Calculation by average (SMA), or simple moving average
Posted: Sun Dec 22, 2024 8:53 am
It is considered the easiest, but at the same time the most common method of demand forecasting, used in many companies. Simple moving average (SMA) is calculated by the formula:
Forecast (t + 1) = (1 / (T + 1)) x [Sales (t) + Sales (t – 1) +...+ Sales (t – T)].
To determine the amount of demand using this method, you need to:
Select the window width – T. This value shows the period for which sales will be averaged. If we are talking about daily demand, then for the last two or three days, seven days, etc., and if philippines whatsapp we are managing demand by months, then for the last such periods.
To calculate the forecast for the upcoming period, you need to take the average for the selected window width. For example, we forecast sales on the tenth day. The window width value is five pieces, respectively, we take the average value for the last five days. We received the sales value for a new day - again we take the average for the last 5 days. This is how the demand forecasting process occurs.
The sales values can be taken either in a row, as they were in the previous period, or at random for a month, year, etc. Here, a flexible approach is applied to the calculation interval of the data that is taken to obtain the average value.
Below is an example of sales forecasting in Excel taking into account seasonality.
Complex Sales Forecasting Methods
There is a sales series on the basis of which a forecast is needed. All sales are broken down by month, and, for example, we want to see the predicted values for monthly periods. What should we do? First of all, select the window width, for which we find the average value for the last two, three, four and 10 months. If we take it equal to two, and the number of sales in November and December was 15 and 40, then the January forecast will be 27.5, and February - 40.
There is a trend - the wider the window, the closer the indicators become to the calculation based on average values for the entire period. This can be clearly seen on the graph, where the blue line shows real sales, and all the rest are forecasts with different window widths.
complex methods of sales forecasting_graph
This method is recommended to be applied only to those goods that are characterized by good and stable sales with minimal fluctuations. In all other cases, this method will give a large error and will be ineffective from the point of view of warehouse stock management, as it can lead to their deficit or surplus. Therefore, instead of this method, they began to use calculations based on the weighted average, which are also characterized by some features.
Forecast (t + 1) = (1 / (T + 1)) x [Sales (t) + Sales (t – 1) +...+ Sales (t – T)].
To determine the amount of demand using this method, you need to:
Select the window width – T. This value shows the period for which sales will be averaged. If we are talking about daily demand, then for the last two or three days, seven days, etc., and if philippines whatsapp we are managing demand by months, then for the last such periods.
To calculate the forecast for the upcoming period, you need to take the average for the selected window width. For example, we forecast sales on the tenth day. The window width value is five pieces, respectively, we take the average value for the last five days. We received the sales value for a new day - again we take the average for the last 5 days. This is how the demand forecasting process occurs.
The sales values can be taken either in a row, as they were in the previous period, or at random for a month, year, etc. Here, a flexible approach is applied to the calculation interval of the data that is taken to obtain the average value.
Below is an example of sales forecasting in Excel taking into account seasonality.
Complex Sales Forecasting Methods
There is a sales series on the basis of which a forecast is needed. All sales are broken down by month, and, for example, we want to see the predicted values for monthly periods. What should we do? First of all, select the window width, for which we find the average value for the last two, three, four and 10 months. If we take it equal to two, and the number of sales in November and December was 15 and 40, then the January forecast will be 27.5, and February - 40.
There is a trend - the wider the window, the closer the indicators become to the calculation based on average values for the entire period. This can be clearly seen on the graph, where the blue line shows real sales, and all the rest are forecasts with different window widths.
complex methods of sales forecasting_graph
This method is recommended to be applied only to those goods that are characterized by good and stable sales with minimal fluctuations. In all other cases, this method will give a large error and will be ineffective from the point of view of warehouse stock management, as it can lead to their deficit or surplus. Therefore, instead of this method, they began to use calculations based on the weighted average, which are also characterized by some features.