Sales of bags every year experience a rapid growth, especially for types of school bags, therefore many ofthe bag shops that not only sell one type of school bag, but a variety of school bags, such as those that areoffered in Paradise Stores. Firdaus bag shop sells 17 types of school bags such as, ripper bags, alto bags,character bags etc. The number of bag sales reaches more than 100 pcs per month from all types of bags,but in making sales reports and reporting data the shop owner still uses a manual system. Asking the storeowner to have difficulty adding or subtracting bag data preparation data. The purpose of this research isto study the bag selling system in the web-based paradise Bag Store. The data in this study use 17 types ofbags with the method used in this study is the method of moving average or moving average. Movingaverage is one of the forecasting methods that uses time series data. The period used is period 6.Forecasting errors can be calculated using the MAD and MAPE formulas. The results of this study are abag sales system in the following month. The bag sales data used are January to December 2018 withmoving averages. The result of applying the moving average method is forecasting in January 2019.Forecasting the most bag sales in January 2019 is the sale of large ripper bags with a total of 127 pcs.Forecasting The level of accuracy is generated using MAD and MAPE. MAD for alto palo bags is 5.83while MAPE for alto palo bags is 58.02%.Keywords: Sales, Moving Average, Forecasting Error