Analysis of Postal Agent Performance in Sidoarjo Seen from External and Internal Factors

: The momentum of COVID-19 should be an opportunity for courier business owners because many people focus on their activities at home and rely on courier services to shop. Two big players such as JNE and JnT have shown an increase in transaction volume in 2020 and 2021. On the other hand, Pos Indonesia has experienced a 15% decrease in profit from usual. This phenomenon is also reflected in the scale of Postal Agents, one of the official partners of Pos Indonesia. Referring to data from the Central Post Office in Sidoarjo , the income of all Postal Agents in Sidoarjo decreased by around 15% during the pandemic, starting in 2020 Rp. 11.3 billion to Rp. 9.6 billion in 2021. This research was conducted to investigate the influence of aspects of external and internal factors on the performance of Postal Agents. All research data comes from questionnaires filled out by 30 respondents who are managers of Postal Agents in Sidoarjo. Then the data was analyzed using the SEM-PLS (Structural Equation Modeling Partial Least Square) method. The results showed that internal factors had a positive influence on the performance of Postal Agents with a T-statistic of 3.115 and a P-value of 0.002 (<0.05). From these internal factors, it is known that finance and service products are the most influential aspects. Meanwhile, external factors did not have a positive influence on the performance of Postal Agents with a T-statistic of 0.123 and a P-value of 0.902 (> 0.05).


Introduction
The presence of digital technology today is able to make human activities much more efficient, and innovation-friendly, and create more diverse business ideas.For example, when people want to shop, who usually have to come directly to the store, now simply from home with a handful of cellular phones and internet networks, the desired item can be directly sent to the buyer.This series of new shopping activities is one form of transformation and utilization of the digital economy.Based on BPS data, although the national economy dimmed in 2020, the information and communication sector experienced the highest growth in the second quarter of 2020 at 10.88% (Rizkinaswara, 2020).This is allegedly because, during the COVID-19 pandemic, people are increasingly dependent on the digital economy.Therefore, in the second quarter of 2021, Indonesia's economic growth, which previously experienced a decline, actually improved, with an increase of up to 7.07% (Ministry of Investment/BKPM, 2021).This result is in line with the government's expectations because it has sought the use of technology to maximize existing economic potential, especially during the pandemic.
Shopping through digital platforms is certainly very closely related to online stores in marketplaces and social media, as well as delivery courier services.The role of the courier in this series of activities is the party most responsible for ensuring the goods arrive at the destination address.No wonder the sellers are very careful in choosing courier services so that the success of the transaction between the seller and the buyer is 100% successful.Based on Top Brand Index data (2021), there are five courier services that are quite famous in the community, namely J&T, JNE, Tiki, Pos Indonesia, and DHL.
Along with the increase in micro, small, and medium enterprises (MSMEs) in the fields of food, beverages, garments, accessories, and others, the demand for delivery services is increasing.Many new players are starting to enter the logistics courier industry to scramble for the market.Even when the pandemic hit various industrial sectors, this logistics courier business grew by 40% (Fitra, 2021).One example is the J&T courier service which occupies the top position of the Top Brand Award, in 2020 it can be seen in Figure 1, this company is able to deliver an average of 2 million packages per day.Then in 2021, through J&T's official news channel, it was officially stated that the volume of shipments increased to 2.5 million packages per day.PT.Pos Indonesia (Persero) is one of the longest courier service companies in Indonesia because of its age.In the past, this company only accepted mail delivery, but Original Article now it has transformed to accept package delivery as well.It is unfortunate if you look at the Top Brand Index data in the Courier Service category in Figure 2, "Pos Indonesia" is in fourth position in the community, below other younger logistics courier companies such as J&T, JNE, and Tiki.Therefore, to increase public interest in the use of PT.Pos Indonesia, in addition to sending letters and packages, PT.Pos Indonesia also provides payment services for PDAM bills, electricity, monthly telephone, insurance, and credit.In addition, PT.Pos Indonesia also expands its business area by embracing the public to be part of this company by becoming a partner called Portal Agent (Agen Pos).Of course, this is a new source of income for PT. Pos Indonesia and open opportunities for companies to win the competition.During the pandemic in 2020-2021, when people's mobility was limited by the government's policy to work from home (WFH), it should be a goldfield for courier/expedition service businesspeople.However, based on research conducted by (Panjaitan &;Siregar, 2021), the impact of the pandemic, which resulted in many companies experiencing crises, was also felt by PT.Pos Indonesia.A study conducted by (Sitanggang et al., 2022) stated that during the pandemic, Pos Indonesia experienced a decrease in overall profit of up to 15% from usual.Written in the study, one of the causes of the decline is the existence of Large-Scale Social Restrictions (PSBB) activities so that air cargo activities are hampered and there needs to be adjustments to land and sea routes.In addition, the existence of sharp competition with many competitors and an uncertain business climate are challenges that must be conquered so that people remain loyal to using Pos Indonesia delivery services.
Marujuk in the 2020 and 2021 financial statements, the revenue of courier  Being able to maintain turnover at that number is certainly due to several influencing aspects.There are aspects that come from internal factors such as human resources (HR), finance, service products, marketing, physical counters, cooperation, technology used, and other investments.There are also aspects derived from external factors such as regulations for the implementation of postal activities, competitors, socio-economic conditions of the community, and the role of PT.Pos Indonesia, and customers.Therefore, the author is interested in conducting this research to find out what aspects are found in external and internal factors that most affect the performance of Postal Agents in Sidoarjo.

