FORECASTING METHOD SINGLE EXPONENTIAL SMOOTHING FOR DEMAND GROCERIES AT THE GUNTUR STORE

: Guntur Store is located on Jl. Sisingamangaraja, Kisaran, Asahan Regency. It focuses on selling basic needs, but the store faces a problem in its sales. The main issue encountered is the uncertainty in forecasting the stock of basic needs, which can lead to imbalance between supply and demand, due to Guntur Store still relying on manual ledger entries for reporting and inputting data into Microsoft Excel. This research aims to predict the demand for basic needs at Guntur Store using the Single Exponential Smoothing Method and to enhance prediction accuracy through the application of time series analysis techniques on historical sales data. The Single Exponential Smoothing Method will be employed to model and forecast patterns of demand for staple goods. The findings of this research can provide insights into future demand for goods, thereby improving supply chain optimization and ensuring adequate stock levels to effectively meet customer demands. The conclusion of this study is to obtain more accurate estimates in anticipating the evolving data patterns of basic needs at Guntur Store, thus providing valuable information for the store and policymakers involved in the distribution of basic needs.


INTRODUCTION
Inventory is one of the most crucial estimates in a company's operations.For companies, inventory is not just an asset, but also a significant value [1].The prese nce of inventory in a company context has major implications, depending on its quantity and availability [2] [3].If the amount of inventory is large enough, this will have an impact on increasing the costs required to maintain and manage the inventory.On the other hand, if inventory is insufficient, the impact can disrupt the production and sales process.Apart from that, the existence of inven tory also has a direct influence on the company's balance sheet and profit and loss statement [4] [5].
Guntur Shop is a shop that sells basic necessities, this shop was founded in 1990.The address is Jl.Sisingama ngaraja No.113, Kisaran Timur, Kec.East Kisaran City, Asahan Regency, North Sumatra 21221.In the context of the retail business, especially in the sale of basic necessities (nine basic commo dities) at the Guntur Store, accuracy in forecasting and managing demand is crucial to maintaining smooth operations and customer satisfaction.However, for the past 2 years, the Guntur store has experienced difficulties in estimating basic food supplies because it only relies on direct observation of stock conditions in the store without data processing.Fluctuating and difficult to predict demand can cause problems in inventory planning, which in turn can result in an imbalance between supply and demand [6][7].
The goods that experienced a drop were rice in June 2022 by 350  This research aims to increase the accuracy of predicting demand for basic necessities at the Guntur Store through the application of the Single Exponential Smoothing Method.By conducting time series analysis of historical sales data, this research is expected to make a significant contribution to improving inventory management efficiency and optimizing the supply chain.Apart from that, this research can also provide valuable insight for industry players in their efforts to improve services to custo mers and avoid inventory imbalances that can be detrimental both from a financial and operational perspective [8], [9].Even though there have been many studies examining demand predictions in the retail sector, the use of the Single Exponential Smoothing Method in the context of specific basic food stores such as the Guntur Store is still relatively minimal [10], [11].Some prediction metho ds that have been developed tend to pay less attention to the unique characteristics of demand for basic goods.
Previous studies discussing fore casting, in research, [12] The Single Exponential Smoothing method can be applied to predict demand for herbal medicines at DS Ria Sari Anggriani, enabling knowledge of the predicted value of demand in the next period.in line with research [13], explaining that by applying the Single Exponential Smoothing method the system created can predict the number of sales of building materials at UD. Karya Mandiri in the July 2022 period.In addition, the Mean Squared Error (MSE) method can be used to evaluate the error level for each alpha value, thereby making it possible to obtain a minimal error value.Therefore, it is necessary to carry out more in-depth research with a focus on implementing the Single Exponential Smoothing method at the Guntur shop by testing the best alpha value [14].By using the Single Exponential Smoothing method, it is hoped that it can produce accurate predictions related to consumer demand to help Guntur Stores in stock planning, marketing strategies and more informational decision making.

Research Stages
This stage it will The research stages and research methodology are described which contain the steps in solving problems at the Toko Guntur

Studying Literature
Search for reference information in the form of books, journals and study materials related to the topic of forecasting discussion.

Data collection
The data collection process was carried out in the form of direct interviews with the shop owner Mr. Ahyar S Pane and observations for data analysis of the process of predicting demand for basic necessities at the Guntur Shop so that the necessary data and information were obtained.Next, the data is analyzed for data processing in the system.

System planning
System design a critical stage in the software development cycle that involves transforming the results of system analysis into a detailed and implementable design.At this stage, the main focus is detailing how the information system will be organized and implemented to meet user needs and achieve business goals.

System Implementation
The software development cycle where the design that has been created is realized into code that can be run.This process involves building an application or system based on previously established specifications and designs.
In this research, the data collected is sales data Groceries at the Guntur Shop from Month December 2021 until November 2023 , which will later be processed using the SES method or desktop-based Single Exponential Smoothing.In application SES method uses formula as following : Ft : Forecast value for period t.Ft-1 : Estimated forecast for the previous period.At-1 : Observation Data for period t -1.α : smoothing constant To measure the forecasting error value, it can be searched using Mean Square Error is an alternative method in a forecasting method.The smaller the MSE value, the more accurate the prediction.

∑( )
Description: Yt : Actual value in period t Ft : Value Forecasting in period t n : Number of data for the forecasting period

RESULTS AND DISCUSSION
Input requirements analysis invol ves collecting and assessing input data that will be processed through calcula tions using the Single Exponential Refinement Method (SES).In the context of this research, basic food sales data from the Guntur Store during the period January 2022 to December 2023 is the main input.The analysis process will lead to understanding sales patterns, trends and fluctuations contained in the data ( table 1).Then the input data will be processed using the Single Exponential Smoothing method, with the formula equation (1).To measure the forecast error value, it can be searched using MSE (Mean Square Error) which is an alternative method in a forecasting method.The smaller the MSE value, the more accurate the prediction.
For calculating rice demand using the Single Exponential Smoothing method with Alpha 0.      Image 1 shows the display of the owner's main menu page.In this system, there are several active menus on the main menu page, namely the forecast calculation menu, change password menu and logout.

Image 2. Display the Forecast Calculation Form
Display of the forecast calcula tion form that will appear after the user selects the forecasting period and alpha.The forecast calculation form is equipped with a print forecast results button.

CONCLUSION
Forecasting system using the single exponential smoothing method for basic food demand at the Guntur Store provides benefits for the store .With this system, forecasting can be more accurate request groceries .The single exponential smoothing (SES) forecasting method is an important method in helping Guntur Shop to predict the number of requests basic necessities in the coming period.This method can be used as evaluation material and help in making more appropriate decisions regarding inventory goods For for sale.
Research methodology refers to the systematic steps used in a study to collect, analyze and interpret data .The methodology contains a research frame work to explain the flow of the problem, analyze the problem, determine the objec tives of the research.The stages in the research method carried out are depicted in image 1.

Table 1 .
Sales Data Groceries on the Moon January 2022 until December 2023

Table 2 .
Rice Demand CalculationTable with Single Exponential Smoothing Method with Alpha 0.9.

Table 3 .
Oil Demand Calculation Table with Alpha 0.3.

Table 4 .
Sugar Demand Calculation Table with Alpha 0.6.

Table 5 .
Egg Demand Calculation Table with Alpha 0.6.

Table 6 .
Calculation Table for Butter Demand with Alpha 0.6.

Table 7 .
Salt Demand Calculation Table with Alpha 0.5.

Table 8 .
Milk Demand Calculation Table with Alpha 0.3.
Image 1. Owner's Main Menu Page View