FORECASTING SPOKE CAKE SALES AT ZAHRA SHOP WITH DOUBLE EXPONENTIAL SMOOTHING METHOD

: Sales of sponge cakes at Zahra's shop have difficulty meeting customer demand when the demand for sponge cakes increases drastically at certain times. This is because the sales of sponge cakes at Zahra's shop are still calculated based on their own estimates, causing excess production to be made and even a shortage of sponge cake supplies. This problem has an impact on the supply control of sponge cake production which cannot be predicted properly by Zahra's shop. The purpose of this study is to build a forecasting system using the double exponential smoothing method to predict sales of sponge cake at Zahra's shop for the next period based on the data obtained. The research method used in this study is a quantitative method. The results in the calculation of sponge cake sales forecasting using the Double Exponential Smoothing method using alpha 0.5, the number of sponge cake sales for July 2023 on Banana Cake is 430 boxes, with a forecasting error rate of 2.75%, on Chiffon Mocca Sponge cake is 330 boxes, with a forecasting error rate of 2.75%. forecasting error of 4.50%, on Pandan Cake as many as 92 Boxes, with a forecasting error rate of 7.21%, on Ordinary Roll Cake as many as 77 Boxes, with a forecasting error rate of 4.78, So the error rate on Banana Cake, Bolu Chiffon Mocca, Pandan Cake, and Ordinary Roll Cake are in the very good category because they have an error value of less than 10%


INTRODUCTION
Almost every company in the industry is currently facing the challenge of increasingly fierce competition.This requires that every company must be able to plan all production parameters properly and correctly including production capacity, in order to meet market demand in a timely and fast manner.This is done in order to increase company profits and reduce losses.On the other hand, developments in information technology affect data to produce information that is more accurate, fast and actual and relevant to increase the precision or accuracy in decision making.
Sponge cake is a food that is quite popular in Indonesia.The sweet and legit taste makes this cake very popular with all people.Especially when we are having a special celebration or event, sponge cake is often used as the main menu.Besides that, sponge cake is also suitable for serving during celebrations of religious holidays, such as Eid al-Fitr.
The Double Exponential Smoothing method for forecasting sponge cake sales for the next period is a good enough method so that the forecasting results can help Zahra Shop owners prepare sponge cake supplies to be sold in the next period.
Zahra's shop is a small and medium business that is starting to develop at this time.Zahra's shop produces sponge cakes using the best and quality selected ingredients.In the process of developing sponge cake sales at Zahra's shop, they experienced a problem, namely they often had difficulty meeting customer demand when the demand for sponge cake increased drastically at certain times.This happens because the prediction method used by Toko Zahra is still conventional, that is, calculations are still based on their own estimates so that the results of sponge cake production can be excess or even lacking.Besides that, with these problems, the zahra shop must be able to predict or predict how many sponge cakes will be sold and how many must be provided in the coming period.This study only discusses the sales forecasting of Banana Sponge Cake, Chiffon Mocca Cake, Pandan Cake, and Ordinary Roll Cake.
The rationale for the Double Exponential Smoothing method is that the smoothing value will be present before the actual data if there is a trend component in the data.Therefore, for single smoothing values, it is necessary to add double smoothing values to adjust the trend.The Double Exponential Smoothing method that can be used to solve linear trends is the twoparameter method from Holt.In the Holt method the trend values are not smoothed by double smoothing directly, but the process of smoothing the trend is done using different parameters from the parameters used in the original data smoothing.
The Dasni Store Sales Forecasting System uses Double Exponential Smoothing "and the results are in the form of a period from the sales forecasting system for the next 3 months, Using MAPE (Mean Absolute Percentage Error) to calculate the level of forecasting accuracy by looking for minimum errors, because the smaller the error, the forecasting period sales Images then it will be more accurate [ Based on the results of this study, the correct alpha value was used to predict next month's bingke production by using alpha 0.3 and the prediction results obtained from alpha 0.3 were 16,057.68cents next month., and the results of MAPE (The Mean Absolute Percentage Error) is 8.13 with the smallest error, namely very good criteria [9].
The purpose of this research is to design The forecasting system uses the Double Exponential Smoothing method which can be used to predict sponge cake sales at Zahra's shop.So that the owner can use it in determining sales planning and supply of sponge cakes in the following week.As well as to make it easier to know sales predictions and stock predictions that must be provided, stores can use the application system automatically without calculating sales with manual calculations.

METHOD
The framework in this study is a step that will be taken by the author to solve the problem to be compiled.The framework in research requires a clear arrangement at each stage.The framework that the author does can be seen in the image below:

RESULTS AND DISCUSSION
In table 1, the sales data that will be used to forecast sponge cake sales is banana cake.Forecasting is done using the Double Exponential Smoothing method with a value of ∝ = 0.5.The following is the calculation using ∝ = 0.5 Implementation of the system interface in forecasting the number of sponge cake sales using the Double Exponential Smoothing method.

Main Menu Page
The main menu page is the page that will appear after a successful login.The main page will display the dashboard page, forecasting category page, period page, category page, change password page, and logout page.

Forecasting Calculation Process page
The forecasting page is a page that contains a form for entering data for the sponge cake category found at the Zahra Store, the Alpha value for the calculation process and the calculate button.
Image 3. Display Forecasting Calculation Page

Period Page
The period page is a page that contains sponge cake sales data for each the First Smoothing Value: ( ′ )  ′ =∝  + (1−∝)−1 ′ (1) b.Determining the Second Smoothing Value: ( ")  " =∝  ′ + (1−∝)′−1 ′ c.Determining Constant Values ( ) :  = 2 ′ −  " d.Determining the Slope Value () :  = ∝ 1−∝ ( ′ −  ") e. Determining Forecasting Value: + =  +   Where:  ′ : Single exponential smoothed value Xt : The actual value of the t-period " : Double exponential smoothed value ∝ request to t Ft = forecasting results to t n = amount of forecasting data The results of this study are the CPI for the City of Yogyakarta for the period January 2014 to February 2022.The CPI data tends to increase because the data continues to increase until February 2022.However, within a certain period of time, namely January 2020, the CPI value for the City of Yogyakarta has experienced a serious decline due to the Covid-19 pandemic.Furthermore, predictions are made using double exponential smoothing which is then searched for the prediction error value using the Mean Absolute Percentage Error (MAPE) calculation.The CPI forecast value for March 2022 is 109.2964,with a MAPE value of 0.76%.The MAPE value is less than 10%, meaning that the prediction with the double exponential smoothing method can be said to be good[7].GoodsSales Forecasting Information System Using the Double Exponential Smoothing Method at the Sayur Palace.The results of this study are using the mean absolute error as a percentage, the value of broccoli 15.05%, eggs 15.78%, green peppers 12.45%, beans 22.22%, cloves 34.69%, garlic 19.53 %, Tempeh with a value of 20.60% Potatoes with a value of 17.58%.Therefore, Caisim broccoli, chicken eggs, green peppers, garlic and potatoes are classified appropriately because their value is between 10% and 20%.And beans, cloves, tempeh just enough because the value is 20% to 50%[8].

Table 1 .
Sponge cake sales data at Zahra's shop for the period January 2022 to June 2023

Table 4 .
Result of Banana Sponge Sales Forecasting Error Value

Table 5 .
Forecasting Results for All Sponge Cake Categories