Methods
This study was conducted to determine what aspects of external and internal factors most affect the performance of Postal Agents in Sidoarjo.The research time will begin on the third Sunday of January 2023 in Sidoarjo.The sampling technique used is a purposive sampling method (intentionally) with the consideration that respondents understand the operational details of a Postal Agent.The types of data used in this study are secondary data and primary data.Secondary data includes data on the number of Postal Agents in Sidoarjo obtained from the Sidoarjo Central Post Office, while primary data includes questionnaire data that has been filled out by respondents.The questionnaire-filling process will be accompanied by direct researchers both offline and online to avoid biased answers.Then to get valid preference answers from respondents, a questionnaire with a Likert scale of 1 -7 will be used.A scale with 7 points according to research (Pearce, n.d.) proved to be more accurate, easier to use, and able to better reflect respondents' evaluations than 5 points.Also stated in another study (Azzara, 2010), a scale of 7 was used in the study because first, it gave respondents more choices in answering questionnaire items and increased point differentiation.
Data processing in this study uses smartPLS SEM (Partial Least Square -Structural Equation Modeling) software.The purpose of using PLS is to assist researchers in confirming theories and explaining relationships between latent variables.

Outer Model Test Results
To find out in detail the aspects of internal and external factors that affect the performance of Postal Agents, researchers have processed other indicators using the SEM PLS method so as to get two results, namely measurement model analysis (outer model) to evaluate the relationship between latent variables and indicators and structural analysis (inner model) to evaluate the strength of the research model as well as the level of significance.
In outer model testing, there are two things that need to be evaluated.The first is to evaluate the validity of the data by looking at the suitability of the standards of the loading factor, AVE, and discriminant validity values.Then the second thing that needs to be evaluated is the reliability of the measuring instrument/questionnaire by looking at the standard suitability of Croncbach's alpha and composite reliability values.
1. Evaluation of Data Validity.
In evaluating the validity of research data, the first value seen is the loading factor value of each indicator.The standard loading factor value for data to be valid is  0.7.The higher loading factor indicates that the indicators in a variable have a strong relationship so that they can be absorbed by the variable.
Based on the initial model of the study, there are 13 aspects (dimensions) that are estimated to affect the performance of Postal Agents in Sidoarjo.Five dimensions are external factors, namely regulations on the implementation of postal activities, competitors, socio-economic conditions of the community, and the role of PT.Pos Indonesia, and customers.The other eight dimensions are internal factors, such as human resources, finance, service operations, marketing, physical counters, cooperation, technology used, and other activities.However, after the thirteen dimensions were tested using smartPLS, there were some dimensions that had to be eliminated because the loading factor value did not match the standard.
In the first PLS running results, it can be seen in Table 6 that there are loading factors with minus values, namely REG1 (-0.950),REG2 (-0.946),REG3 (-0.939) on the first order, and SOE-X13 (-0.506) and  on the second order.A minus value indicates that the item/indicator is negatively related to the variable, so it must be eliminated from the research model.In addition, very small loading factors (< 0.5) such as KSM-X26 (0.448), SDM-X21 (0.432), and REG-X11 (0.410) in the second order must also be eliminated because the construct/variable can only account for less than 50% of the items/indicators.Some of the reasons that cause low loading factor values include because some statement items in the questionnaire have double/ambiguous meanings (Benitez et al., 2020).
Table 1.First PLS Running Results Source: Processed Primary Data (2023) In addition to being caused by loading factors that are less than standard, the five aspects (dimensions) that are eliminated if analyzed empirically do have less influence on the performance of Postal Agents.The first explanation is related to the dimensions of social and economic conditions around the postal agent counter (SOE-X13), many think that if the postal agent counter is located in an elite area or the center of the crowd, the amount of income will increase because the number of customers present is increasing.In fact, what determines the Postal Agency income is having customers who regularly send packages with large volumes or destinations with expensive shipping.Usually, customers like this do not always live in elite areas or crowded centers, but can also be in areas that enter narrow alleys.Therefore, one way to get customers like that, the Postal Agent must be active in picking up the ball, not waiting for customers to come to the counter.

