on direct observation and preliminary interviews with the construction work directors of the Badung Regency Public Works and Spatial Planning, the realization of physical progress is not in accordance with the time schedule, lack of administrative order, and lack of understanding of administrative documents are still commonly found. These problems affect the performance of supervisory consultants, which can cause a bad impression in the future. The purpose of this study is to analyze the factors that influence the performance of supervisory consultants and how the relationship between supervisory consultant performance and customer satisfaction and the performance of supervisory consultants that needs to be improved or maintained in order to increase customer satisfaction. This research was conducted by distributing questionnaires to project owners involved in project implementation at the Badung Regency Public Works and Spatial Planning. The data from the distribution of questionnaires will be analyzed using factor analysis and PLS-SEM methods. Based on the analysis results with factor analysis, the factors that affect the performance of supervisory consultants are cost, quality and time control, work inspection and correction, coordination meetings and documentation, administrative preparation, understanding contract documents, and problem-solving factors. The relationship between supervisory consultant performance and customer satisfaction results in a positive and significant relationship. Efforts to increase service user satisfaction are to improve the performance of supervisory consultants by understanding contract documents, controlling time project management, pay more attention to document completeness and service quality.


Introduction
A construction project that has good performance is caused by good supervision (Harsian and Zaitul, 2020). The role of the supervision consultant is to monitor, inspect/review, and evaluate the contractor's performance. Consultant supervision performance is declared good if the implementation of project supervision is in accordance with the request or expectations of the owner as customer. Supervisory consultant performance is used as a measurement of the level of effectiveness that links the quality of work products and consultant productivity. The services offered by supervisory consultants are not always assessed for quality from the final product, namely in the form of a supervisory report, but also from the supervision process during implementation (Abas et al., 2021). Quality consultant services must be able to provide satisfaction to customer. Satisfaction is measured based on an expression of pleasure or disappointment with what is obtained by Based on direct observation and preliminary interviews with the construction work directors of the Badung Regency Public Works and Spatial Planning, it was found that the realization of physical progress was not in accordance with the time schedule and lacked orderly administration. In 2022, construction projects in Badung Regency amounted to 36 construction projects for buildings, roads, and water/irrigation buildings with 38 work packages for supervision consultancy services, of which five projects experienced delays, so an extension of the project implementation time was required. Lack of understanding of supervisory consultants of contract documents is also still found. From the results of interviews with six supervisory consultants, there are four consultants who do not understand the scope of work and administrative provisions. These problems affect the performance of supervisory consultants, which can create an impression of poor performance in the future. Poor performance will affect the satisfaction of customer (Damayanti et al., 2022). Assessment of customer satisfaction with the performance of supervisory consultants results in factors that are prioritized to improve performance so that they can provide breakthrough improvements and quality results as expected in construction projects within government agencies, especially the Badung Regency Public Works and Spatial Planning and can still compete in the construction market.

Supervisory Consultant Performance
Supervision consultants have a vital role in assisting customer and especially in terms of supervision of construction implementation, both from the aspects of human resources, tools, materials, costs, time, quality, and HSE. Supervision aims to ensure that the process and results of the work carried out by the contractor are in accordance with the qualitative and quantitative requirements stated in the contract and provide administrative, technical documents during implementation as a means of monitoring work progress (Apriliasari and Indryani, 2010).

Customer Satisfaction
According to Kotler, quoted by Tjiptono and Diana, (2019), customer satisfaction is the level at which a person feels that the performance or outcomes have met his expectations. Quality service products play a significant role in determining customer satisfaction; as a result, the higher the level of service performance and the products offered, the higher the level of satisfaction felt by customer. This can be advantageous for service providers as well, as it is hoped that by achieving a certain level of satisfaction, there will be repeat purchases from customer who decide against switching service providers.

Method
This study uses the factor analysis method, which is used to find relationships between mutually independent variables, which are then grouped into groups so that one or more sets of variables can be formed which are fewer than the initial number of variables. In this study, the type of factor analysis used is exploratory factor analysis or Principle Component Analysis using the Statistical Package for the Social Sciences (SPSS) program tool and the PLS-SEM method, which is used to analyze how the relationship between variables.

