Efficiency Distribution Analysis with Data Envelopment Analysis ( DEA ) Approach

Distribution channels have an important meaning for achieving company success in the field of marketing so that company management is required to always be responsive and able to adapt to environmental changes. Input and output data processing is done by giving weights to the input and output using the DEA CCR primal model by maximizing the input-orientedbased objective function. The results of processing the efficiency scale show the relative efficiency level of the scale of each DMU in the company. The efficiency scale is obtained through the formulation of the DEA CCR primal model between each DMU input and output. If the DMU gets an input and output efficiency value of less than 100%, then the DMU is said to be relatively inefficient. Meanwhile, if the efficient value is equal to 100%, then the DMU is said to be relatively efficient. Of the 5 distribution cities, Sidoarjo, PaluKendari, Bandung and Lamongan which were analyzed, there were 2 cities that were inefficient or experiencing waste in their input and output variables, so the company needed to reorganize the level of use of inputs and outputs it achieved and utilize them optimally to get output that is optimal. targeted.


INTRODUCTION
One of the obstacles that occur in product delivery, especially fertilizer products, is the distribution channel. This distribution channel has an important meaning in the field of marketing. Goods will reach consumers through distribution channels, either direct or indirect. The distribution channel is a set of organizational participants performed by all functions and is needed to deliver the product/service from the seller to the final buyer. For companies, competition can be an opportunity to develop a company or can be a threat to the company. Therefore, company management is required to always be responsive and able to adapt to continuous environmental changes so that they can survive in competitive competition, including developing effective distribution strategies and appropriate steps. This study aims to determine the efficient level of distribution channels in the company and to determine the inefficient distribution channels of the distribution area.

METHOD
Data Envelopment Analysis (DEA) is a tool used to evaluate and improve the performance of a manufacturing or service. DEA is widely applied in performance evaluation and benchmarking in educational institutions, hospitals, bank branches, production plans and others. The units used in the DEA are referred to as DMUs. This technique can be used to find out how efficiently a DMU is used with the utilization of existing equipment to produce maximum output. In the DEA model used, it is known as the Charnes, Cooper and Rhodes ratio (CCR) and is a non-linear equation. In determining the number of distributors, producers are faced with two alternatives as proposed by BasuSwastha (Swastha, 1990) as follows: a. Intensive Distribution, which can be done by producers by selling conventional goods. Manufacturers try to use dealers, especially retailers as much as possible to approach and reach consumers. This efficiency ratio is more widely used when a unit or process has one input or one output. In fact, a process or organizational unit has various inputs and various outputs (incommensurate). To overcome this, Relative Efficiency is used, namely the efficiency of an object is measured relative to the efficiency of similar objects with the notation used as follows This method does not require a production function and the result of the calculation is called the relative efficiency value. DEA is a multifactor analysis method to measure the efficiency and effectiveness of a group of homogeneous Decision Making Units (DMU). Efficiency Score for multiple outputs and inputs can be determined as follows: = total output weight total input weight In this study, we will measure the efficiency value of 5 cities for 6 months using a non-parametric approach. The efficiency score of this study was obtained from the results of the calculation process using the WinDEA software which is the relative efficiency score between each DMU in the object of research. This software gives a score of 0-1 which is then converted into a percentage of 0-100% for each DMU. This study uses an input-oriented approach to see how much input can be reduced so that the DMU becomes efficient. In addition to showing the efficient score, the WinDEA software will also show the target value. The target value is the value suggested by the DEA calculation to make the company more efficient.

RESULTS AND DISCUSSION
The DEA input variable uses the details of the number of shipments and the amount of distribution costs. To find out or calculate the efficiency value on the input variable in the WinDEA software, the only data that can be processed is monthly data. Variables in calculating output at DEA using sales data from distributors, sales from customers, profits from distributors, and profits from customers, as well as input data in calculating output data also taken 5 samples of distribution cities, namely Sidoarjo, Palu, Bandung, Kendari, Lamongan. The DEA method uses WinDEAP software, the calculation uses CRS which is oriented to the input approach. WinDEAP is a software or tool to calculate the efficiency level of the DMU. This software will give a score of 1.00 if the DMU's performance is efficient and vice versa if the DMU's performance is not efficient then the score is less than 1.00. Based on the results of efficiency calculations using DEA, the efficiency level of 5 cities can be seen in the following table: Table 3. Distribution Efficiency Level The results of processing using WinDEAP software, there are only 2 cities that are inefficient for 6 months, namely Palu and Lamongan, both are inefficient in February and less than 1.00..Palu City has an efficiency score of 0.800 and Lamongan city has an efficiency score of 0.965. While the city of Sidoarjo, Bandung, Kendari has a score of 1.00 out of 6 consecutive months. The following can be seen the level of waste or inefficiency in each city of Palu and Lamongan based on the input and output variables of the company:  The results of data processing using WinDEAP software show that for the city of Palu, the company has experienced corporate waste for 6 months. In January from the input variable, the number of shipments for the original value and projected value is efficient because they are the same. However, in the second input variable, namely production costs, the original value is Rp. 2,450,000 higher than the projected value which is only Rp. 760,000. So to get a more efficient value must be reduced by Rp. 1,690,000.
While the output variables of distributor sales, consumer sales, profits from distributors, profits from consumers in January show that the original value and projected value are the same, so it can be said to be efficient.
In February, the results of data processing showed inefficiency, as seen from the input variable for distribution costs and the input variable for the number of shipments. The distribution cost variable shows the original value of Rp. 1,275000. bigger than the projected value which is only Rp. 1.122.000. So in order to be efficient must be reduced by Rp. 153,000.
Meanwhile, in the second input variable, namely the number of shipments, the original value of 1020 Kg is greater than the projected value which is only 897,600 Kg. If the company wants to be more efficient in managing the number of shipments, it must be reduced by 122,400 Kg.
For the distributor's sales output variable and the distributor's profit, it can be seen that the original value and projected value have the same value, this can be said to be efficient. However, the consumer sales output variable and consumer profit are inefficient, it can be seen in the consumer sales variable that the original value is only Rp. 24,360,000. smaller than the projected value of Rp.25,092,000. so to get a more efficient value must be added by Rp. 1,690,000.
The second inefficient output variable is the profit from the consumer which has an original value of Rp. 8.160,000. While the projected value is Rp. 9,792,000. The original value of the consumer's profit is still less than the projected value, so to get an efficient value, it must increase the profit of the consumer by Rp.1.632.000 Likewise in the following inefficiency months, the problems and solutions are almost the same as in January and February.