Indicator
Furthermore, the second is the dimension of the physical condition of the postal agent counter (FLK-X25).Not always an agent counter that is large, has a large parking, and is located on the side of the highway will guarantee a large income.In Sidoarjo itself there are several Postal Agents in Sidoarjo who have small counters, only use the terrace of the house, and do not have a seat or parking lot, but the income per month can be more than 20 million rupiah.Based on this case, it means that the customer does not really mind this aspect, the delivery of the package is smooth, there are not many conditions, and it arrives quickly.After all, from the perspective of the owner of the agency, renting a large and located place on the side of the road will increase the operating load because the price is quite expensive.
The third dimension that is eliminated is government regulations regarding the implementation of posts when there is an emergency such as the pandemic in 2020 -2022 (REG-X11).Indeed, it has been explained earlier that during the pandemic, logistics delivery lines are clearly disrupted, especially those using air transportation.The existence of the PSBB makes packages that were originally sent by plane undergo adjustments by land and sea.This makes the delivery travel time delayed from the schedule that should be.Not a few complaints were received by the Postal Agency about this at the beginning of the pandemic season, but after people began to adapt to the limited circumstances at that time, they began to understand if there was a delay.For example, usually delivery can only take one day to two days, then what is usually three days to five days.Even for some overseas parcel deliveries, there are some that are suspended for uncertain time.
The fourth dimension that needs to be eliminated is cooperation (KSM-X26) with other expeditions or fellow Postal Agents.This cooperation, when viewed from the surface, is quite profitable at first glance.Please note that in general, if there is cooperation like this, the Postal Agent and the third party will share the benefits.Competition is quite tight in the expedition world making it almost impossible for any expedition service agent who provide customers/share transaction volumes for free.Even if there is something to give, the volume shared is only small.Therefore, if the volume of packages that come from the cooperation volume is not much, the profit obtained by the Postal Agent is not much too.
Then the fifth dimension that is eliminated is human resources (HR-X21).Actually, in all companies, this aspect is the main one, but the indicators written for the criteria of a good counter manager/clerk are, friendly when greeted by visitors, responsive when handling complaints, and empathy with customers.In fact, to lure customers into the courier business world, the HR capabilities needed are more than that.Based on the experience of agents who have succeeded, the special characteristics of their human resources include actively picking up packages to the customer's home/warehouse/store/office, quickly adapting to the situation, being clever in setting counter opening hours, and helping to ensure that the customer's package/mail has arrived at the destination address without damage.
After the first elimination, proceed again to run the second PLS.The results of this second run can be seen in Table 7, the loading factor of the indicator in the first order has met the standard of all > 0.7.However, there is a loading factor in the second order that is less than 0.7 and more than 0.5, namely from the dimensions of MKT-X24 and KEU-X22 with the same value of 0.540.The researchers chose to eliminate MKT-24 because its AVE value was smaller (0.859) than KEU-X22's AVE value of 0.893.A smaller AVE value indicates that the mean of a construct/variable explains less variance of the indicator.Source: Processed Primary Data (2023) The sixth elimination process in the marketing dimension must be carried out, not only to meet the requirements of loading factor values of 0.7 and AVE 0.5 but also there are several other considerations.It is clear for some types of businesses, marketing is needed to introduce a company/store to the public.Whether it's digitally, through social media platforms, or the installation of large billboards on the roadside.In fact, in practice, some postal agents in Sidoarjo, who are classified as successful, do not have roadside signage to show the location of agents.Even the existence of an agent counter on Google Maps is just a formality and there is no significant digital optimization like Google Ads.One important thing that a courier business does is to dare to pit directly to prospective customers.Make sure the customer provides most of the volume of his package using Pos Indonesia services.If it is successful, don't forget to maintain customer loyalty by maximizing services such as giving special discounts.
Running the third PLS, after removing the MKT-X24 dimensions, it was found that the loading factor on the second order was still less than the standard 0.7, namely KEU-X22 (0.622), TEK-X27 (0.673), and VAP-X28 (0.605).As can be seen in Table 8, the AVE value of the variable X2 (0.423) in the second order is also still less than the standard 0.5, so it is necessary to eliminate it again to get a more appropriate loading factor and AVE value.Based on the data, the eliminated dimension is VAP-X28 because it has the smallest value compared to other dimensions.Source: Processed Primary Data (2023) Then the seventh elimination is another dimension of investment outside the Postal Agency (VAP-X28).The standards provid by Pos Indonesia for partners such as Postal Agents, in addition to providing package/mail delivery services, they must also provide postal objects such as seals, stamps, and postcards.As for services that are not linear with the courier world, agents are equipped with applications for bill payment channelsmonthly household bills such as water, electricity, telephone, gas, BPJS, and PBB.There is a possibility that this variation in practice still does not attract market interest because the function has been unable to compete with minimarkets such as Indomaret/Alfamart.Therefore, other new variations are needed that can reach more markets so that they can help agents increase their revenue.
The next run for the fourth time, showed better results than before.Presented in Table 9, the loading factors that do not match are only two, namely from the dimensions of TEK-X27 (0.572) and KEU-X22 (0.696).The AVE value of the internal factor (X2) has increased to 0.471, but it is still less than the standard.While the loading factor of the firstorder and second-order external factor variable (X1), is in accordance with the standard.Therefore, the dimension with the smallest value of TEK-X27 (0.572) will also be eliminated to increase the AVE value and loading factor.Source: Processed Primary Data (2023) The final elimination is the technology used by the Postal Agency (TEK).When compared to existing systems in competitors, maybe the postal system is far behind, even though everything is online.On the bright side, with this basic application system, Postal Agents do not need to prepare expensive devices for operations.Enough computers/laptops at least Windows 7, ordinary printers, internet networks, and Android mobile phones, all can run.So there is no need for technology that is too sophisticated to run a Postal Agent, the important thing is to have a connecting media such as WhatsApp to communicate with customers and prospective customers.In the fifth PLS run, finally obtained results according to standards, loading factor > 0.7 and AVE > 0.5 in all orders.You can see Table 10, the remaining 3 variables, 5 dimensions 18 indicators that survive.Source: Processed Primary Data (2023) There are three dimensions of external factor variables, namely competitors (PES-X12), and the role of PT.Pos Indonesia (PPI-X14), and Customers (PEL-X15).The dimensions of PES-X12 consist of three indicators, including the population of other expedition services (PES1), the population of expedition services among Pos Indonesia partners (PES2), and the price of competitor shipping costs (PES3).The dimensions of PPI-X14 consist of four indicators, including PT. Pos Indonesia is the main infrastructure provider of the Postal Agent (PPI1), PT.Pos Indonesia is the controller of the activities of all its partners (PPI2), PT.Pos Indonesia is the promoter of the Postal Agent (PPI3), and PT.Pos Indonesia is an intermediary for cooperation with the marketplace (PPI4).The dimensions of the PEL-X15 consist of two indicators, including the type of Postal Agent customer (PEL1) and customer preferences in choosing expedition services (PEL2).Furthermore, there are also two internal factor variables that persist, namely finance (KEU-X22) and Products and Services (PUJ-X23).The KEU dimension consists of three indicators, including bookkeeping (KEU1), routine financial evaluation (KEU2), and monthly budget (KEU3).While the PUJ dimension consists of four indicators, including pickup service (PUJ1), tracking service (PUJ2), completeness of shipment protection (PUJ3), and operating hours (PUJ4).With the existing dimensions and indicators, the latest research model can be illustrated in figure 6.In a two-level study, to evaluate the validity of the measuring instrument must be done twice.In the first order level, the value seen is the loading factor between the dimensions and indicators as shown in Table 10.While at the second level / second order, the value seen is the loading factor between variables and dimensions.It can be seen that the numbers presented in the table are in accordance with the standard  0.7.This indicates that these variables are quite good at explaining the indicators.
From the same source, Table 10, the value of Average Variance Extracted (AVE) is level one and level two.At level one, the AVE generated by the competitor dimension is 0.837, the role dimension of PT.Pos Indonesia 0.844, customer dimensions 0.871, financial dimensions 0.894, product and services dimensions 0.863, and performance variables 0.856.At level two, the AVE produced by the external factor variable is 0.593 and the internal factor is 0.607.Based on the reference set, all AVE values of each variable are in accordance with the standard  0.5 (Hair Jr. et al., 2017).This shows that latent variables in the research model are able to explain more than half or 50% of the variance in the indicators.
Furthermore, the third value seen when evaluating data validity is the discriminant validity value.By looking at this value, it can be known to what extent a variable is completely different from other variables.The approach to the first order to get the discriminant validity value is to compare the loading value of an indicator against its variables with the loading value of the indicator against other variables often called crossloading.As can be seen in Table 11, the loading value of each indicator in its latent variable at level one is always greater than the value in the other latent variable.For example, the loading value of one of the external indicators of PES1 in its latent variable (PES-X12 dimension) is 0.917.This number is greater than the value of loading PES1 on other latent variables, namely KIN-Y (-0.061),PPI-X14 (0.571), PEL-X15 (0.453), KEU-X22 (-0.009), and PUJ-X23 (-0.213).Similarly, the loading value of one of the internal indicators KEU1 on its latent variable (KEU-X22 dimension) is 0.954.This number is also greater than the loading value of KEU1 on other latent variables, namely KIN-Y (0.378), PES-X12 (-0.127),PPI-X14 (0.365), PEL-X15 (0.082), and PUJ-X23 (0,367).Source: Processed Primary Data (2023) In the previous approach, cross-loading could only display discriminant validity values at the indicator level.Therefore, to display values at the variable level, another approach, Fornell-Larcker, is used.The basic principle of Fornell-Larcker is to compare the AVE root of a latent variable with the correlation value between that latent variable and other variables.It can be seen in Table 12, that the yellow part of the table shows the value of discriminant validity between variables at the first-order level.While the blue part shows the discriminant validity value between variables in the second order.Taking data from Table 10 that has AVE values of external and internal factors, the root value of AVE (0.593) of the external factor EKS-X1 is 0.770 The number is the same as presented in the Fornell-Larcker table and is greater than the correlation value with the internal factor INT-X2 which is 0.126.Similarly, the root AVE (0.607) internal factor of INT-X2 is 0.779.The figure is also greater than the value of its correlation with external factors EKS-X1.Based on the values in the cross-loading table and Fornell Lrocker, the discriminant validity criteria in this research data have been met.After the research data has been confirmed to be valid, the next evaluation is the reliability of the measuring instrument/questionnaire.Table 7. Fornell-Larcker Value on the Outer Model.

