Sales Forecast in Excel or Google Sheets with our Free Template
Posted: Thu Feb 20, 2025 6:55 am
There are several time series based formulas you can use for this purpose, such as the "naive method", simple moving average, weighted moving average, and exponential smoothing:
Naive method : Just assume that sales will be the same next month as they were this month or last month.
Simple moving average : For example, to calculate sales forecasts for the next month, we take the average of the sales results for the previous x months.
Weighted Moving Average : For example, to calculate afghanistan cell phone number list sales forecasts for the next month, you need to add up the (monthly sales results * weighting factor) for the previous x months. Typically, this weighting factor is higher for recent months and gradually decreases for less recent months.
Exponential smoothing : This is usually a weighted moving average with a weighting factor that decreases exponentially. However, there are more complex versions that can also handle trends or seasonality (just google “double exponential smoothing” and “triple exponential smoothing” – it can get a bit more complex).
In short: This sales forecasting method uses an extrapolation based on historical results that sounds more fanciful than it actually is.
Regression analysis
This sales forecasting method does not use only time extrapolation.
It uses a mathematical equation to model the relationship between a dependent variable (sales) and one or more independent variables (such as advertising expenditures or economic indicators). The basic formula for a linear regression is:
y = a + bx
where y is the dependent variable (sales), a is the intercept, x is the independent variable (advertising expenditures), and b is the coefficient representing the change in y for a one-unit change in x.
This type of forecasting method is useful if you have identified factors that accurately predict your sales results and you can accurately predict them yourself in the future. In addition, you need someone who is well versed in statistics to build the forecasting model. Otherwise, you are unlikely to get reliable results. Or with an emoji:
Naive method : Just assume that sales will be the same next month as they were this month or last month.
Simple moving average : For example, to calculate sales forecasts for the next month, we take the average of the sales results for the previous x months.
Weighted Moving Average : For example, to calculate afghanistan cell phone number list sales forecasts for the next month, you need to add up the (monthly sales results * weighting factor) for the previous x months. Typically, this weighting factor is higher for recent months and gradually decreases for less recent months.
Exponential smoothing : This is usually a weighted moving average with a weighting factor that decreases exponentially. However, there are more complex versions that can also handle trends or seasonality (just google “double exponential smoothing” and “triple exponential smoothing” – it can get a bit more complex).
In short: This sales forecasting method uses an extrapolation based on historical results that sounds more fanciful than it actually is.
Regression analysis
This sales forecasting method does not use only time extrapolation.
It uses a mathematical equation to model the relationship between a dependent variable (sales) and one or more independent variables (such as advertising expenditures or economic indicators). The basic formula for a linear regression is:
y = a + bx
where y is the dependent variable (sales), a is the intercept, x is the independent variable (advertising expenditures), and b is the coefficient representing the change in y for a one-unit change in x.
This type of forecasting method is useful if you have identified factors that accurately predict your sales results and you can accurately predict them yourself in the future. In addition, you need someone who is well versed in statistics to build the forecasting model. Otherwise, you are unlikely to get reliable results. Or with an emoji: