Schreibfeder's method, or the weighted average method
Posted: Sun Dec 22, 2024 8:53 am
It differs from the previous one in that instead of average sales values, it has different weights by month. What you need to do to determine the forecast:
It is necessary to determine how many sales were made per working day in the past period. If there were weekends and holidays, they are not counted.
Then you need to choose a weighting philippines whatsapp system for the previous months. So, some data will be of primary importance for the calculations, and some - of secondary importance.
Next, a forecast is calculated for one working day of the future period (month), based on the values of past sales and weights.
Finally, the value for the coming month is determined based on the number of working days in it.
Below is an example of sales forecasting:
Monthly consumption Number of working days Consumption per working day
November 560 28 20
October 310 31 10
September 450 30 15
August 372 31 12
July 310 31 10
The first column shows the initial data for each month, which we need to use to build a forecast for December. November sales were 560 units over 28 working days. This means that the average consumption per working day is 20 units. Once we have determined the consumption for each month, the weighting system comes into play. In our case, we use a model that is characterized by a strong influence of recent sales on the forecast.
Monthly consumption Number of working days Consumption per working day
November 20 3 60
October 10 2.5 25
September 15 2 30
August 12 1.5 18
July 10 1 10
10 143
The weight in September will be 2, in November – 3, etc. The maximum indicator is for the month that is closest to the one for which the calculation is being made. In our case, this is November. We need to multiply each month by its weight – 20 x 3 = 60. Then we find the sum of monthly consumption multiplied by the weight, which will be 143. In this case, the sum of the weights is 10.
The December forecast will be 143/10 x 28 = 400 units, where 28 is the number of working days.
In his book "Efficient Inventory Management" Schreibfeder offers a large number of scale schemes that are used for various types of goods. For example:
A simple six- or three-month average, which is a calculation of the average with a window of 6. In other words, you take a six-month period and calculate the average.
For goods that are not consistently sold during the year, a seasonal average, simple or weighted, is used. So, if you need to build a forecast for December, you need to take data for the last winter or a seasonal weighted average.
Weighted average coefficients. According to our example, these are 3, 2, 5, and decreasing.
That is, the choice is offered to use some supposed set of weights, or the organization can develop its own scheme. This forecasting method is characterized by the presence of a large number of restrictions.
Below is an example of an Excel calculation based on weighted average sales.
Schreibfeder's method
There is a series of sales and the number of working days in each period, it is necessary to determine the probable values. Forecasts are available for five months with weights from three to one and for three months with weights from five to one. First, we divide the sales by the number of working days, then multiply the result by the weight of the month. This number is divided by the sum of the weights, then multiplied by the number of working days. All calculations are easier to do in Excel spreadsheets. Algorithm: a weight formula is selected and forecasts are determined based on it.
It is necessary to determine how many sales were made per working day in the past period. If there were weekends and holidays, they are not counted.
Then you need to choose a weighting philippines whatsapp system for the previous months. So, some data will be of primary importance for the calculations, and some - of secondary importance.
Next, a forecast is calculated for one working day of the future period (month), based on the values of past sales and weights.
Finally, the value for the coming month is determined based on the number of working days in it.
Below is an example of sales forecasting:
Monthly consumption Number of working days Consumption per working day
November 560 28 20
October 310 31 10
September 450 30 15
August 372 31 12
July 310 31 10
The first column shows the initial data for each month, which we need to use to build a forecast for December. November sales were 560 units over 28 working days. This means that the average consumption per working day is 20 units. Once we have determined the consumption for each month, the weighting system comes into play. In our case, we use a model that is characterized by a strong influence of recent sales on the forecast.
Monthly consumption Number of working days Consumption per working day
November 20 3 60
October 10 2.5 25
September 15 2 30
August 12 1.5 18
July 10 1 10
10 143
The weight in September will be 2, in November – 3, etc. The maximum indicator is for the month that is closest to the one for which the calculation is being made. In our case, this is November. We need to multiply each month by its weight – 20 x 3 = 60. Then we find the sum of monthly consumption multiplied by the weight, which will be 143. In this case, the sum of the weights is 10.
The December forecast will be 143/10 x 28 = 400 units, where 28 is the number of working days.
In his book "Efficient Inventory Management" Schreibfeder offers a large number of scale schemes that are used for various types of goods. For example:
A simple six- or three-month average, which is a calculation of the average with a window of 6. In other words, you take a six-month period and calculate the average.
For goods that are not consistently sold during the year, a seasonal average, simple or weighted, is used. So, if you need to build a forecast for December, you need to take data for the last winter or a seasonal weighted average.
Weighted average coefficients. According to our example, these are 3, 2, 5, and decreasing.
That is, the choice is offered to use some supposed set of weights, or the organization can develop its own scheme. This forecasting method is characterized by the presence of a large number of restrictions.
Below is an example of an Excel calculation based on weighted average sales.
Schreibfeder's method
There is a series of sales and the number of working days in each period, it is necessary to determine the probable values. Forecasts are available for five months with weights from three to one and for three months with weights from five to one. First, we divide the sales by the number of working days, then multiply the result by the weight of the month. This number is divided by the sum of the weights, then multiplied by the number of working days. All calculations are easier to do in Excel spreadsheets. Algorithm: a weight formula is selected and forecasts are determined based on it.