Evaluation of Reliability of Measuring Instruments
To evaluate the reliability of a measuring instrument/questionnaire, Cronbach's Alpha standard is used with a value of  0.7.As seen in Table 13, it can be seen Cronbach's Alpha value on latent variables at level one/first order such as competitor (PES-X12), role PT.Pos Indonesia (PPI-X14), customers (PEL-X15), finance (KEU-X22), products and services (PUJ-X23), and performance (KIN-Y) have exceeded 0.7.The same is true of latent variables at level two/second order, Cronchbach's Alpha value at external factors is 0.912 (EKS-X1) and internal factors are 0.891 (INT-X2).If the data has been declared valid and the measuring instrument/questionnaire is reliable enough, it can be continued with the inner model test.This test is carried out to determine the direct influence between latent variables and determine whether or not the research hypothesis is accepted using bootstrap analysis.The first hypothesis of the study is to test whether external factors have a positive effect on the performance of the Postal Agent organization in Sidoarjo.Based on the test results in Table 14, the value of the coefficient (original) of external factors to the performance of the Postal Agent is 0.027 and T-statistics is 0.123.With the value listed, it can be stated that the T-statistic is not significant because it is less than the T-table 2.06 and the P-value is greater than 0.05, so the first hypothesis is rejected.This confirms that external factors are not proven to have an influence on the performance of Postal Agents in Sidoarjo.
The second hypothesis of the study is to test whether internal factors positively affect the performance of the Postal Agent organization in Sidoarjo.Based on the test results in the table below, the value of the coefficient (original) of internal factors on the performance of Postal Agents is 0.511 and T-statistic is 3.115.With the value listed, it can be stated that T-statistically significant because it is more than T-table 2.06 and P-value less than 0.05, so the second hypothesis is accepted.This confirms that internal factors are proven to have an influence on the performance of Postal Agents in Sidoarjo.Source: Processed Primary Data (2023) Furthermore, in testing the inner model, there is an R-square value that needs to be considered to determine the variability of endogenous variables that can be explained by exogenous variables.As shown in Table 15, the adjusted R-square value is 0.210.It can be explained that exogenous constructs (X1 and X2) simultaneously affect Y by 21%.If interpreted based on criteria (Pham, 2010), the value falls into the weak category because the nominal is less than 33%.
Table 10.R-square value on inner model testing Source: Processed Primary Data (2023) The next value to note in inner model testing is the relevance of predictions (Q2) using the Blindfolding procedure to validate the research model.The Q2 result is said to be good if the value is > 0 which indicates exogenous construct variables have predictive relevance for endogenous construct variables.As can be seen in Table 16, the Q2 value of the research endogenous variable (Y-KIN) is 0.092.The value is more than 0.02, but less than 0.15, so it can be said that the exogenous latent variable as an explanatory variable is able to predict endogenous variables in small levels.Based on the results of outer and inner model testing, it was found that internal factors had a positive and significant effect on the performance of Postal Agents in Sidoarjo (P-Value 0.001 > 0.05).There are two dimensions described by internal factors, namely products and services, and finance.The loading factor value of both is quite high, for products and services at 0.873, while for finance at 0.779.
It has become a secret recipe in the business world if you want to win in the competition the business entity must have added value to the products and services offered, the same applies to the Postal Agent.There are four added values commonly applied by Postal Agents who have been operating for more than three years, namely pickup customers, help track, provide protective accessories, and open counters for more than 9 hours consistently.These four things are not something that the Postal Agency must do when signing an initial agreement with PT.Pos Indonesia.However, based on years of experience in the courier business world, the four added values are applied to make it easier for customers when sending packages/letters.Not only that, but customers are also able to monitor the actual location of packages, ensure the physical security of packages until they arrive at their destination, and use the services of Postal Agents at any time.That way customers will be comfortable and more loyal to the Postal Agent, even voluntarily referring the services of related agents to people around them.From the Postal Agent's own side, the agent's income will increase because it gets new customers or at least stable because old customers still use its delivery services.This is in accordance with research conducted by previous research (Rahmani, 2022;Sanjaya, 2018) that customer convenience will increase customer loyalty.If customer loyalty increases, then the company's revenue will also increase (Dameria, 2016;Panjaitan & Siregar, 2021).