The Proposed Hypothetical Model
Each component was selected based on the literature review. Previous research reveals that service user satisfaction is influenced by project performance (Kotler, (2017); Harsian and Zaitul, (2020)) and communication (Tai et al., (2009) ;Kusuma, (2014)). The hypothesis of this research are Supervisory consultant performance has a positive and significant effect on customer satisfaction. For the detail can be seen in the following Figure 1

Validity Test
The validity test in this study used a construct validity test with a degree of freedom (df) of 50 respondents minus 2 to 48, with the significance level (a) used being 5% (five percent). The r (correlation) table value is 0.284 (Ghozali, 2021). The r (correlation) calculated value was obtained using the IBM SPSS V.26 program. Based on the validity test of supervisory consultant performance from 26 statements, the results for two statements KIN18 and KIN25 are invalid because of the r count (0,168 and 0,280) under the r table. The invalid statement will not be included in the further analysis. For customer satisfaction, there are 16 statements declared valid so that the whole can be used for the next stage, the Reliability Test.

Factor Analysis of Supervisory Consultant Performance Variables, Supervisory Consultant Team Communication and Customer Satisfaction
In processing the factor analysis data, this research uses the IBM SPSS 26 program and the analysis steps are as follows:

Kaiser-Mayer-Olkin (KMO) Measure of Sampling Adequacy (MSA) Value and Bartlett's Test of Sphericity
The first step is to determine the KMO MSA value, where the KMO MSA value must be more than 0.5 and the Bartlett's Test of Sphericity significance value is less than 0.05 (Malhotra and Birks, 2007). Based on the analysis results, the KMO MSA value of supervisory consultant performance and customer satisfaction is 0.729 and 0.824, where the values are more than 0.5, so the instrument used has met the KMO MSA and Bartlett's Test of Sphericity significance criteria and can be said to be reliable.

Factor Extraction (Determining the Number of Factors Based on Eigenvalue) and Factor Rotation
Kaiser (1960) recommends that the factor to be used is the factor that has an eigenvalue of more than 1. The percentage of the total variation that can be described by the number of tobe-produced components is used to determine the second criterion. The Total Variance Explained value of Supervisory Consultant Performance analysis results obtained from component one to component six has an eigenvalue of more than I with a value of 9,372; 2,622; 1,866; 1,583; 1,102; and 1,088. Components one to six can explain the variance of 39.051%; 10.926%; 7.775%; 6.596%; 4.590% and 4.533% respectively so the total variance is 73.471%. The total value of Variance Explained Customer Satisfaction shows the results of component one to component four having eigenvalues 1 with a value of 7.491%; 1.477%; 1.347% and 1.119%, respectively. Components one to four can explain the variance of 46.819%; 9.228%; 8.420%; and 5.991%, respectively so that the total variance is 71.459%. When viewed, the variables correlate with each factor, but the resulting factor loading has not been able to provide the meaning as expected. Therefore, the factor cannot be interpreted clearly, so it is necessary to rotate with the varimax method.

Factor Rotation and Factor Naming
The rotation procedure seeks to produce factor loadings that are comprehensible. Compared to the component matrix, the correlation matrix known as the rotated component matrix presents a clearer and more prominent distribution of variables. The rotated factor loadings have given the expected meaning and each factor can be interpreted clearly. After the formation of factors, each of which consists of the variables studied, factor naming is carried out based on the characteristics that match its members. The results of naming the factors for each variable are:

PLS-SEM (Structural Equation Model -Partial Least Square) Analysis
The evaluation stages consist of measurement model evaluation, structural model evaluation, and mediation testing.

Loading Factor (LF)
The loading factor value is the correlation between each measurement item and the variable. This measure illustrates how well the item describes the variable measurement. According to t Hair et al., (2021) LF value ≥ 0,70 is acceptable, but according to Chin, (1998) LF value is acceptable when its ≥ 0,60. In the test, three stages of factor testing were carried out to achieve the minimum value, with the final test results as the Figure 2 and the    Figure 2 and Table 4, the results of third stage testing show that the LF value of all statements is above 0.7 so that the evaluation can continue to the next analysis.