In addition to these four things, there is one other added value that has been applied by most Postal Agents who have been operating for more than years, namely the variety of products and services offered.Some Postal Agents in addition to providing delivery of packages/letters using the flag of Pos Indonesia, they also accept delivery of packages/letters using the flag of other expedition services such as Wahana and Lion Parcel to reap the market for large packages/coli.There are also several Postal Agents who add business at their counters in the form of products and services that are not related to the courier service business such as selling fast food, contemporary drinks, pet equipment, money transfers to all banks in Indonesia, photocopies of documents, studio photos, and receiving payment services that are more than just monthly bills that are currently needed by the public (credit installments, train tickets, airline tickets, e-commerce, and e-money top-ups).The operation of other products and services can be run by maximizing tangible and intangible assets at the agent counter because most of the needs of the business still intersect with the needs of the Postal Agent.All of these side businesses are carried out as an effort to back up revenue if one day the Postal Agent courier service business is less profitable, there is still income from other sources (Carter et al., 2004).Based on the experience that has been passed by agents, not always the income obtained by these Postal Agents is stable.There is a time when customers switch to using other expeditions because there is a shipping cost promo so that the price becomes cheaper.Unreliable, the volume of packages sent using Pos Indonesia services will decrease and will have an impact on decreased income.With cross-subsidies from other sources of business income as previously explained, at least it can help ease the burden of operational costs.In addition, by providing many variants of shipping services, customers have the freedom to choose expeditions according to the desired criteria, which will increase customer convenience when transacting with the Postal Agent (Anjani, 2021).
The next dimension is finance.In general, the owner of this Postal Agent has limited capital and comes from personal funds.Therefore, daily capital turnover is very important to monitor.Long-standing Postal Agents always have books to monitor the daily cash inflow (cashflow) of Postal Agents.In addition, monthly budgets and regular budget evaluations can monitor agent expenditures so as not to over-budget and financial conditions remain healthy.This is similar to the results of research (Wijewardana, 2018) which states that finance in MSMEs must always be maintained so that the business continues to run even with limited capital.The Influence of External Factors on the Performance of Sidoarjo Postal Agents On the contrary, what happened to external factors, this factor did not prove to have a positive effect on the performance of Postal Agents in Sidoarjo (P-Value 0.901 > 0.05).There are three dimensions described by external factors namely competitors, the role of PT.Pos Indonesia and customers with each have a good loading factor value against external factors, namely 0.854, 0.877, and 0.733.The influence of external factors has no effect on the performance of Postal Agents in Sidoarjo is supported by research (Islami et al., 2018) which states that internal factors have a greater positive influence than external factors.This is because the internal factors tested (operating cost savings, improving the quality of products/services, HR management, and the use of technology) can improve the company's image in the eyes of customers.The better the company's image, the more customers will make it the first choice when compared to competitors.In addition, external factors that basically cannot be controlled by the company, require great efforts to make changes.Based on research (Alkhateeb, 2019), it is enough just to maximize the function of internal factors that can be controlled, then company performance can also increase.
Reflecting on the direct experience of Sidoarjo Postal Agents, external factors such as customers, the role of Indonesian Post, and competitors do not have much influence on the performance of Postal Agents.This is shown by the existence of twenty-five Postal Agents who managed to survive more than three years.They are agents who have ability to maximize their internal strengths, such as providing maximum service towards customers from picking up packages of any amount, free package repacking, operating hours of more than 9 hours from morning to night for 6 days a week, tracking until the package is confirmed to its destination.Even to maintain customer loyalty, these postal agents also do not hesitate to provide other services such as opening money transfer services, paying water bills, electricity, and telephone to open expedition service franchises other than Pos Indonesia.It can be seen here, that the Postal Agents try not to depend entirely on Pos Indonesia so that they can still move even though competitors are holding massive discounts and later the market reached will be wider.Slowly, these postal agents are also building their own image/branding as a place that accepts all types of shipments and payments.

Conclusion
Based on the description above, it can be concluded that regarding the performance of the Postal Agent in Sidoarjo, namely first, internal factors consisting of financial aspects and service products, have a positive and significant influence on the performance of the Postal Agent.second, external factors consisting of competitor aspects, the role of PT.Pos Indonesia, and its customers, have no influence on the performance of the Postal Agent.

Suggestion
For future research, it is better if researchers can increase the number of samples so that the analysis results are more accurate.If possible, research methods can be combined between quantitative and qualitative.That way, researchers will get a stronger source of data when conducting analysis.The third suggestion.Every time you determine the indicators, you should study the literature in more detail so that a strong research model is obtained and produces questions in the right questionnaire.

Figure 2 .
Figure 2. Top Brand Index 2021 for Courier Service Category services PT.Pos Indonesia from 2019 (IDR 2,913.72 billion) to 2020 (IDR 2,646.70 billion) decreased.In fact, it decreased again in 2021 (IDR 1,992.02 billion) when compared to 2020.This decline in revenue was also reflected in a smaller scope, namely the Postal Agency.Data obtained from the Sidoarjo Central Post Office, overall, the income of Postal Agents in Sidoarjo decreased by around 15% during the pandemic.Initially, the revenue in 2020 was IDR 11.3 billion to IDR 9.6 billion in 2021, as shown in Figure3.If revenue decreases, the profit earned by the Postal Agent from revenue sharing per transaction will also decrease.

Figure 3 .
Figure 3. Income Graph of All Postal Agents in Sidoarjo 2018-2021 (data source: Sidoarjo Central Post Office) Recorded at the Sidoarjo Central Post Office, the total number of Postal Agents in 2019 was 67 units.As time goes by, in 2022 the number of agents decreases to 46 units.The decrease in the number of Postal Agents occurred a lot during the COVID-19 pandemic.Of the 46 surviving agent units, there are 25 units that are agents that have been operating since 2019.While the remaining 21 units are agents who were only established in 2020 and 2021.Based on the observations of researchers, Postal Agents are able to last a long time (more than three years) because they have good company performance, as shown by their average monthly turnover above 15 million rupiahs.Being able to maintain turnover at that number is certainly due to several influencing aspects.There are aspects that come from internal factors such as human resources (HR), finance, service products, marketing, physical counters, cooperation, technology used, and other investments.There are also aspects derived from external factors such as regulations for the implementation of postal activities, competitors, socio-economic conditions of the community, and the role of PT.Pos Indonesia, and customers.Therefore, the author is interested in conducting this research to find out what aspects are found in external and internal factors that most affect the performance of Postal Agents in Sidoarjo.

Figure 4 .
Figure 4. Latest Research ModelIn a two-level study, to evaluate the validity of the measuring instrument must be done twice.In the first order level, the value seen is the loading factor between the dimensions and indicators as shown in Table10.While at the second level / second order, the value seen is the loading factor between variables and dimensions.It can be seen that

Table 2 .
Second PLS Running Results.

Table 3 .
Third PLS Running Results.

Table 4 .
Fourth PLS Running Results

Table 5 .
Fifth PLS Running Results.

Table 6 .
Cross Loading Value on Outer Model.

Table 8 .
Cronbach's Alpha Value On Outer Model

Table 9 .
Results of Hypothesis Test Using Bootstrap Analysis

Table 11 .
Q2 Value Using Blind Folding Procedure