Average Variance Extracted (AVE)
According to Hair et al., (2021) AVE value ≥ 0.50 indicates that the average variance of the measurement items contained by the variable is above 50%. The results of the internal consistency reliability value and convergent validity can be seen in Table 5  The results of the AVE value in this test can be concluded as; (1) The AVE value for the supervisory consultant performance variable is 0.594, which means that the amount of variation in the measurement items contained by the supervisory consultant performance variable is 59,4%. Because the AVE value of project performance is 0.594 > 0.50, the convergent validity evaluation is fulfilled; (2) The AVE value for the customer satisfaction variable is 0.662, which means that the amount of variation in the measurement items contained is 66.2%. The AVE value of customer satisfaction is 0.662 > 0.50, so the convergent validity evaluation is fulfilled.

Discriminant Validity
Discriminant validity test in PLS can use three methods, namely cross loading, square root of AVE and HTMT.

Cross Loading
Evaluation of discriminant validity at the indicator level is fulfilled, where each item / measurement dimension correlates more strongly / higher with the variable it measures. The

Square Root Value AVE or Fornell Lacker Criterion
The fornell lacker criterion value is acceptable if the AVE root of each variable (on the diagonal axis) is higher than correlation of that variable with other variables. The results of the Fornell Lacker Criterion can be seen in Table 6 Table 6 Based on the results of the analysis of the Fornell Lacker Criterion value, it can be concluded that the supervisory consultant performance variable has an AVE root value of 0.771, this value is higher than its correlation with customer satisfaction (0.658). Customer satisfaction has an AVE root value of 0.814, this value is higher than its correlation with supervisory consultant performance (0.658).

Heterotrait Monotrait Ratio (HTMT)
The HTMT measure introduced by Henseler et al., (2015) recommended the value of HTMT is below 0.90. Based on the analysis results, the HTMT value on the performance of supervisory consultants and customer satisfaction is 0.759, where the HTMT value shows that all variables are below 0.90, indicating that discriminant validity is met. The HTMT value is more recommended to be reported because it has a higher level of sensitivity than fornell lacker and cross loadings (Hair et al., 2019).

Reliability
Cronbach's Alpha and Composite reliability are two ways to measure reliability in PLS. Cronbach's Alpha should be better than 0.7, and composite dependability should be greater than 0.7. Based on the results of the analysis of the internal consistency reliability value, it can be concluded that the Cronbach's Alpha value for the supervisory consultant performance variable is 0.828 and the customer satisfaction variable is 0.898. The Composite reliability value for the supervisory consultant performance variable is 0.880 and the customer satisfaction is 0.922. Overall, the Cronbach's Alpha Composite Reliability value meets the minimum value requirement of 0.7. So it can be concluded that all measurement items that measure each variable are said to be consistent / reliable.

Evaluation of Reflective Measurement Models
According to Ghozali, (2021) structural model evaluation consist of R-Square, Effect Size f 2 , Q 2 preditcive relevance, and significance (two tailed).

R square
According to Hair et al., (2021)    The Q Square value of the customer satisfaction is 0.275, the value is above 0.25, indicating moderate prediction accuracy. The Q Square value of customer satisfaction has a value above 0 which indicates the model has predictive relevance.

Significance and Relevance of The Path Coeffcients (Two-Tailed)/Direct Effect Testing
To determine a hypothesis can be accepted or rejected, it can be done by observing the significance value between constructs t-statistics and p-values. If the p-value <0.05 and the tstatistic value> 1.96 , the hypothesis will be accepted (Hair et al., 2014).  Based on the results, it is known that the path coefficient value (original sample of supervisory consultant performance on customer satisfaction) is 0.658, which is positive, which means that performance has a positive effect on customer satisfaction. The results of the t-statistic value obtained are 8.465 > 1.96 and the p-value 0.000 <0.5, which means that these results are significant; it can be concluded that the performance of supervisory consultants has an effect on customer satisfaction (hypothesis accepted).

Conclusion
Based on the results of the study, the factors that affect the performance of supervisory consultants at the Badung Regency Public Works and Spatial Planning are cost, quality and time control factors, work inspection and correction factors, coordination and documentation meetings, administrative preparation factors, understanding of contract documents and problem solving factors. The relationship between supervisory consultant performance and customer satisfaction is that the performance of supervisory consultants has a positive and significant relationship to customer satisfaction, this means that the higher the performance of supervisory consultants, the higher the perceived customer satisfaction. Based on the results of the analysis that has been carried out, it can be concluded that the supervisory consultant's efforts in increasing customer satisfaction by improving performance in studying and understanding contract documents, improving time management project supervision, as well as improving administration completeness and service quality to maintain satisfaction and compete in the construction market.